## Crevasse rescue

This topic has been archived, and won't accept reply postings.

I'm now trying to get my head around crevasse rescue.

I've spent a whole day on the internet, I knew the theory anyhow,

and I know how to build various hauling systems from big walling and

teaching self rescue, but I still don't have a goto workable system in my head.

Even the Petzl stuff is unclear, or contradictory. If you have the time to give short comments to the following it would be fab. Thanks!

In the following

https://www.petzl.com/INT/en/Sport/Approaching-a-glacier-as-a-roped-team-of-two

it recommends about 20m between the climbers, and the use of knots.

I'm assuming the following - two people, knots in rope. 12m of rope in

knots themselves (6 knots minimum, 2m rope per knot). 20m between the

climbers. Hence ( 60 - 12 -20 ) /2 = 14 rope in coils on each climber. I

understand there is no one size fits all solution as the person might be

injured or unconscious and there are other variables, but most of the

solutions out there seem close to impossible. I have only fallen in a hole

once, and I got myself out.

1. some sources pad and prepare the edge exactly above the climber, (a) how

easy is it to prepare this without pushing ice and snow on top of the

victim? (b) how do you get a sack under a tight rope in snow? (c) is it

just better to go to one side and prepare that area, then use the spare

length of rope to perform the rescue. Some videos show this.

2. many sources show the rope the climber is hanging from as part of the

haul system. EG 3:1, 5:1, 7:1. However assuming the rope has knots that

worked, dragging them back through the snow might be hard, and passing them through the haul system not covered in anything I have bumped into. (I know how to pass these knots, but I doubt everyone does). Hence it seems

again logical to use the spare length of rope. As some sources do.

3. A 2:1 assisted hoist with the spare rope seams by far the best idea. I assume 9 times out of 10 the victim if fully functional? An one would almost always you an assisted hoist on rock if you had the choice. For one it means much less force on the anchor, but also only takes seconds to initiate. And Petzl show

how to do this

But! you only have 12m of spare rope to the victim and

back to the anchor - you don't even have enough rope to get to the victim.

With the redirect at the anchor shown on the Petzl link, you need 20 x 2 =

40m of spare rope, which you can't possibly have.

3. The Petzl approach has the rescuer pulling toward the crevasse as the

rope is redirected through a pulley at the anchor. I can see no reason for

this pulley, it just adds friction with no increase in mechanical

advantage. And I would want to walk away from the hole, back along my

tracks!

4. If doing an unassisted hoist, a 5:1 or 7:1 might seem attractive, and

with all that friction I can see them being needed. But there is a big

difference between 7:1 being used because it means you have to put less

effort in, and 7:1 because you need 7 times the force just to move

something. In the latter case, you are putting 6 times your max pulling

force onto a wet glacier snow anchor. It can't see it holding in anything

but ideal conditions. Or a dry glacier – in which case hopefully no one would fall in.

Sorry for the long post, and any comments would be really useful as I'm

off into the Alps in two weeks! Thanks

Think you lost me by point 2, there again I'm not good with word descriptions, so all I can say is this is how I like to cross wet Glacier's

Normally your climbing on 2 half ropes. Rear person puts one rope in sack or draped over. Both tie into the ends of the other rope, rear person then takes a couple of coils only and clove hitches off on belay loop. Lead person takes coils to just before the middle mark in rope and ties off the same. Make your knots in-between the 2 of you until you have your 20m gap, note length of rope will dictate number of knots. With this system should the lead person fall in and assuming your both not in the hole, then rear person makes safe with connecting rope (hence couple of coils to play with) They then have the other rope in bag to effect rescue if needed! Also this way if rear climber falls in hole then the lead climber should have enough rope on the coils to effect the rescue! One thing to note is the only time I've had a partner fall into a crevasse (not that experienced and I like to send them first!!) it was on a sloping glacier. The cracks tend not to be vertical but break at 90 degrees to the surface meaning as in our case my uninjured partner was able to climb out on the downward side of the crack! As with most things all of this only works if you both stay on top!

In reply to pass and peak:

Perfect. I like that. And have not seen a video or a webpage with that system.

Keen to see replies to this as I often ponder the same issues.

Not that it's much use to you, but my conclusions tend to be that if the victim is unconscious or incapable of helping then they're staying down there!

Both from theory and practice I struggle to see how one person (especially a weak, lightweight like myself) can extract a deadweight partner in 'real' conditions (rope cut 1m deep in to snow, etc).

However, there are a few things I would throw in to the mix in reply to some of the specific issues...

>  it recommends about 20m between the climbers, and the use of knots.

I personally would use less rope between climbers, especially as I tend to use knots. This would usually be 12-15m. This gives more 'free rope' to play with in your calculations. I guess 20m is recommended to reduce the chance of person 2 being pulled in to the hole by person 1, but if you have 3-4 big bulky knots, a relatively taught rope, and are reasonably alert(!) then this risk should be pretty low. Obviously there are variables that may affect this... wet vs dry glacier, crampons vs skis, gradient, etc Also, if skiing downhill roped (rather than skinning) then knots are a nuisance and a longer rope probably sensible.

> 1. some sources pad and prepare the edge exactly above the climber, (a) how

> easy is it to prepare this without pushing ice and snow on top of the

> victim? (b) how do you get a sack under a tight rope in snow? (c) is it

> just better to go to one side and prepare that area, then use the spare

> length of rope to perform the rescue. Some videos show this.

> 2. many sources show the rope the climber is hanging from as part of the

> haul system. EG 3:1, 5:1, 7:1. However assuming the rope has knots that

> worked, dragging them back through the snow might be hard, and passing them through the haul system not covered in anything I have bumped into. (I know how to pass these knots, but I doubt everyone does). Hence it seems

> again logical to use the spare length of rope. As some sources do.

Both times I have done a crevasse rescue there was no chance of doing anything with the 'live' rope, other than belaying it. One time it was knotted and just as the knots did the job they are supposed to in adding friction to help arrest the fall, they did the same job in stopping the rope from moving the other way! The other time was with no knots (skiing) but the rope was so deeply cut in to the lip there was too much friction to move it. So this comes back to leaving enough free rope to be able to reach the victim. The use of the free rope assumes the victim is conscious to attach to it... unless you abseil in to do that for them!

> 3. A 2:1 assisted hoist with the spare rope seams by far the best idea. I assume 9 times out of 10 the victim if fully functional? An one would almost always you an assisted hoist on rock if you had the choice. For one it means much less force on the anchor, but also only takes seconds to initiate.

Particularly the 'less force on the anchor' part... whatever that anchor is it's not going to be like hauling off a pair of shiny bolts.

> 3. The Petzl approach has the rescuer pulling toward the crevasse as the

> rope is redirected through a pulley at the anchor. I can see no reason for

> this pulley, it just adds friction with no increase in mechanical

> advantage. And I would want to walk away from the hole, back along my

> tracks!

I'm no expert, but I think the pulley at the anchor is there less as a pulley and more as a progress capture device to reduce the chance of dropping the victim by letting go of the rescuer end of the loop? (Correct me if I'm wrong though, as your knowledge of systems is a bit better than mine!) I see what you mean about it adding friction for no increase in pulling power though.

> 4. If doing an unassisted hoist, a 5:1 or 7:1 might seem attractive, and

> with all that friction I can see them being needed. But there is a big

> difference between 7:1 being used because it means you have to put less

> effort in, and 7:1 because you need 7 times the force just to move

> something. In the latter case, you are putting 6 times your max pulling

> force onto a wet glacier snow anchor. It can't see it holding in anything

> but ideal conditions. Or a dry glacier – in which case hopefully no one would fall in.

A couple of winters back I buried a ski as an anchor for my partner who had fallen in a crevasse (no indication it was there... he was unlucky to do a snappy turn exactly in the centre of the snow bridge!) Even in reasonably consolidated snow, with a textbook set up (though I say it myself!) the anchor was shit! It was solid enough as a belay to keep me safe, but there was no way I could have hauled anything bigger than a toddler off it. I ended up digging down about a metre and a half until I hit the glacial ice and could place a screw (which, to be fair, you probably could have safely hauled several times max pulling power on).

Sorry for a long reply that doesn't really help! I think the answer I always come back to is:

If your mate is conscious then he/she can do most of the work climbing out of the hole. If they are unconscious then that's why you carry a knife...

I'm keen to see if anyone has any more constructive advice!!

Some good info on how to tie your in-between knots near the end of this video  https://www.youtube.com/watch?v=IgNR-VZMwHo& Note on skinny rope each re-wrapped knot can take 1.5m or more.

In reply to pass and peak:

Like that! Quite an elegant solution.

There are plenty of days I'd go out with just one rope (easier lines, moving together, no abseils) but for the rest I think I'll be adopting your plan... especially as it makes the weight split a bit more even!

As per the previous replies:

• If you are on your own it's very difficult
• If you are on your own and the person in the crevasse can't help themselves (injured/incapacitated), you WILL need extra help.
• If the rope is knotted you're going to need another rope to get the person out.
• Even if the rope isn't knotted, the "live" rope is going to cut into the crevasse wall and there isn't a hope of getting something underneath it.
• If you don't have a second rope, all the coils on the second person so they have enough rope to stand a chance of getting the first person out (and hope the second person doesn't sniff out the crevasse first).

Never hard to try in anger but have practised and the upshot was it's all a lot harder than it looks in the Petzl catalogue......

I have tried in anger and failed even though I was fully aware of the theory.  It was a good job another pair of climbers turned up.

I would agree that it's almost impossible for a single climber to extricate another unconscious climber from a crevasse on his own. I have never seen the technique demonstrated with a real crevasse and a dead weight climber with a rucksack on his back. If someone knows of a video like this please provide a link, I'm happy to be proven wrong.

Al

Just considering your maths further... (I've a lot of work to do so enjoying the opportunity to prevaricate!)

Do you need 6 knots between climbers? I would normally have 4 -2 sited toward each climber. If the knots are going to work at all then you shouldn't need any in the middle of the rope?

What knots are you tying to take up 2m of rope? Even tying a super bulky monstrosity I can't make it take up more than 1m of rope. What am I doing wrong?!

So with my normal tactics I would use less than 4m of rope on knots -a lot less than the 12m lost in your original calculation.

The way I was taught:

If in a party of two, each climber should have coiled/in backpack twice the length of rope that is between the climbers.

Steps for rescue of a climber unable to ascend the rope:

Set up an anchor, transfer weight onto anchor/escape system.

Prepare edge and put backpack/ski/axe down for the rescue rope to run over. Attach rescue rope to anchor at both ends, so there is a drop loop going to the casualty. Abseil in on the rescue rope and clip casualty in to the rescue rope with a pulley and ideally progress capture. I use a micro traxion here but could use pulley/prussik (I think the physics means if you only have one pulley it should be used here though I might be wrong). Ascend rope.

Have practised this with a climber who didn't assist, though obviously a little different to completely unconscious. Worth adding that I have been attached to a climber who fell in a crevasse and 100% of the load was taken by a knot in the rope (butterfly), so I always use knots now. They can also be useful when ascending as you can clip directly into them.

Hi Tom, I was using the knot and number of knots from the video linked above.

Extracting a fallen climber from a crevasse fall is VERY difficult,unless they're able to help themselves with prussiks. I know because I've been in that situation. You can read all the advice and theory on how to do it, but it's something that you need to practice in a "safe" environment either on a "dry" glacier or somewhere where there are fissures and cracks in rocks like Southern Sandstone (where it's important to ensure that your practising does not harm the rock).

Having lost two friends together in a suspected crevasse fall many years ago (their bodies have never been found) I made it a rule that I would never cross a "wet" glacier solo or as just a pair, and only do it with a minimum of 3 of us on the rope. Mountaineering is full of risks, but sometimes the risk is unacceptable. Of course the point at which the risk becomes "unacceptable" is a personal choice, but my perspective changed dramatically when it actually happened to close friends and I watched the agony of uncertainty and worry their relatives went through following their disappearance.

> If in a party of two, each climber should have coiled/in backpack twice the length of rope that is between the climbers.

This is one of my issues.

Worse case 50m rope, 12m in knots means 38m of rope. Divide by 5 to match your rule. Distance between climbers is only 7.6m. Whereas the common recommendations is 20m

Best case. 60m rope 9m in knots. Gives 10.2m!

Hi Dave.

Have e mailed a response to you,

Mark

I guess you could skip the drop loop and use a single strand and 3:1 for hauling with a shorter rope? Though I'd think using alpine butterflies it couldn't be more than 50cm per knot, which might be worth considering.

Edit to add: usually do my best to have two ropes between two peopl. Recently got a 30m petzl RAD line which makes this easier

Post edited at 14:03

Here is one of of the worst offenders

The blocking knots suddenly disappear by magic.

If companies like this and petzl can't get this to work, what chance is there for punters like us.

In reply to pass and peak:

Thinking about this more, is the obvious solution to have two ropes with one person tied in the middle of one rope and the end of the other, and the other climber doing the reverse. Half the knots go in one rope half in the other. Each climber now has enough rope to mount any kind of rescue, and the knots do not have to be passed or pulled through the snow. One rope can even act as the guide to get the assistanced haul loop down

Anyone seen this written up or a video?Please

Maybe, but I think your working from the premise that your GOING to fall in a crevasse, when actually it's a rare event. I think having to drag 2 lengths of knotted rope through wet snow, possible up hill would rather tire any leader out, especially at higher altitude! Plus both your ropes are getting wet and UV'd

M

Getting slightly more insane. Why not put knots in both ropes, then clip a snow stake between the two ropes. Auto braking.

In reply to David Coley: Good topic this, I have crossed a few glaciers over the years and luckily never had an issue but I'm keen to improve my glacier technique if some good suggestions come out. If you cross a glacier in the Alps you see a variety of systems in use and to be honest I don't see that many people with knots tied in the rope, I would say they are in the minority. You also see a worrying number of people either with no rope or with a couple of metres between them.

My system is for Alpine summer, that's not meant to demean the risks, its just I think the situation would be more risky in Alpine winter or in the Greater Ranges and you may need a more robust system in those circumstances.

My "system" if you can call it that is:

• 60m rope
• 15m between climbers if 2 people(slightly less if a rope of 3)
• No knots
• 20m+ coiled by each climber, tied off then a prussic attached to allow a bit of slack to escape the system. Also means the direction of pulling force is from the harness.
• If one of the party falls in the other party(assuming he hasn't been dragged in as well?) gets an anchor in and escapes the system.
• Next steps are:
1.  Is the guy in the crevasse able to climb out!
2. If not, shout very loudly for help, is anyone nearby?

3. If not call rescue services

4. If rescue services are not available/contactable then can you go and get help reasonably  quickly and reasonably safely(this might be quicker than setting up a haul system alone)
5. If not....well you are going to have to set up a haul system for someone who cant climb out so he must either be injured or unconscious so it is going to be extremely difficult

I've always assumed(hoped) that steps 1 to 4 would be successful so I have neglected step 5.

Just regarding point 3 in your original post.

It's setting yourself up to be able to see what's going on at the other end of the rope.

Even if the haulage is working smoothly, the person below is likely to get stuck under the lip of the crevass

This comes up periodically.  I think you are right all the various systems are, to say the least, highly optimistic if there are two people and one rope.  It's one of those topics, like ice-axe braking, that don't actually work in practice but people obsess about endlessly theoretically.

One approach that might actually work was presented by"jon" a few years ago:

- One rope.  On person ties on the near the middle, the other to both ends

- One strand between climbers with no knots.  It is slack with a few hand coils each.

- Another strand between climber with knots.  Kept taught.

-The knotted strand catches the fall

-The unknotted strand is free with sufficient slack to arrange a hoist without the need to make an anchor, while the knotted strand takes the load throughout and aids climbing out.

> This comes up periodically.  I think you are right all the various systems are, to say the least, highly optimistic if there are two people and one rope.  It's one of those topics, like ice-axe braking, that don't actually work in practice but people obsess about endlessly theoretically.

I think it's a lot more theoretical than ice-axe braking, which I've done successfully in serious situations a couple of times, as have others I know. I've never met anyone who has managed to rescue an unconscious partner from a crevasse unaided. It would be fascinating to hear from anyone who has. If on a dodgy glacier nowadays I only relax if there's three of us with a fat lad at the back.

If there are two people then the only realistic option is that the person in the crevasse rescues themselves while the person on the surface makes as good an anchor as possible - with a sensible amount of rope (20m) and knots just lying there with an axe or ski through a Prusik is enough.

Having an unconscious or injured "victim" isn't that likely assuming that there wasn't much slack - getting an unconscious body out of a hole is well beyond basics and requires more than one person.

For the victim to extract themselves they need to master the more troublesome phase which is getting past the lip of the crevasse where the rope has cut through the snow and for this a micro-traxion is worth more than its weight in gold.

If it's a ski touring scenario then also practice taking skis off whilst hanging on a rope as losing them would be annoying and will involve a helicopter ride.

> This comes up periodically.  I think you are right all the various systems are, to say the least, highly optimistic if there are two people and one rope.

my beef isn't so much that this is all theoretical, but that major companies are putting out safety educational material that shows impossible things, like magically vanishing knots. I then have to spend a long time discovering that it can't work, or go into the hills thinking I have a working solution, then find out too late my rope isn't long enough

Thanks. Shame their rope is 112m long.

> One approach that might actually work was presented by"jon" a few years ago:

> - One rope.  On person ties on the near the middle, the other to both ends

> - One strand between climbers with no knots.  It is slack with a few hand coils each.

> - Another strand between climber with knots.  Kept taught.

> -The knotted strand catches the fall

> -The unknotted strand is free with sufficient slack to arrange a hoist without the need to make an anchor, while the knotted strand takes the load throughout and aids climbing out.

Now we are getting somewhere. That's awesome. Has Jon tried it? I feel the need to buy him a pint

I dont know. He still appears sometimes here. As I recall, he got it from ENSA, so it has pedigree.

Post edited at 19:23

As I remember it was a method that originated in Norway. It was discussed favourably at ENSA during a course that I attended. I never tried it and I don't know if ENSA ever adopted it in any way other than to propose it as a viable alternative to other equally viable methods. Whether it gets any mileage now, I don't know. I'm not sure you're right about the not needing an anchor bit though, or have I misunderstood (or forgotten something)? The advantage as I see it is that there is an unknotted rope instantly available that is already attached to the victim with enough slack to effect a pulley system of some sort.

Might have been mentioned above but it’s a knot every 2m, not 2m of rope to make a knot. The knot is just an overhand or similar so uses up at most half a metre of rope.

Yes, on reflection I think you are right.

Hi,

No the knot recommended in the video is a figure of 8 with an tucked in loop. I tried and it takes at least 1.4m per knot. They recommend only 6 knots, so minimum of 8.4m

As you say, Petzl recommend one per 2m. So 10 knots. Even with an overhand, this is 10x0.6m=6m

However the knots are not the main issue lengthwise. It is the need to run the rope 3 times to the climber that is in https://www.petzl.com/NL/en/Sport/Crevasse-rescue-no--3--haul-systems-for-crevasse-rescue

this requires an >80m rope plus the rope in the knots and the tie in knots, so 90m.

> Just regarding point 3 in your original post.

> It's setting yourself up to be able to see what's going on at the other end of the rope.

> Even if the haulage is working smoothly, the person below is likely to get stuck under the lip of the crevass

Hi, just looked at the diagram again

https://www.petzl.com/NL/en/Sport/Crevasse-rescue-no--3--haul-systems-for-crevasse-rescue

If you are right, then there are 3 rope lengths to the climber = 3 x 20 = 60m, then a 4th extending to the edge, so 75m in total, plus any coils on the victim plus the knots so >100m . It just doesn't work. It can't work with a team of two (as shown) and one rope (as shown). I would suggest this is a great method if carrying 2 ropes and one had already joined the ropes. Yet there is no indication this is so.

Clearly you can only do an assisted hoist where the fallen climber is conscious and if you have enough rope. In other words, where they haven’t fallen very far and perhaps there was less than 20m of rope out in the first place. Also in practice people often don’t bother with the knots for the reasons you mention.

Otherwise you do a 3:1, which you can do even if you don’t have much rope available. There are variations on the 3:1 which I guess you’d be familiar with. But as you say might not work with the knots.

Post edited at 21:09

I think I need to mention something here as there's a lot of talk about an unconscious climbing partner in the hole. Priority 1 after making your anchor is to establish communication with your partner in the hole! If their unconscious then ideally you need to get down to them and make sure their in a position that doesn't compromise their breathing and sort them out. This is NOT a simple task and needs a lot of practice in order not to jepordise yourself! As mentioned above, if the person in the hole is not able to give themselves a lot of assistance then things are bad. If they feel their injury's are not life threatening and you know help is on the way, it may be best to wait for help! Either way, reality if is often different from theory, but without the theory we all would be in a worse position

As drawn, the distance they fell is irrelevant. it is the tie in distance that matters. for it to work and either climber be able  to do an assisted haul, you need to be so close to each other you blow all the other recommendations in their other docs. It can't work. It needs a second rope, or for you to join your ropes before you fall in. Or you to be no more than 60/5 = 12m apart (without knots), or if you have a 50m rope 10m. The use of knots reduces this further. Madness.

Post edited at 21:42

there's only one thing for it - zip-line across the glacier.........

Post edited at 22:08

I don’t really see what the issue is. The assisted hoist is one of the options presented. You might be able to use it (second rope, no knots, only 12m apart, etc). Or you might not. In which case use one of the other options.

I guess your point is it’s not possible if the other recommendations are followed. But in reality the other recommendations are not always followed.

Exactly. So if you present a method which needs lots of other things to be in place, you need at least to hint at this. And not produce material that kind of contradicts other advice or doesn't logically stack. The need to be tied in at 12m for this system to work should be stated. As 12m is far from the norm, including the diagram I would say is strange. One could train using this and then waste a lot of time in an emergency trying to use it before realising it can't work.

Yes, you could use an assisted hoist with a spare rope. Hence this diagram should show this. It doesn't, it clearly shows the original rope being used. With an second rope you would rig it differently anyhow.

I think youre expecting a bit much from this guidance. These are examples of how things can be done.

I agree with your point regarding the disappearing rope knots. But you might consider how a knot could end up below the lip of the crevass and how this could be avoided.

Yes, the examples are discrete. There are many variables, it's not going to tell you a step by step breakdown of where to put your hands.

I'm not sure they are meant to be totally discrete. But if anyone can explain how the assisted hoist in the petzl diagram can be completed when tied in at the recommended distance of about 20m, I'll give a tenner to the local rescue team. I can imagine someone trying this for real and waisting valuable time.

I'd also like to know if a 7:1 would hold with a buried axe as shown in many videos, unless you got lucky with supper good snow.

teaching materials should within practical limits be realistic. It is because they are not realistic that the authors are not spotting some of the issues and then modifying their materials to reflect this. This is why when laying out a kitchen one draws to scale, or uses graph paper; by doing so one sees what doesn't fit together.

Video from ENSA about the strength of various deadman anchors. Obviously only valid for those specific snow conditions but gives an idea.

Thanks. Yep I'd watched that. The were getting 200 to 300 kg equivalent I think . So I'm not sure a 7 to 1 would be a good idea

Personally, when I read about knotted ropes in Peter Cliffs 'Alpinism', it was the first time using a rope for a party of 2 made sense; I've done it religiously since. In his words, 'you will be pleasantly surprised how easy it is to hold the fall.' I'm prepared to take his word for that; using the knotted rope to climb out seems feasible as well.

I'm not sure that self rescue is possible with a party of 2 and the  victim unconscious, if there's no prospect of help I suspect you are probably in Touching the Void territory.

> Thanks. Yep I'd watched that. The were getting 200 to 300 kg equivalent I think . So I'm not sure a 7 to 1 would be a good idea

The load on the anchor mainly depends on that needed to lift the person down the hole and overcome the friction on the lip. At a guess up to double body weight?

Using a 7 to 1 enables a haul with a reasonable pull load for the hauler, it does not put more load on the anchor.

That said, have never heard of any case of one person hauling out an unconscious casualty.

The only time I have been involved with a crevasse rescue, we fixed an anchor, ice screws in a dry glacier, and went to the edge.  The unfortunate one had slipped in boots without crampons, burst through a snow bridge, leaving a highly amusing M…… shaped hole, he does not take kindly to being reminded of the incident. However to his credit he insisted on improvising a harness with slings and prussicked out under his own steam, took a good hour, he was almost hypothermic, whilst we sunbathed on top. Most of the time was taken moving the first few inches out of the constriction in the  crevasse narrowing that stopped him.

Post edited at 18:38

Bet he was glad he’d been on the pies and beers ;-)

> The load on the anchor mainly depends on that needed to lift the person down the hole and overcome the friction on the lip. At a guess up to double body weight?

> Using a 7 to 1 enables a haul with a reasonable pull load for the hauler, it does not put more load on the anchor.

Yep I get that. But with the friction lost in a 7:1, and the friction on the ice/snow, possibly with brake knots which might be caught on the lip the possibility of pulling such as to weight the anchor with more than 200kg equivalent isn't theoretical.  I haven't done any hauling on ice, put I have on rock, and if there is any friction (from the stance edge, runners etc.) you can be pulling with a 5:1 with full body weight and go red in the face and still not move them.

Excellent discussion!

Petzl says, somewhere in one of their articles, that one drawbacks of the 2:1 haul system is, that it requires a lot of rope. So, they are fully aware of this, and somewhere else they suggest tying together two ropes. They have a solution to hauling a rope with knots. All in all, I think they cover all bases.

ENSA Chamonix has made some excellent videos about the subject, and very possibly they work together with Petzl. They have filmed good evidence for the benefits of having big knots in the rope. Inconsistently, one 2012 video shows 4 knots where the newer ones show 6. Supposedly they have become wiser.

Outdoor Research in collaboration with AMGA has made several excellent videos too, and they are the most instructive. Unfortunately they don't support the use of knots and they suggest shorter distances between climbers (10-n).

What I would do when in a team of two with a 60 m single rope:
1. Find the middel, tie 6 big knots 2 m apart, tie a fig 8 knot 3 m from last big knot, and tie in to the end as normal. Around 20 m rope is used for this, which is leaving around 20 m for each climber to coil. (This is the ENSA/PETZL method)
2. Kiwi coli the rope until reaching the fig8 knot, and end with an overhand around the coil. Then clip the fig8 to the belay loop with a directional locker. (AMGA/OR method)
The knots help stop the fall, and the fig8 makes it easy to transfer the load to the anchor once made.
3. Walk towards the lip protected by a prusik on my end of the rope. Asses the victim and the crevasse. If needed, I would enter the crevasse to perform first aid.
4. I might just drop my end of the rope to the victim and protect the lip with my backpack and leave the victim to prusik up the rope.
5. To facilitate a 2:1 assisted haul, I might make another anchor closer to the crevasse.
6. Otherwise, I would shove an axe under the rope as close to the crevasse as possible, and then apply a 3:1 or a 6:1 haul (I don't like tending two prusiks in a 7:1) using the method mentioned by Petzl to pass the knots. Actually, passing knots is fairly straightforward.

What I would do when in a team of two with two 60 m half ropes:
1. Tie the ropes together with a double fisherman, then tie 6 big knots etc... (Mentioned by Petzl)
2. As the excess rope is around 40 m I might put it in the backpack. Otherwise as above.
3. As above.
4. I would skip aiding victim self rescue.
5. I would clear the lip a meter or so from the victim to facilitate extraction, then combine 2:1 assisted haul with 3:1 haul as mentioned by AMGA/OR.

Thanks. A very clear discription.

One thing I have yet to resolve is whether the 3 to 1 or 7 to 1 haul can work with knots. Unless the victim is very lucky, at least one knot will have passed the edge and the rope behind this cut a deep groove. Some commentators think this means the drop loop is compulsory if knots used. The existence of the petzl passing the knot instructions indicate otherwise. But I don't know if they have evidence for this, or just invented a system on Friday to solve the problem.

What is driving me a little mad is that this isn't a little more joined up. Video and pdfs take time and money so this surprises me. Petzl do point out the issues elsewhere, but many will not spot it, and I see little reason not to comment on the issues in the pdf or video. We are talking one sentence. In fact if they had said the drop loop haul needs a second rope or anchor they would have not drawn an impossible diagram.

The ENSA knot video looks great, until you ask yourself, can I still use my normal 3 to 1 method to pull someone out? They should have tried and not published the video until they at least had a working  knowledge to comment on how knots might impact the whole rescue sequence.

When I wrote the self rescue chapter for my book I hauled real people. On real cliffs. On real climbs. It was only by doing so that I could discover some of the issues not mentioned by many sources.

In the arc'teryx video they simply untie the knots rather than passing them. If they really had 80kg on the line they could not do this.

The ENSA snow anchor video ends in a table in dNa. With no comment on the likely hauling forces in a crevasse rescue. And even if not technically correct, use kg not dNa.

I agree. The videos and pdf's are leaving us behind without knowing if we can haul the knots through the snow groove at the lip. I have now looked through some of the comments, and found an answer from ENSA saying that: "Knots are certainly a big hindrance for hauling someone out of a crevasse".

However, on a wet glacier with more than a meter of snow cover, I would still want to use knots, because they will help stopping the fall and thereby they will limit the number of casualties to one. This calls for having enough rope for an alternative haul line. And with only 60 m of rope you will need to setup a hauling system from above (3:1, 6:1, 7:1) as there is not be enough for a 2:1 assisted haul. Because 2:1 "Requires a lot of rope", as Petzl put it quite drily.

So, my take is, that on a wet glacier with deep snow cover, each climber should have a coil of 20 m rope, and between them they should have 6 knots consuming 6 m rope leaving enough distance (14 m) and knots between them to stop a fall. Then they have just enough for setting up a hauling system from above. But if the snow cover is less than one meter, knots are not of any help, and then distance between you is more crucial. And if the glacier is bone dry, you will not be able to stop a fall, but will be able to see the crevasses from a distance.

In reply to pass and peak:

Either way, reality if is often different from theory, but without the theory we all would be in a worse position

I agree but it is somewhat misleading, indeed some might say downright dangerous, to suggest in any way that a single climber can successfully rescue another unconscious climber, with a pack on don't forget, from some way down a crevasse unaided by anyone else and only using what would typically be carried by a pair. There are dozens of videos out there claiming to demonstrate this but my point is that I have not seen one that does so accurately or realistically.  Having tried it myself, and being fully aware of the theory, I feel fairly confident in suggesting that IMO it cannot be done but I am happy to be proven wrong.  That's the challenge, prove me wrong.

Al

OK this is my take, having watched and read everything I could and read the comments here and on MP, and then using ideas from my rock climbing book. Comments welcomed. 2 people, high risk deep snow situation, big packs but much applicable elsewhere. He and she 20m apart, 6 monster knots. Kiwi coils, but with a fig8 clipped to belay loop to make clipping into the anchor easier. Assumption being that the rope with the brake knots cannot be hauled due to one knot being over the lip, and an assisted is normally better anyhow. The aim is to reduce the amount of rope needed, and the force on the belay, and the friction. There is no one size fits all, so this is just 1 possible approach, but I hope a working approach that takes onboard the experience of those that have been in this situation for real.

1. hopefully she gets herself out. If she can she heads for the far side of the crevasse, not back towards her partner! Otherwise she will still have to cross. (unless she was at the back when she fell in)

2. her first act is to move sack to belay loop. Not clipped to form a 2:1 loop as shown in many books, as the brake knots will bugger you up later. This requires the sack to be pre rigged with a 120 sling. Clip chest coils to prussik on rope to keep upright. Particular if you are feeling faint or injured.

Place ice screw or axes to unweight him (the belayer) and make his life easier (he will not know anything about the condition, thoughts etc. of her). Clip into the screw with sling or prussik, which can be cut if need be (e.g. 12 people turn up and just start pulling. She removes snowshoes and probing walking pole(s). This will require some 60cm slings to stop her dropping them. Extract crampons and axes from sack to help getting over the lip. Best if the axes are tied to sack, not between sack and shoulder when crossing glaciers as you risk losing them.

3. Assuming she can't climb out, she awaits communication from him. Best not to prussic before anchor built, unless it is a very short distance until she can reach ice; often best to await an assisted haul which is faster than prussiking.

4. Meanwhile. He builds an anchor. If possible two. clips fig8 to it and escapes the system. Removes coils. and using this spare rope keeps himself safe using a prussik or retying knots and approaches the edge. In many cases using the loaded rope as a safety line will not work as it is buried deep and has 6 knots in it. Be very careful as the axes are now in the belay, so don't fall in.

5. If she is hurt, phone. Otherwise:

6. Have a comforting chat. And mention that you once read a very long thread on UKC and all agreed crevasse rescue was as easy as pie. Clear the edge, clear deep, clear well, as the lip is the tough bit. Do this slightly to the side as at least one victim has been killed due to falling ice. Pad the edge. Drop the loop. The pad goes under the loop, not under the main rope, as the latter is deep in the snow and using this method will not be moving. Secure the pad with a walking pole or whatever is to hand.

7. Now this is the new bit. She puts HER trax on the loop and clips to HER belay loop. This does two things compared to the Petzl method. It means you don't need an 80m rope; it reduces the force on the anchor. She will need to know which rope is the pull side of the loop. He could preplace his trax so she don't need to understand the orientation bit, and can't drop the trax, but this would leave him without a trax should plans need to change.

8. if she is but light and he a thug, he pulls and she pulls the other rope. This is only a 2:1 not a 3:1 assisted, so this will be hard. If this doesn't work, she puts a 120 sling or prussik on the drown rope to remove her weight during each stroke.

9. Bob's your uncle, she is out. Time for a cup of tea.

10. if she drops the trax, use a locker if the distance is small and she only needs a bit of help. If not, and this is another new bit compared to many videos. Use a prussik on the main rope between the anchor and edge to form a new master point or one of the brake knots. The prussik will need a very large number of turns. By moving the master point, the amount of rope needed it once more much reduced.

Note this whole thing needs trax's, prissiks and quite a few slings. If she is unconscious and no helicopter, you are going to have to rap to stabilise her and haul. But haul using the free rope because of the brake knots.

Reasonable?

Post edited at 14:36

> 7. Now this is the new bit. She puts HER trax on the loop and clips to HER belay loop. This does two things compared to the Petzl method. It means you don't need an 80m rope; it reduces the force on the anchor.

I don't quite follow how connecting the haul loop to the victims belay loop with their own trax will
a) reduce the need for an 80 m rope?
b) reduce the force on the anchor? (maybe because of the reduced friction?)

This seems to be somewhat overcomplicating things.

An alpine butterfly uses between 40 and 50cm of rope so with five knots that's two and a half metres gone.

Most of the Petzl diagrams show two people to avoid complicating the situation but are, in reality, intended for three or more. For example the usual case in a team of 3 is for the 3rd person to create the anchor between the second and the victim, ideally as close to the crevasse as possible to have more free rope.

I'd advise getting a proper book such as Bergsport Winter - Technik / Taktik / Sicherheit from the SAC which explains far more than the mechanics.

> 1. hopefully she gets herself out. If she can she heads for the far side of the crevasse, not back towards her partner! Otherwise she will still have to cross. (unless she was at the back when she fell in)

This isn't going to happen. You climb out of a crevasse on the side the rope is.

> 2. her first act is to move sack to belay loop. Not clipped to form a 2:1 loop as shown in many books, as the brake knots will bugger you up later. This requires the sack to be pre rigged with a 120 sling.

No need to pre place as it's easy enough to put a sling through the belay loop then clip the sac. 60cm "alpine" quickdraws are ideal for this.

>Clip into the screw with sling or prussik, which can be cut if need be (e.g. 12 people turn up and just start pulling.

Where do we start...

Multiple people pulling is a good way to crush someone against the lip of a crevasse and should only be attempted after stabilising the situation and with a lookout on the lip to communicate.

> 3. Assuming she can't climb out, she awaits communication from him. Best not to prussic before anchor built, unless it is a very short distance until she can reach ice; often best to await an assisted haul which is faster than prussiking.

It's much quicker and more comfortable to Prusik out than to await being hauled. As a victim one can assume that if the partner is still on the surface then the anchor is already good enough. As a team of two both people have to be able to self rescue - otherwise they shouldn't be on a glacier.

> 5. If she is hurt, phone. Otherwise:

Always phone unless it's a large group and the situation isn't serious (i.e. you know what you're doing and you can still see the head or hands of the victim). Far better to call back after ten minutes to say that all is well and no help is needed - the helicopter may not even have take off yet.

> 6.  Pad the edge. Drop the loop. The pad goes under the loop, not under the main rope, as the latter is deep in the snow and using this method will not be moving. Secure the pad with a walking pole or whatever is to hand.

It's actually a pole or spare axe that you'll be sliding under the rope to prevent it sawing into the snow. Good luck trying to shove something soft under a tight rope.

> 7, 8, 9 and 10.

Have a look for the Österreicher-System / Austrian haul and the various tweaks possible.

> Reasonable?

Not really alas.

Post edited at 18:53

The petzl method requires the pull rope to be returned to the anchor. As that is where their tax is hence the need for a rope longer than 60m

Again with the tax on the anchor, all the load is on the anchor. With it on the belay loop only half is, the rest is on the person pulling

Hi Ewan,

> Most of the Petzl diagrams show two people to avoid complicating the situation but are, in reality, intended for three or more.

I'm not criticising most of their diagrams, just one, the assisted hoist. Which is impossible and I will still put money in the local mountain rescue tin if anyone can show how this can be done AS DRAWN with a >=15m distance between the climbers and a 60m rope.

> For example the usual case is a team of 3 is for the 3rd person to create the anchor between the second and the victim, ideally as close to the crevasse as possible to have more free rope.

I have only been talking about a team of 2, as that is what I wanted to learn.

> This isn't going to happen. You climb out of a crevasse on the side the rope is.

I think that depends on the crevasse. I have climbed out the other side (it was a small hole), so have others. Doesn't Mick Fowler climb the far side of the crevasse in his autobiography, as the rope side was overhanging?

> No need to pre place as it's easy enough to put a sling through the belay loop then clip the sac. 60cm "alpine" quickdraws are ideal for this.

I have always found it a good  idea to pre-do this on rock routes where I am going to clip the sack to the belay. So I guess I just brought it over from there. However, you will note I said big packs. Once you have 20kg on your back you sure are want to do this quickly! Possibly upside down.

> Multiple people pulling is a good way to crush someone against the lip of a crevasse and should only be attempted after stabilising the situation and with a lookout on the lip to communicate.

Good point. Thanks!

> It's much quicker and more comfortable to Prusik out than to await being hauled. T

Is this true? They will also be prusiking towing a 20kg sack and having to tunnel through a metre of snow. For all the rock experiments I did for my book, a drop loop assisted hoist was an order of magnitude faster. On rock I could never imagine asking someone to prussic over a roof rather than tossing them a rope and just pulling. The only way I can see it being faster is if they pussik before/while the anchor is being built. This might be possible if they are just over the lip.

> Always phone unless it's a large group and the situation isn't serious (i.e. you know what you're doing and you can still see the head or hands of the victim). Far better to call back after ten minutes to say that all is well and no help is needed - the helicopter may not even have take off yet.

Great point! Thanks

> It's actually a pole or spare axe that you'll be sliding under the rope to prevent it sawing into the snow. Good luck trying to shove something soft under a tight rope.

This is where you lost me. This is a 2:1 drop loop haul, not using the loaded rope, so the padding is placed under an unloaded rope. This what Petzl show and it looked reasonable.

> Have a look for the Österreicher-System / Austrian haul and the various tweaks possible.

I tried, but failed to find a diagram.

Post edited at 20:20

There's a diagram of the Austrian system at the end of https://www.sac-cas.ch/fileadmin/Ausbildung_und_Wissen/Sicher_unterwegs/Sicher_unterwegs_Hochtouren/SAC-Merkblatt-Bergsporttechnik-2010.pdf

I agree that the  assisted hoist as drawn by Petzl won't work without lots of rope - it's very much a technique for teams of more than two as one needs to build the anchor as close to the crevasse as possible. On rock close to the belay it's a great technique.

I can happily climb out of a crevasse before a surface team has finished installing their system but it's down to practice. There's also a huge "user experience" difference between being pulled up rock and through snow.

The Petzl diagrams don't show the most efficient system which is to have the rope going through a micro-traxion on the belay loop then up to a tibloc and pulley on the rope and back down - wrap the rope around your foot for a relaxed leg powered ascent. This is also how one gets over the lip of a crevasse so better to do it from the beginning.

As an aside, 20kg sacks are very much in expedition territory and should assume that the person unfortunate enough to carry such a load has trained for such a situation. 10kg is easy to take off and clip a carabiner through the haul loop.

OK. I have now watched 30+ crevasse rescue videos. Most have felt slightly unimpressive, with various bits even I spotted that seemed unlikely to work (like untying the brake knots with the climbers weight on the rope). Others seemed to have what looks like to me not the most modern advice, particularly for a team of 2. For example, not discussing use of the free rope, prepping the edge, asking the victim to use their legs to make the final edge easier to navigate. Many of the possibly unrealistic things came I felt from not completing a complete scenario and not using a real weight. For a relative novice like me, the following set covered things really well, using modern tools (a trax and a reverso). I know that there are many ways to skin a cat, and every case will be different, but I belive for me at least it really helps to run through the whole thing once in a realistic as possible way, then add tools to the tool box, rather than the other way around. The gender reassignment John goes through makes it even more modern. https://www.youtube.com/watch?v=bDcUnocXhpw&