Modern Anchor Theory and Snow Anchors

So I learned about snow anchors about 10 years ago, around the same time I was learning about rock climbing anchors.  At that time, the concept of "equalization" was something impressed upon me.  Times change, lab tests get done, and the cultural tide evolves in terms of how we think about anchor priorities.   My understanding of modern rock climbing anchor-craft "best practice" is to let go of the concept of equalization, focus on bombproof pieces, simplicity of construction, and thoughtful ways (if possible / practical) to share the load across the pieces (assuming that's even a priority for the climbing party relative to the available pro).

Based on what has seeped into the popular consciousness of climbing about how equalization isn't something that can be achieved (even load sharing can hard to accomplish depending upon what you're doing), how should I update my thinking (if at all) regarding snow anchors?  

For reference, I live in the PNW, and only have cause for using snow anchors for either belaying a sketchy crossing of a crevasse or feature, or the nightmare scenario of having to hoist a victim after a crevasse fall.   I'm more worried about what I do in the latter situation.  

My intuition is that ensuring as much "sharing of the load" across anchor points is far more desirable  since it can be hard to assess the "bomberness" of a given snow anchor.  Yes yes yes only build an anchor if it's bombproof.  But seriously, if someone fell into a crack and is all banged up, and we (maybe just I?) am up top going through the steps of building an anchor to which to transfer the load while also being mindful of setting up a haul system, where in my prioritization should sharing the load across two anchor points (such as a picket and an ice axe) be?  Theoretically, you're laying in the snow in this situation and once you transfer the load in some fashion it's gonna be real tough to rejigger the works if there's some slack in the system (e.g. the load is resting on one point despite ones' best efforts to distribute the load with slings or whatever's available).    Perhaps use of systems with some adjustment (such as munter mules) on pieces can allow finer tuning, or something.  

It's a rainy day in Seattle and debating hypotheticals that are rare occurrences sounds like a fine thing to do today.

I think equilization is far more important in snow than on rock because the anchors are so much weaker.  Even a good picket falls at way lower loads than a good cam let alone a good bolt. (granted you're usually asking a lot less of them....)

There are techniques to effectively back up a weaker component in an anchor after transferring the load to it and still achieve decent equilization, but it's not really worth it to try and explain it via the internet.  Guide's in the NW practice this stuff all the time.  I'd hire a guide for a day of snow anchors and crevasse rescue and show up with exactly these questions.  You'll almost certainly get higher value information than you will from this thread.  

Equalization is still an important part of quality anchor building. The rhetoric surrounding it lately is unfortunate and could give newer climbers the wrong idea. Of course you arent going to get perfect 33% load distribution on a 3 piece anchor, but partial equalization will still have a significant effect on its overall strength.
Yes, each piece should be bomber, and no, 10 garbage pieces will not equalize into a bomber anchor, but there is certainly a nuanced middle ground here.

I'm in Spokane, also stuck inside for the next couple of rainy days, so here I go:

I also climb in the PNW.  I'm very wary of pickets, which are really weak in all but the hardest snow - I think I've placed one in my life that I would actually trust as bomber lead protection.  You can T-slot a picket for additional strength, but they are still not nearly as strong as a decent screw or piece of rock gear.  In conditions where the snow is hard enough that pickets work well, I typically don't feel the need for a belay.  In snow conditions sloppy and steep enough where I want a belay, pickets just aren't that strong and aren't worth placing.  I still carry 1 or 2 pickets on snowy alpine climbs, but they tend to be a last resort for me for use in anchors or crevasse rescue situations.

If I was in a situation where I needed a snow-only anchor on technical terrain, I would T-slot and equalize two pickets.  I would also protect the anchor by belaying off my harness on a good stance.  In reality, I almost always am able to equalize a picket with an ice screw or a piece of rock gear, which makes me feel much better.  Sometimes this requires some downclimbing or simulclimbing to reach a place where a good anchor can be built.  Ultimately, if you don't have strong pieces, your anchor is not strong.  

There is no reason to believe that equalization is any more achievable on snow than it is on rock; however, redundancy is. That said, once you're at the point of questioning the snow anchor quality, it's doubtful redundancy will do anything for you.

I have done some load testing on snow anchors, and i can generally say that load distribution between multiple anchors should always a top priority.  in my experience, a single snow anchor, like a properly placed vertical picket, is only good for about 400 lbs before it pulls.  the problem is, when it pulls, it pulls suddenly and violently.  a fluke, for instance, has a more benign failure mode.  when properly placed, they fail around the same loading, but will "dive" into the snow pack and still support some load (assuming of course they don't find a weak layer of facets and pull right through it).

in many scenarios where i need to solely rely on snow anchors, this is my general approach:

1. build the appropriate anchor based on the expected load and snow conditions (vertical picket, t slot, fluke, deadman etc.)
2. load share with another appropriate anchor (again, based on conditions)
3. try to reduce the load as much as possible
4. when the load is applied, have someone monitor the anchors

rarely will i ever use a 3:1 on pure snow anchors.  2:1 is much more common because i can reduce the load the anchors see.

http://web.mit.edu/sp255/www/reference_vault/McMillan_how_strong.pdf
page 9ish

Folks seem to get hung up on load "equalization" and true enough, equalization is unlikely. But load sharing is a thing; even if the shared loads are not equal, sharing buys a significant margin. An 8 kN fall arrest load that is shared 5 kN and 3 kN is easily sustainable by most trad anchors.

One of the challenges of snow anchors is that the snow can easily be much weaker than the anchor.

FosterK wrote: There is no reason to believe that equalization is any more achievable on snow than it is on rock; however, redundancy is. That said, once you're at the point of questioning the snow anchor quality, it's doubtful redundancy will do anything for you.

I am not sure if I really agree with this.  In general, when you look at something in terms of 3 dimensional stiffness you will get better distribution through decreasing the stiffness.  Snow is obviously less stiff than rock.  It will depend on how the pickets interact with the snow.  Does the snow increase in resistance as the picket starts being dredged through it, or does it immediately blow out?  Does it increase in resistance to a certain point and then lose it's resistance?  It will depend a lot on the snow.

I think the current zeitgiest isn't that there is no point the equalization as much as it is the least important and hardest to achieve in actual practice.   You shouldn't sacrifice things like redundancy or "no extension"  or timeliness to get equalization.   I haven't seen people saying you shouldn't do things like tie your masterpoint in the expected direction of pull.  It costs nothing time wise and adds no weakness.... And it might help. However, when it comes to rock it is bad practice to put up a bunch of crappy anchors with the hope equalization will save your bacon as acheiving it is unreliable.  Falls happen at unexpected angles and a line up is almost never straight.  That said it may be all you have and in those cases you best try and equalize as it is better than nothing.

 Multipoint snow anchors are a different story than most rock anchors in that they are usually a rescue anchor with a somewhat static load in a known  direction of pull that is unlikely to change.  It is much easier to share and equalize an anchor in this situation.  Due to the weak substrate it is more important as well.

slim wrote:

I am not sure if I really agree with this.  In general, when you look at something in terms of 3 dimensional stiffness you will get better distribution through decreasing the stiffness.  Snow is obviously less stiff than rock.  

This is fair I think: deformation of the snow will result in a more distributed load than would otherwise be expected on rock anchors.

 It will depend a lot on the snow.

This is the crux of the issue - if the snow is weak, I'm not sure any level of redundancy or equalization will fix this. In the context of good snow with short pickets in t-slots or deadman anchors (buried bags or axes), then attempting to distribute the load across multiple points is probably worth the effort as smaller pieces are weaker. I wouldn't necessarily do this if I was burying my skis in a t-slot. But if I have bottomless facets, or isothermal snow, there is no strength in the system that a redundant or distributed anchor is going to recover.. 

If the snow is isothermal mush you just can’t fall. Seriously a pickets surface area isn’t large and flukes structural rigidity is questionable.

I also worry about vertically placed pickets functioning as intended. There was a good study by a New Zealand group on vertical pickets and clipping location, it’s worth looking up.

In good snow a t slot (deadman) picket is great. I fell through a bridge looking for a partners errant ski on baker a few years ago. Was around a 25 foot fall total with the rope cutting FAR into the snow (reducing load on anchor I’m sure). Had one picket in a t slot around 30” down in firm snow, cut a long slot to allow runner to not levarge picket up. Partner had me on belay in a stance, was ripped right out of the hole and slammed onto the anchor when I fell in. I did have them back that one picket up with their other ski before I got out. 

jaredj wrote: For reference, I live in the PNW, and only have cause for using snow anchors for either belaying a sketchy crossing of a crevasse or feature, or the nightmare scenario of having to hoist a victim after a crevasse fall.   I'm more worried about what I do in the latter situation.  

You need to also consider expediency.  For example, building an equalized multi-piece snow anchor for every significant crevasse crossing may be completely unrealistic on some routes, due to the massive amount of time it might add to a climb.  Because it should be a low factor fall, for things like crevasse crossings I usually will simple get into the self arrest position, or do a simple fast anchor like a stomped-in ice axe (obviously it's very situational).

Hauling is potentially a much higher load activity, so it requires a stronger anchor.  I think equalization is reasonable and valuable in this specific context.  I'd also note that even in a crevasse fall scenario, hauling is usually the last resort, there are other better ways to resolve the situation that don't load the anchor nearly as much (prusik out, climb out on belay, lower to the bottom, etc.).

There's a decent amount of research on snow anchors and strength as well as snow travel. This folder has a reasonably summative pile of the publicly available research:

https://drive.google.com/drive/folders/1lIY2C_y3Pkg_Jy2OeepIqXhvZEMwJaSo?usp=sharing

The general summary for snow anchors:
-A picket can generally hold no more than 4 to 5kN.
-For 1F to P hard snow, use vertical mid-clip pickets oriented 30 degrees back from perpendicular to snow surface.
-For P to K hard snow, use vertical top-clip pickets, roughly plumb. (P to K hard snow is generally supportive of an individual in boots; crampons may be needed to travel on it.)
-For work-hardenable snow (moist/warm; can make a snowball) , 4F snow can prove viable for mid-clip pickets after it's been work-hardened (i.e. stomp on it) and easier/faster than T-slots.
-For all other conditions, use a T-slot.
-Don't count on most ski anchor configurations (I, N, X) for more than about 2kN. T and H anchors are more comparable to picket strengths.
-Upright skis in hard snow buried to the binding and configured similar to a mid-clip picket can be quite strong.
-When moving roped together on snow, don't count on being able to hold a fall on terrain steeper than about 30 to 35 degrees depending on snow, number of people on the rope, and direction of travel (uphill/downhill).

Note the mid-clip pickets mentioned above. These are different from traditional top-clip pickets both in construction and the point of attachment. They are referenced in the literature I linked. As far as I know, Yates is the only place you can get them in the U.S. See the "cable picketts" here: http://yatesgear.com/climbing/ice/index.htm#6

For the specific question of crevasse falls, as Kyle notes, exhaust all the non-hauling options first. If you must haul, a 6:1 haul will bring the load on the anchor dangerously close to the strength limit in good snow (about 4-5kN); however, other factors such as friction of the rope against the snow can serve to reduce the load. Consequently, if you need the increased mechanical advantage of a 5-6+:1 hauling system, adding an additional anchor to share the load isn't a bad idea. Here's how you might set that up:

1. Hold the initial load via self-arrest and then with your legs, per normal.
2. Create initial anchor (t-slot, vertical picket, etc. as appropriate; don't use your axe as you'll need the thing for the next anchor point)
3. Carefully shift load off your body and onto the anchor. Verify it's holding the load appropriately. In general, this anchor will only see a static load of one person, or perhaps double the weight of one person if they bounce around while ascending.
4. Now freed from the load, walk back from the first anchor and build the second anchor. This will ideally be 4+ feet directly behind the first anchor and in line with the load.
5. Using an untied cordelette, clip a bight to the second anchor and then run it through the carabiner currently holding the load on the first anchor. Run the cord back through the carabiner on the second anchor, then back to the first, then back to the second, etc. until no material remains. When you're done you should have a continuous loop between the two anchors; this creates a block and tackle mechanical advantage system.
6. Apply a small amount of tension to the block and tackle (just pull on the end of the cord) to ensure load sharing to the second anchor. Secure with a mule hitch, mariner's, or your favorite way to hitch a load.
7. Do whatever else it is you're gonna do.

In step 4 above, you'll want some distance between your anchor points. Basically, the snow that supports the load is a cone shape with the apex at the anchor point and cone extending outward into the snow. The distance the cone extends is somewhere in the ballpark of twice the length of the picket/object used (so 4+ feet for a standard 2ft picket). If you don't have enough cord to cover that distance to create the load sharing, you can offset the second anchor to the side, but this will not be ideal as an increasing angle reduces the maximum load the two pickets can support when combined.