Twenty feet? Forty feet? Sixty feet? It seems like there should be a clear-cut answer to the question, but unfortunately there's not.
Some years ago a friend of mine was coming down the Coleman-Deming route on Mount Baker late in the season. He was at the back of the rope team. The person at the front of the team slowly began to work his way across a snowbridge. Approximately half-way across, the bridge collapsed.
The leader dropped into the crevasse. The guy in the middle of the rope team did not attempt to arrest the fall at all. And my friend immediately dropped into a self-arrest position. Each of the two climbers at the front of the team were essentially lowered to the bottom of the crevasse. But unfortunately, my friend was dragged in and fell to the bottom.
As the crevasse wasn't that deep, no one was seriously hurt in the incident. But it was a very close call. One could make a very good argument that if there were just a few more feet between these individuals, that then the fall could have been arrested before it became as serious as it became.
There are three things to consider when deciding on rope length:
1) How big are the crevasses?
Obviously, you will need enough rope out to make sure that two members of the team are not on the same crevasse. If you're in the Alaska Range or the Himalaya, this is significantly more rope than it is if you are in the Cascades, the Canadian Rockies, or the Andes.
2) How many people are on the team and what kind of room will you need to arrest?
The more people on the rope, the more weight there is. On a team with five people, I've seen a person fall into a crevasse and stop without a single member of the team self-arresting. While larger teams tend to be slower and more difficult to manage, they are better when it comes to arresting a fall.
It is also important to make sure that there is not only enough room between each member of the team to arrest a fall, but that there is also enough rope out to arrest the fall before getting dragged into the crevasse. Essentially, this means that there should be more rope out between people when there are smaller teams. Having lots of rope out between people doesn't matter as much with larger teams.
3) Is there enough rope to perform a rescue?
Why not just put all the rope out? Won't this ensure that you always have enough room to arrest?
Certainly there are places where having all the rope out is good. In ranges with giant house-eating crevasses like the Himalaya and the Alaska Range, it's probably best to put all of the rope out. But this does make crevasse rescue more difficult and it doesn't give you a lot to work with if someone gets hurt in a fall.
One of the things that we teach at AAI is that, if possible, you should have some rescue rope on either end of your team. This rope should be long enough to reach the next person on your team. This is so that if there is a major injury in a crevasse fall, that you have enough rope to rappel down and perform first aid before pulling a person out.
If you plan to travel to a range where you will need to have all of your rope out, it is good to practice crevasse rescue without rescue coils. Unfortunately, it is a slower and more tedious technique.
Rope Measurement for Smaller Ranges
I have a simple system for measuring rope length in the Cascades. I'm six feet tall and so my wing span is also approximately six feet. I will generally separate people by measuring the rope with my arms. Here is a team breakdown:
Two person team - 8 arm lengths - 48 feet between climbers
Three person team - 7 arm lengths - 42 feet between climbers
Four person team - 6 arm lengths - 36 feet between climbers
Five person team - 5 arm lengths - 30 feet between climbers
In a pinch, it's possible to have a larger team, but it is not optimal. And you should never go below 30 feet between climbers.
When you get to teams with five or six people on them, generally there is not enough rope on either end to perform a rescue. In such a case, you should have the leader (i.e. the strongest/most experienced person on the team) carry any rescue rope that's available.
Conclusion
Certainly, the amount of rope out between people is a personal and team choice. Some of you who are reading this are probably shocked at how little rope I suggest between people. And others are just as shocked about how much I suggest. Either way, I can say comfortably that I personally feel as safe as is reasonable with these lengths on the glaciers of the Northwest...
--Jason D. Martin
Update --
The guide who fell into a crevasse back in 2000 responded to this post. Following are his comments on the topic:
Hey Jason, I was checkin out your AAI blog this morning and I saw your post on rope distance on glaciers. Nice article - well written and a great topic. I thought about trying to write about that topic recently but got side-tracked.
Interesting to see an anecdote told about my experience. I learned a few things from that, and as such I might disagree with a couple of your premises - or at least have a few other factors to consider. After my accident, some people told me (Crusty old Tom Bridge being one of them) that I could have benefited from the rope distance being a bit longer. I saw that you put that in your blog as well. Keep in mind however that had there not been a little "floor" in that crevasse, an extra few meters would not have mattered, and we probably all would have gone in anyway - and died. Yeah, more rope distance = more "time to react" but only to a point. The amount of rope out was for me a tiny little variable. The other key thing worth pointing out (and you did anyway) is that it was a small, visible bridge that failed over an open crevasse, rather than the failure of a soft blanket of snow over a hidden crevasse.
My clients and I were tied the "standard" 35 or so feet apart. In the following season or two, I heeded Tom B. and others' advice that I should go more like 40-45 feet on cascades glaciers. It wasn't until 2005 or 2006 or so when I realized, that all things being equal, I should have gone way closer than 35 feet. When I started going through the AMGA alpine program, that confirmed it for me. For me rope distance ceased to become a function of anticipated crevasse width. I now go even closer in AK too. I think the most likley consequences of a crevasse fall is usually trauma to the victim - not "the whole rope team getting sucked in". The circumstances of my crevasse fall in 2000 should never be used as an example of how far apart to tie in for glacier travel. We should have been in short rope mode - and thats what I would do if I could go back to that moment on Sept 28th, 2000 (at about 1:45pm in the afternoon to be exact).
At some point that day as we descended, the snow got firmer and the crevasses became more open. It was probably around 8000 feet on the coleman glacier. S-R mode would have allowed me to route-find better and just end-run the crevasse (which was the most logical and "safe" way of solving that particular guiding problem that day. Long-roping is a great technique for crossing glaciers where hidden crevasses comprise the greatest hazard, but in our case most everything that could open was already open. Short-roping would allow a guide to routefind much better than trying to long-rope with a client 80 feet in front of you leading the way (on that piece of glacier, I thought
about going first but the risk of slips and falls was high - thus another reason to S-R and not L-R.
The other thing to remember about your blog topic is that crevasse falls are not the only big hazard on glaciers. Slips and falls on steep terrain are sometimes more severe, and a rope team moving together while tied in far apart is not well prepared (in my opinion) to deal with those hazards.
So, to make a long story longer, when I am trying to figure out how far apart to tie my rope, I think about a few of the following variables (above and beyond crevasse size, numbers of people, etc...)
- Is long-roping more appropriate or less appropriate than short-roping right now? (L-R when soft snow, bad vis, lots of hidden slots, etc... S-R when late season, Firn, most slots are open, route-finding in complex (but not whiteout) terrain.
- Is the snow firm or soft? I might short rope routes these days in the morning, but then L-R them in the afternoon when they have softened up.
- How fit are the clients and how important is my communication with them and with each other? - if client safety benefits more from communication, pacing, direction, etc then I might go shorter rather than longer. Its a lot easier to remind your clients to keep the rope tight when they are 25-30 feet away than it is when they are 40. Its also easier to keep the rope tight, as well as direct, warn, caution, etc... I have gotten so sick and tired of tying my clients in 40 feet apart only to see them tripping on my rope with their crampons. They sometimes do such a bad job monitoring rope tension that the additional distance ceases to become a benefit. And you know this is true for recreationalists (comprising much of your blog audience) as well.
- Is adjustability important? Build a system into the rope team that allows one or more members to drop an intermediate knot - thus extending themselves temporarily - It isn't that hard to teach even the greenest of clients. If we are approaching a monster slot where I am afraid of its strenght and of having 2 or more people on it at once, its easy to go quickly from 25 feet to 40 feet if neccessary - or even switch to a belay-from-anchor. Furthermore, I think there are hardly any monster slots out there (bigger than 20-25 feet or so) that don't manifest themselves on the surface some how. The biggest slots I have ever seen outside of AK are down in Antarctica, and they always reveal themselves one way or another.
Instead of asking myself "how long should I go with my rope distance" when I rope up with my clients, I now try to ask my self 1): Is a rope even necessary? (because sometimes it isn't of course - or sometimes the presence of a rope makes things more dangerous than less - I asked my clients to unrope on a relatively crevasse-free glacier on Mt Blanc this summer because I felt the risk of rockfall from above far outweighed the likelihood of an unroped crevasse fall. 2): How short can I safely go? I bet there are plenty of days on Baker late in the season where one could short rope the vast majority of the route and argue logically that it exposed the team to less risk than long-roping.
The caveat of all of this, and the reason that it is on my mind so much (besides me being the lucky survivor of the story you told) is because of all the 14 guide fatalities in France last year, many of them were due to crevasse falls. Many euro guides can be seen short roping here and there on heavily crevassed glaciers, and it really makes me wonder sometimes... I find myself long roping lots of terrain that fellow french, italian, or swiss guides might be short roping on. There was an inquiry at ENSA (the french guide school) last year after many of the accidents and the ENSA instructors were asked if they teach new guides to S-R the glaciers. "No" they said. "We teach our candidates to use longer distances of rope when the risk of crevasse falls is high" they said. "We don't know why these guides learn one way with us then do something completely different (and much less safe) when they complete their UIAGM diploma".
That about wraps her up... Sorry this isn't very short but as you might know I have a very personal connection to crevasse falls and crevasse-related risk management. Plenty of people probably don't agree with me by the way, but I'd happily maintain that there is little or no evidence to support the claim that more rope in the team equates to less crevasse fall risk.
I witnessed two crevasse falls in a 30 minute period while guiding the Dufourspitze in Switzerland last summer. I rescued both victims myself. One team (a czech team) tied in a about 35 feet, and did a poor job of watching their tension. A girl popped through a weak bridge in the dark and yanked the other two guys off their feet. She went deep-until she corked. If she hadn't corked she might have dragged them both in. 15 minutes later an Italian climber passed me as I was probing a suspicions area (he wasn't concerned). He and his (single) partner were tied in 25 feet apart. He fell in - and the tension went immediatey onto his partner - who self arrested (crampons on, toes in the snow! - It works!). His partner was dragged a little bit but not much, and successfully stopped the fall. We hauled that guy out too. One of your blog points was that smaller teams should tie in further apart than big teams. I disagree (with some exceptions). I think they can go just as close (and use heaps of butterfly knots). I honestly believe the benifits outweigh the disadvantages, and overall - if people are skilled and aware - going a little closer is often safer.