Snow Anchors

Don Bogie

5/07/2005 Page 1 of 20 Snow Anchors

Background
The Coroners court finding for the four fatalities that occurred on Mt Tasman in December
2003 asked the New Zealand Mountain Guides Association and the New Zealand Mountain
Safety Council to look into whether current practices with snow anchors in New Zealand are
adequate. This document is a look at what is currently known about snow anchors in New
Zealand. Over the past 12 months I have been involved in snow anchor testing with Ruapehu
Alpine lifts, NZMGA and the DOC Aoraki mountain rescue team. That testing looked at ideas
from the Fortini presentation “On the Use of Pickets and Flukes as Snow Anchors” given by
Art Fortini of the Sierra Madre Search & Rescue Team to the International Technical Rescue
Symposium, Denver, CO, November 2002 and at current snow anchor techniques in New
Zealand. Fifty-one tests were carried out at these three sites in a variety of snow conditions.
Tests were carried out in fresh soft moist snow, wet spring snow, hard cold snow and low to
moderate strength cold dry snow. All testing was done with gradually increasing steady pulls,
none were done with dynamic loads. While the number of tests carried out can not be
considered high enough to give good statistically valid results for individual pieces of
equipment, they did show the phenomena described in this document and give us an idea of
the likely performance of different styles of anchors. It also highlighted a number of issues
with the material strength of some anchor and attachment materials and with the orientation of
V profile stakes.

Introduction
The strength of snow anchors is dependent on a combination of
• The strength of the snow in compression and shear
• The strength of the materials the anchor is made of
• The strength and placement of the attachment system
• The mode of use, angles and orientation of the anchor and its attachment system

We need to look at each of these and at their effect on each other when looking at anchor
systems. We also need to examine how strong anchors need to be for different
mountaineering and rescue tasks so that we can build appropriate anchors for those tasks.

Definitions
The following terms used in this document are defined as:-
Standard sized stake:- This is 60 cm long and is made of right angle section aluminium
(referred to as V throughout the document) with 5 cm wide sides which gives it an overall
width of 7 cm. Its area is 0.04 m2. The area is calculated by length times overall width less the
area of the points. The surface area of the sides is not taken into account.
Top clip:- Any anchor attached at its top.
Mid Clip:- Any anchor attached at or near its middle
Upright:- Any anchor put in perpendicular to, or at an angle back from perpendicular to the
snow surface.
Horizontal:- Any anchor that is put in horizontally at right angles to the direction of load.

5/07/2005 Page 2 of 20 Snow Anchors

How strong does an anchor need to be?
For rescue this is relatively simple. The anchor should meet the NZ LandSAR standard of
being 10 times the rescue load. For a rescue sized load of 2 kN this means an anchor should
be at least 20 kN. This means that 2 by 10 kN anchors or 3 by 7 kN would satisfy these
requirements if they were tied together with an equalised system.

For climbing things are more complex as we are dealing with dynamic loads. The UIAA
standard for a climbing rope is that no more than 12 kN of force can go onto an anchor for a
fall of 4.8 m with 2.6 m of rope with a weight of 80 kg. This is based on the maximum short-
term load a human body can handle with out damage. Most modern climbing ropes out
perform the standard by a significant amount with some claiming figures of between 7 and 8
kN. In real life the actual load on an anchor is affected by a range of things some of which
lower the load and others that make it worse. They include friction, age of rope, number of
previous falls it has held, how wet the rope is, what the actual weight of the load is, slope
angle, type of belay and fall factor.

Fall factor, which is the distance of the fall divided by the amount of rope that is between the
belayer and the falling climber is the key predictor of load. With falls of the following fall
factors the likely maximum loads are; ff 0.5 -- 6 kN, ff 1.0 -- 9 kN, ff 2.0 --12 kN.
(Information comes from the Petzel web site) The figures are for a vertical fall for the
standard UIAA 80 kg mass. There is no reduction for friction between the falling climber and
the ground. If friction and the effects of slope are taken off then loads can be 30% less for a
45o slope. That means the range of maximum loads on a snow anchor is likely to be between
4 kN and 8kN. If used as a runner, force is 1.6 times the maximum that is expected on the
belay anchor.

The sort of uses that snow anchors get make it unlikely that one would receive the maximum
sized loads. They are mainly used for belaying on moderate angled slopes, belaying over
crevasses and abseil anchors. It would be unusual for a snow anchor to be used as the only
anchor for a situation where a lead climber could take a fall factor 2 fall on steep terrain. The
uses snow anchors are put to would give us the likely following loads for these situations.
- Leading on 45o slopes and belaying over crevasses- 4 kN to 8 kN
- Runners - 8 kN to 12 kN
- Top roping 1 person or abseiling- 2 kN to 3 kN
- Top roping 2 people- 3 kN to 4 kN
Ideally anchors should be able to handle 12 kN. However there are material strength issues
with cables and stakes themselves which make this hard to achieve. It would therefore seem
that although there are some circumstances when snow anchors could receive over 10 kN that
this is unlikely for the type of use they normally receive so a target of 10 kN for the strength
of the components in a snow anchor would be a more realistic target to set. Some snow
conditions will limit anchor strength to less than 4 kN when using conventional sized anchors.

5/07/2005 Page 3 of 20 Snow Anchors

 A particular circumstance when
Force on a person in an Avalanche
o
a snow anchor could be
•40 cm thick slab on a 45 slope
•Assumes a person goes into a arched backwards position hanging from a waist
subjected to loads of greater than
2 2
harness that gives an area 0.4 m and debris density starts at 350 kg m and drops to
2
10 kN is if an avalanche hits the
200 kg m as the snow accelerates
person being belayed or tied to
14
an anchor. For large paths forces
12 are likely to be higher than in the
diagram to the left but most
10
climbing avalanche accidents
8 typically occur when the
kN

6
climbers set off an avalanche
themselves. These avalanches
4 are usually not particularly large
2
and the climbers are normally
high in the start zone. If someone
0
0 5 10 15 20 25 30 35 40 45 50
is less than a rope length from
Distance M the top of where an avalanche
started this is probably a
reasonable representation of likely loads. Therefore a 10 kN anchor is likely to hold someone
if they are being careful with how they are traveling and belaying in avalanche paths.

Key point – Be aware of the maximum load your anchor could be subjected to and build them
strong enough to cope with that.

Snow Strength
Snow fails either in shear or compression with snow anchors.

Compression Failure In compression failure the anchor pulls forward

through the snow. Under a steady load this can be a
fairly slow movement. The compression strength of
a snow anchor is dependent on the compression
strength of the snow, the size of the buried object
and whether the load is evenly spread over the buried
Top View object.

The Anchor pulls forward through the snow

In a shear failure, a stress cone in the snow is formed around the buried object. It goes out
from the sides of the object at approximately 45o and up from the bottom of it at
approximately 30o. The stress cone phenomenon was described by Fortini in his presentation.

When it fails it does so fast and the

snow cone and anchor come out of the
snow in an explosive manner. The shear
w strength of a snow anchor is dependent
d 45o
θ Side View φ on the shear strength of the snow and
30o Top View the surface area of the stress cone. The
size of the shear cone is a lot larger than

5/07/2005 Page 4 of 20 Snow Anchors

 the buried object. It can be over 50 times the size of the buried object depending on its depth
and width. The size of the stress cone can be calculated by working out the surface area of the
three surfaces that make it up by using the following formula from Fortini’s presentation.

Wd d 2 tan φ d2
As = + +
sin θ [tan θ sin θ ] [cos φ tan θ ]

Observations of stress cone failures such

as the photograph to the left in snow that
has been manually compressed do not
show clean shears of the entire cone.
Weaknesses obviously exist inside the
compacted snow, which mean that the
actual shear surface is less than the
theoretical area and therefore the strength
will also be weaker than the maximum
theoretical strength. However failures
under shear are still relatively strong
when compared to anchors failing under
compression.

Upright Horizontal An important feature of this is that

Stress Cone Area in

8 increasing the depth of an anchor has a far

larger effect on the size of the stress cone
6 than increasing its width. The graph to the
left shows the theoretical stress cone size
2

4
M

for a standard sized snow stake. When

2 used as an upright anchor it is 7 cm wide.
0
When used as a horizontal anchor it is 60
cm wide. A T-slot using a 60 cm stake
40 50 60 70 80 90 100 would need to be 52 cm deep to have a
Depth of bottom of anchor in cm bigger stress cone than the stake in as an
upright mid clip.

If we look at the compression strength of snow as described in the Avalanche Handbook we

can see that it varies hugely from < 1kPa for fist hardness to >1,000 kPa for knife hardness
snow. Each step in the snow hardness scale is an increase by an order of magnitude. Snow is
generally about ten times stronger in compression than in shear. As the shear cone provides a
greater surface area for the force to work on this can in certain conditions make for a strong
anchor even though shear strength is only about 10% of compression strength.

5/07/2005 Page 5 of 20 Snow Anchors

 Compression strength Theoretical Compression
strength, if load evenly
spread for 60 cm x 7 cm
2
stake with area of 0.04 m
Hardness Pa k Pa Range k N Range
3
fist 0 - 10 - 1 - 0.04
3 4
4 finger 10 - 10 1 10 0.04 0.4
4 5
1 finger 10 - 10 10 100 0.4 4
5 6
pencil 10 - 10 100 1,000 4 40
6
knife > 10 1,000 > 1000 40 >40

Average of Strength Snow Water

Snow Hardness when tested Dry Moist Moist to Wet Average
4 finger to 1 finger 4.7 4.7
1finger to pencil 3.5 3.5
pencil 12.1 7.7 9.5
knife 12.1 12.1

The test results to date of mid clip similar sized anchors shown above are in line with the table
above them although the wetter, less hard snow gave a higher average than the strength table
indicated it would. The failures above 10 kN were mainly the stakes attachments breaking or
shear failures.

Snow will fail when compressive stress exceeds compressive strength or when shear stresses
exceed shear strength. In weak snow compression failure is the predominant failure
mechanism. In strong snow shear failure is the predominant failure mechanism if the load is
evenly spread across the anchor. If the load is not spread evenly then this usually means that
the compressive stress at the point of greatest load exceeds the compressive strength of the
snow and failure occurs in compression. This is the main failure mechanism in upright top
clips.

Snow strength is dependent on the number of bonds (the more the better) and the strength of
each bond. The key factors are that; larger crystals have fewer bonds than small crystals for a
given volume of snow, bonds are weaker in the presence of water in the snowpack, wet grains
are strong when in their frozen state and bonding occurs rapidly close to 0o. We can in some
circumstances increase the strength of the snow and therefore the strength of a snow anchor
by compressing the snow that we build the anchor in. Compressing snow increases the
number of bonds by pushing crystals closer to each other. The useful strength of this
compressed snow then depends on the speed of bonding. In moist snow (which is at 0o) this is
very rapid. In wet snow (which is also at 0o) bond strength will usually deteriorate over time
even though more bonds have been created through compression. In colder snow < -5o
bonding is very slow and may not occur fast enough to be of any assistance in making an
anchor stronger. Strength generally correlates to hardness and hardness generally correlates to
density. The key exception being for wet snow where although it is dense it also has low bond
strength.

The best test of whether compressing snow will make the anchor stronger is to make a
snowball. If squeezing hard will make a solid snowball then you will create strong snow and

5/07/2005 Page 6 of 20 Snow Anchors

strong anchors. If the snow crumbles which occurs with cold snow it is unlikely to produce
stronger snow and may in fact destroy existing bonds and make it weaker. If water drips from
a snowball then compressing the snow may give it higher density but bonds are likely to be
weak and they may break down rapidly.

Key Points
• If you can make the snow stronger by compressing it then do so as you will get a stronger
anchor
• If you find yourself needing to use snow anchors a lot in weak snow, then you should be
using anchors with a large surface area

Snow Anchor Materials

It is important that the materials used for a snow anchor and the attachment methods used
with them are strong enough to handle the potential loads on them and to maximise the
strength of the snow they are in. As snow anchors should be able to handle a load of 10 kN
then the materials used need to be able to cope with this. Observations of snow anchor
failures shows that many currently used snow anchors bend or break at substantially less than
this.

In order to keep weight down on snow anchors aluminium alloy is usually used. Ski areas on
Ruapehu use a mixture of stainless steel and mild steel for their rescue stakes in order to cope
with the icy snowpacks that are common there. The 1.5 mm thick V section stainless steel
anchors appear to have very similar strength characteristics as 3 mm V section aluminum
stakes, while the 6 mm V section mild steel stakes are significantly stronger but are too heavy
to be considered for mountaineering use.
Curved inner corner is Aluminum comes in a variety of different strengths. The
likely to be high tensile
higher strength 6261 T6 alloy used by Aspiring Enterprises in
6261 T6 angle
its snow stakes comes close to the 10 kN strength being sought
but falls just short of this when used as a top clip in hard snow
and is well short of this with the point of the V to load as a mid
clip in weaker snow.

The shape of the stake section did not seem to matter from the
point of view of gaining maximum strength from the snow
when used in strong snow. However not enough tests have
been carried out to confirm this. It is possible that the shape of the object could change the
angle the stress cone comes out from the object at. Shape may be more of a factor in weaker
snow but it has not been possible to test this yet. The theory has been that by placing a V
stake point of V to load in all circumstances is that it is more stable when being pulled
through the snow and that it can create a bow wave effect that compresses and strengthens the
snow in front of it, which makes the anchor stronger. However if it were possible to
strengthen the snow by compressing it, then it would be better to do it manually when placing
the anchor and know that you have created a stronger anchor than to rely on an unknown
amount of compression from a moving stake to do this.

5/07/2005 Page 7 of 20 Snow Anchors

Width and length of the anchor are important for getting the area of snow that gives
compression strength or producing the size of the stress cone. The shape and orientation of the
section is important for determining the structural strength of the anchor material when under
load. In very strong snow when an anchor is pulled from a mid clip the strength of the stake is
not an issue as it is supported by the strength of the snow and the weakest link becomes the
strength of the attachment system. In weaker snow the snow does not give this support, so
stakes that do not have sufficient strength (stiffness) will buckle and pull out through the
snow. This has been observed at loads of around 7 kN with standard sized stakes.

In hard snow where a top clip is being used stakes pull forward under load as the anchor
bends and the snow fails in front of the upper third of the stake. The testing at Plateau in May
2005 in knife hardness snow showed that the stronger the material used in a snow stake the
higher the load it could handle. With the weaker materials and the narrower (5 cm wide) MSR
Coyote the failure was in compression. With the strong wider stakes shear failures were
observed. In order for either of these to happen the upper part of the snowstake has to bend as
its lower half is under very little load and is held firmly in the hard snow.

Five upright topclips in knife hardness –10o C snow. Left to right:- Test 1 Al alloy (6261 T6) failed in shear at
9.4 kN. Test 3 Al Alloy failed in compression at 6.7 kN, Test 5 was open part of V to load and failed at 4.4 kN
through buckling of the stake. Test 6 was 3mm mild steel which failed in shear at 11.3 kN. Test 12 was a MSR
coyote that failed in compression 7.8 kN

To get maximum structural strength from a V section stake the point of the V needs to point
towards the load when used as a top clip and the open part of the V needs to point to the load
when used as a mid clip in weak to moderate strength snow. In very strong snow orientation
does not matter for mid clips as the snow provides the strength. This orientation would also
apply to other types of open sections such as C-section.

There is a difference in strength between the 6261 T6 alloy and the weaker alloys. Anyone
using snow anchors made from the weaker alloys needs to be aware that their maximum
strength will be well below the 10 kN target figure.

The length of a snow stake contributes to its overall area and when used as an upright midclip
a longer stake gives a bigger stress cone. In order to gain the benefits of longer stakes the
material the stake is made of has to be strong enough to counter the effect of greater leverage.
Tests at Ruapehu showed no significant difference between 60 cm and 90 cm stakes used as
upright mid clips with point of V to load, both lengths folded in the middle at about the same
load. Stakes longer than 60 cm will also not make much difference to hard snow top clips
unless they are made from materials that are strong enough to offset the effect of higher
leverage on them.

Holes are often drilled into snow stakes in order to provide attachment points, to lighten them
and some people advocate it to provide grip. Holes can structurally weaken a stake so care
needs to be taken in order to not effect its structural strength. In testing at Plateau this year
several stakes that had holes 1/3 of the width of the side of the stake, bent and tore at the

5/07/2005 Page 8 of 20 Snow Anchors

attachment holes placed at the stake mid points. Any holes placed for grip or lightening
purposes may in fact reduce a stakes holding power in compression as they potentially reduce
the area of snow being compressed.

Any attachment point to a stake needs to be able to handle the anticipated maximum load.
This means any slings or cords need a breaking strain of over 10 kN after going over a sharp
edge and any knots is taken into account. The only way to be certain with whether a particular
combination of tape or cord will be strong enough is to carry out a strength test. In general
having tape or cord go over sharp metal edges is not recommended due to the possibility of
them being cut on that edge. If they are used then they need to have a large safety margin in
their strength in order to compensate for the edge issues. When stakes are used as top clips,
clipping a karabiner directly into the stake provides the most secure attachment point. With
the Coyote stakes karabiners can also be used for secure mid clip attachments.

Holes large enough for karabiners or to push webbing or cord through in the center of V or C
section stakes will tend to weaken the stake and compromise its ability to hold larger loads. If
slings are larks footed or clove hitched around the outside of a stake they can squash the stake
and cause it to bend at lower loads than other attachment methods. It is important to attach at
the middle in order to spread the load evenly. In weaker snow pulling from off centre could
lead to rotation which will cause the stake to pull out easily. A 4 mm wire cable passing
through small diameter holes and then swaged gives a mid point attachment that will stay in
the middle and not effect the strength of the stake. Using 4 mm cable limits the maximum
strength of the attachment to that of the cable, the Aspiring Enterprises web site rates their 4
mm cables to 11 kN. An alternative is to larks foot a wire strop through a pair of center holes
in the stake that are just big enough to get the strop through. This minimises weakening the
stake but unless care is taken with placement it can at times lead to the stake being twisted off
center from its long axis which results in less surface area in weaker snow situations.

Key points
• Stake material needs to be structurally strong enough to handle anticipated loads.
• Longer stakes are only stronger if the materials are strong enough to handle the higher
loads caused through leverage
• Minimise the numbers of holes in a stake as they could weaken it
• Only use attachments that have been tested and found to be greater than 10 kN
• If using V section have the point of the V to load for top clips and open part of the V
to load for mid clips

5/07/2005 Page 9 of 20 Snow Anchors

Placement of Snow Anchors
Top clip or mid clip attachments
Snow anchors can be placed in several different modes, as up right stakes, with attachments at
the top (top clips) or in the middle (mid clips) that can also be tilted back at different angles
and horizontally with attachments in the middle. (T slots)

One of the most critical things that effects

snow anchor performance is where you pull
from. An upright top clip is in engineering
terms a laterally loaded pile. In the Foundation
Engineering Handbook by Winterkorn and
Fang they say the following about laterally
loaded piles. “Piles are rather slender
structural elements, usually vertically inclined,
and therefore cannot carry high loads which
act perpendicularly to their axis.” If we look at
how the load is spread in the snow for an
upright top clip in the diagram to the left, then
we see that the majority of the load will be on
the snow in the upper third of the stake. The
actual cross over point from one side to the
other will be dependent on the stiffness of the
stake and the hardness of the snow. The
stronger they are the further down it will be,
which will increase the anchor strength as the
load is spread over more surface area.

If you pull from the center the load is more evenly spread. If the stake is strong enough it will
be even, but if it flexes then there will be higher pressure in the center. Pressure on the snow
with a standard sized stake being pulled from the middle with a load of 6 kN is around 150
kPa which is within the range of pencil hardness snow.

During testing at Plateau some upright top clips were observed to start failing in compression
at 6 kN in knife hardness snow. As knife hardness snow has a strength in compression of at
least 1000 kPa this meant that a pressure of greater than 1000 kPa must have been applied to
the snow in its upper third in order for this to happen. This is at least six times the pressure the
same load would be applying if pulled from the middle.

5/07/2005 Page 10 of 20 Snow Anchors

The angle of upright midclips
The angle a stake is placed in the snow has a different effect depending on snow strength.
What increases strength in strong snow has the opposite effect in weak snow.

The effect of stake angle on the area of the stress cone size (As) The depth of the bottom of
45o 0o the upright mid clip has a
major effect on the size of a
stress cone in stronger
snow. The size of the stress
cone decreases the further a
D Side View stake is leaned back. This
would weaken the anchor
As the angle of a stake is tilted back its depth decreases. The difference in depth
(D) means that the size of As is significantly reduced
in strong snow.

0o In weaker snow where

β compression is the
dominant failure
AL = 2L mechanism, leaning the
stake back will increase the
α anchors performance. As
L
the stake moves towards
Side View the load as it begins to fail
in compression, the angle
of the stake influences whether it pulls straight forward, dives or lifts up out of the snow.
• If α = 90o it pulls straight ahead
• If α < 90o it produces lift and the anchor will come up
• If α > 90o it produces dive and the anchor will go down
If a stake is leant back about 15o from perpendicular to the snow surface with a stake in at its
full length and an attachment coming out of the snow at twice the length of the stake (AL =
2L) it will form the right angle where the attachment meets the stake. Lifting up or pulling
straight ahead will both cause the anchor to fail at relatively low loads. If the anchor pulls
straight ahead the top of the anchor comes out of the snow and reduces the surface area of
snow in the upper half of the stake, which causes it to rotate forward and fail. Tilting the stake
back to make angle β, 45o makes the stake dive down into the snow and therefore put more
snow in front of itself to pull through. Under load they have been observed to travel down
slope by several metres and go down into the snow by more than a metre.

If being placed at 45o to encourage diving, users need to be aware that it can hit harder layers
and have the angle flatten out. This usually causes them to pull down slope at relatively low
loads until they pop out. Probing the snow with an object longer than the stake would allow
the user to identify possible hard layers that could cause this to occur.

There should be an angle somewhere between the 15o and the 45o that causes the stake to stay
just below the snow surface as it drags forward and prevents diving down into deeper layers.
This needs testing to confirm whether this can be controlled.

5/07/2005 Page 11 of 20 Snow Anchors

 This series of photos of an upright mid clip in at 10o back from perpendicular to the snow surface shows it
lifting from the snow at around 3 kN. A similar sized stake set back 45o pulled down into the snow and
failed at
Effect ofaround 7 kNanchors
multiple when the stake folded in the middle and the wire cable pulled out of it. The snow was
cold <-10o C and between pencil and 1 finger in hardness. It was not possible to compress it to make
stronger snow.

Multiple anchors
When there is a need to
Multi tool Anchor Separate tool Anchors produce a stronger
anchor than can be built
with a single piece of
Top View equipment a multiple
point anchor can be
built. If the pieces of
equipment are put in
close to each other then
Side View
an issue occurs with
overlapping stress
cones. Although
combining two tools produces a stress cone larger than one tool it produces less total shear
surface area than two separate anchors would because the stress cones overlap. There is also a
potential issue with using an upright axe in the multi-tool anchor as it is being pulled from the
top, which creates uneven load so it is possible that it is not adding to the size of the stress
cone, but is instead contributing to the compression strength of the anchor in an inefficient
way. It is also important with anchors that are likely to fail under compression to set them up
so that they do not pull through where another anchor was.

When building anchors with multiple pieces of equipment, maximum strength can be
achieved by separating them by at least twice the distance of the depth of the deepest anchor
and by making sure that you tie multiple anchors together with an equalised system in order to
achieve the full combined strength of the multiple anchors.

5/07/2005 Page 12 of 20 Snow Anchors

Effect of an upward force on snow anchors
Snow anchors are designed to be loaded along the surface of the snow. With the buried
anchors with midpoint attachments some upwards load will not be a big issue but it could still
weaken an anchor. With any top clip anchor any upwards force is an issue. We need to be
aware of the ways that an upwards force can go on a snow anchor and take measures to
minimise them. The following three situations could all cause lift and weaken the anchor:-
1. Having the rope to the load go up over an object like a pack or the belayers leg if belaying
off of the anchor.
2. Having someone lead past a snow anchor and fall after placing a runner or fall over the
other side of the ridge.
3. Belaying off of your harness in a sitting position with your feet in a good stance. When
the load goes on, the belayers feet act as a pivot point and their legs as a lever which
causes their waist to lift up from the slope.

There have been instances of numbers 2 and 3 above causing anchor failures leading to
serious accidents. Number 1 can be managed through good rope management and 2 through
building a multidirectional anchor (requires two snow anchors) if upwards as well as
downwards loads are expected. At Plateau in May 2005 an experiment was carried out to
simulate the effect of belaying off a persons harness while tied to an upright top clip anchor.
A piece of timber the length of an average sized persons foot to waist height was used to
simulate a belayer. This was done with the stake placed at 0.0 m, 0.5 m and 1.0 m away from
the person’s hip.

Simulated belayer in the system. From left to right. 0.0 m, 0.5 m, 1.0m back from belay point. All photos
were taken at point of failure

Anchor distance back from Approximate angle of pull Failure Load in knife
belay point above snow surface hardness snow
No belayer in system 0o 9.4 kN
1.0 m from belayer 20o 7.8 kN
0.5 m from belayer 30o 5.8 kN
0.0 m from belayer 50o 2.5 kN

This confirms what people are being taught, that if you are belaying off of your body with an
upright topclip then keep it some distance away from you. This experiment shows that an
anchor needs to be more than 1 m behind the belayer, probably at least 1.5 m to avoid
reducing the strength of the anchor.

5/07/2005 Page 13 of 20 Snow Anchors

The other mode for placing a snow anchor is horizontally, otherwise known as a T slot or
horizontal mid clip. If the snow can be compressed in front of them to produce stronger snow
they are very strong anchors. As stress cone size is heavily influenced by depth they would
need to be dug nearly as deep as the length of the anchor in order to get greater strength than
an upright mid clip that uses the same object. This would require digging and compressing a
large volume of snow. In snow that can be strengthened they will take longer to build than an
equivalent strength upright mid clip. In weak snow that cannot be compressed to make
stronger snow they become the only option for a snow anchor if you do not have an anchor
with a wire cable that can be pulled into the snow. In this sort of snow where moving the
snow damages snow bonds, a narrow trench needs to be dug for the cord or tape attachment.

For anchors failing under shear the main features that give it strength is the depth of part of
the anchor and evenly loading the surface of the anchor. For anchors failing under
compression the main features that give an anchor strength are its surface area, evenly loading
the surface of the anchor and having as much snow as possible to pull though before it comes
out of the ground.

Key points
• Pull from the center of anchors if at all possible
• Get the anchor as deep as is possible
• To get maximum strength from multi tool anchors separate them by at least twice the
depth of the deepest anchor and tie them together with an equalized system.
• In strong snow lean upright stakes back by no more than 10o in order to maximise the size
of the stress cone.
• In weaker snow upright stakes failing under compression will either:-
- move forward and out of the snow if the angle the attachment meets the stake is 90o
- move forward and lift up out of the snow if the angle the attachment meets the stake is <
90o
- or move forward and go down if the angle the attachment meets the stake is > 90o
• Leaning upright stakes back 45o from perpendicular to the snow surface will make them
dive and put more snow in front of the anchor, but be aware that the angle could flatten
out if it hits a hard layer, which would weaken the anchor.
• If belaying off your harness with an upright top clip then place the anchor at least 1.5 m
behind you.
• If using a T-slot in snow that will not get stronger through compressing the snow,
minimise disturbance of the snow in front of the anchor and cut the narrowest slot possible
for the attachment

5/07/2005 Page 14 of 20 Snow Anchors

Four different categories of snow
From the perspective of snow anchors there are four different categories of snow. Different
techniques are needed for each sort in order to get the strongest anchor for the conditions.
1. Snow that can be compressed to make denser snow
2. Snow that is hard that can not be compacted back into a trench but can have a slot cut
into it
3. Snow that is hard that can not be compacted back into a trench and can not have a slot
cut into it
4. Snow that is weak and cannot be compressed to make stronger snow. That is bonds
will not form easily through pressure. (Very wet snow or very cold snow)

1. Snow that can be compressed to make denser snow

There are two choices with this. Either the upright mid clip or the horizontal mid clip (T-Slot)

Upright mid clip

ke
o f the sta
ngth
ic e the le
e tw
m e nt to b
h
Attac
Side View

• Cut a trench that goes from the depth of the bottom of the anchor to the surface with a
shovel or ice axe adze in the direction of the load
• Compress the snow in the base of the trench and in the front of where the stake will go
in an area larger than the stress cone (out at 45o from the stake and up at 30o from its
base)
• Have an attachment (cable or sling) that is twice the length of the stake attached to the
mid point of the stake.
• Place the stake at 10o back from perpendicular to the surface so that the top of the
stake is below the snow surface. Have it well below the surface if you want to increase
its strength further.
• Backfill the trench compressing the snow in it taking care to make sure it is
compressed evenly from its base to the snow surface.

5/07/2005 Page 15 of 20 Snow Anchors

T-Slot

Side
View
Top View

• Dig a trench with a shovel or ice axe adze at right angles to the direction the load will
come from. It needs to be nearly as deep as the length of the buried object in order to
get as big a stress cone as could be achieved by putting the object in, in an upright
orientation.
• Place the stake horizontally in the trench at right angles to the load.
• Have an attachment (cable or sling) that is at least twice the length of the depth of the
buried stake attached to the mid point of the stake
• Compress the snow in front of the trench and in an area larger than the stress cone (out
at 45o from the stake and up at 30o from its base)
• Backfill the trench compressing the snow in it taking care to make sure it is
compressed evenly from its base to the snow surface.

Testing indicates that both of these should produce strong anchors of greater than 10 kN with
a standard sized stake if the snow makes a good solid snowball. If the snow pack is wet the
strength range is likely to be less than 10 kN. Actual strength will depend on the size and
strength of the anchor materials, how deep part of the anchor is and how well it is placed.

2. Snow that is hard that cannot be compacted back into a trench

but can have a slot cut into it

This style of anchor requires a wire cable attachment to the centre of the stake.
• Cut a slot with an iceaxe pick or snow saw for the cable
• Hammer the stake in 10o back from perpendicular to the surface until the attachment
point reaches the bottom of the slot. 10 cm to 15 cm of stake will be left protruding
from the snow if using an iceaxe pick.
• Pull the cable tight at the front
The reduction of depth of the stake through losing some length is made up through pulling
from closer to the centre, this more evenly spreads the load and reduces the pressure on the
snow near the surface. This now means that the anchor will most likely fail in shear. Although

5/07/2005 Page 16 of 20 Snow Anchors

the stress cone has been reduced by only having 45 cm to 50 cm of stake in the snow the shear
strength of the cone produced will be far higher than the strength of the full stake pulled from
its top.

Testing indicates that this should produce a very strong anchor of greater than 10 kN with a
standard sized stake in snow of knife hardness. Actual strength will depend on the size and
strength of the anchor materials and how well it is placed. The weakest link is likely to be the
attachment cable.

3. Snow that is hard that can not be compacted back into a trench
and can not have a slot cut into it

• Hammer the stake in 10o back from perpendicular

• If using a belay off of your waist with this style anchor attach to it at least 1.5 m away
so that any upwards force on it is minimised. If belaying off of anchor ensure rope
stays close to surface when under load and does not go up over anything.

Testing indicates that in knife hardness snow with a standard sized stake that this should
produce a reasonably strong anchor of 7 kN to 10 kN depending on the strength of the anchor
material and how well it is placed.

4. Snow that is weak and cannot be compressed to make stronger

snow. That is bonds will not form easily through pressure. (Very wet snow
or very cold snow)
There are two choices with this. If you have an anchor with a cable attachment they can be
used as snow pigs where an upright mid clip is leaned back so that it will dive down into the
snow. Leaning the stake back 45o will produce this. This diving effect can also be a problem
as the anchor could hit a hard layer and lean back further which will cause a loss in strength.
The other choice is a horizontal mid clip (T-slot). This is the only choice available when a
wired stake is not available or in very weak snow when a large object needs to be buried.
With both sorts of anchors in this sort of snow avoid damaging the snow in front of the
anchors.

Snow pig

5/07/2005 Page 17 of 20 Snow Anchors

 • Push stake into snow 45o to the surface and pull wire down into snow.
• Be aware that the anchor will pull down into snow under load and that it could lean
back further and become weak if it hits a hard layer.

T Slot

Side View

Top View

• Dig a trench a few degrees less than perpendicular to the slope as long as the object to
be buried as deep as is practical but at least 40 cm preferably 60 cm
• Cut a narrow slot at right angles to this for the attachment. The deeper the trench and
the longer the slot the more snow the anchor has to pull through before coming out.
• Backfilling is optional but may help the anchor sit in the right orientation when it is
not under load.

Strength is very variable with either style of anchor in weak snow that will not compress to
form stronger snow. Testing indicates you can get strengths of up to 7 kN with a standard
sized stake depending on the strength of the anchor materials, how well it is placed and the
bond strength of the snow. In snow softer than pencil hardness, 4 kN is likely to be the
maximum strength with a standard sized stake. There will also be times particularly in soft
cold snow conditions or in very wet snow when it will not be possible to produce a workable
anchor with standard climbing equipment. Burying a larger object as a T-slot, such as a ski,
snowboard or pack or large bag full of snow and getting that as deep as practical can be
effective as strength in this sort of snow is dependent on surface area.

Because snow pigs dive when the loads exceed the compression strength of the snow this
means that they will have more snow to pull through than a T-slot, which has a fixed amount
of snow to pull through. While this means they should be able to sustain a longer load than a
T-slot there is also a risk that they will strike a hard layer and flatten out which will reduce
their strength.

5/07/2005 Page 18 of 20 Snow Anchors

Conclusion
There are many snowstakes in use currently that would not be strong enough to handle the
upper end loads that a snowstake could be placed under. Some of the techniques that are in
use at present such as, not disturbing the snow in front of an anchor in any circumstances or
putting multiple pieces of equipment in close together or having the point of the V to load
when building a mid clip anchor do not let the users derive the maximum strength from their
placements. Although I have no direct evidence to prove it, I am fairly sure that a lot of users
do not have a good understanding of the strength of the snow anchors they are using and are
probably over estimating the holding power of the anchors they are building.

I think that because snow anchors do not come under the upper end of the loads they could be
subjected to (6 kN to 10 kN) very often, that catastrophic failures are infrequent with users.
However I would think that there are many people who are operating very close to the failure
limits of their snow anchors without realising that. If users of snow anchors were to adopt the
following objective and strategies then the chances of snow anchor failures would be
minimised.

Objective
To quickly produce a snow anchor that will not fail under the expected loads

Strategies to achieve this

• Be aware of the factors that effect snow strength
• Be aware of the likely loads for the situation
• Know a variety of techniques that will cope with the combinations of snow and loads
• Increase snow strength if possible
• Get anchors as deep as possible
• Pull from the centre of the anchor as much as possible
• In weak snow have as large an anchor as possible
• Use anchor materials and attachments that exceed 10 kN in strength
• Use good quality rope when climbing
• Use a 10:1 load safety factor for rescue

The three key things are to increase snow strength, get anchors deep and pull from the middle.

Further testing requirements

More testing is required in order to quantify aspects of snow anchor placement and strength.
The things that have been identified when doing this document are:-
• Working out the best angle to lean an upright anchor back in order to keep it below the
snow surface as it fails under compression.
• How much influence the shape or orientation of the buried object has on shear failure.
• How much influence the shape of the buried object has an on compression failure.
• The best materials to use for snow anchors and their attachments.
• The most appropriate size for a snow anchor.
• Whether further refinements to the snowball test or other simple tests can give users a
better idea of likely anchor strength.
• Whether anchors behave differently when subject to dynamic loads.

5/07/2005 Page 19 of 20 Snow Anchors

Don Bogie
[email protected]
DOC Canterbury
NZMSC Snow and Avalanche Committee
NZ LandSAR Technical Rescue sub-committee

Acknowledgements
• Ruapehu Alpine Lifts
• New Zealand Mountain Guides Association
• DOC Aoraki Alpine Rescue Team
• Twizel-Farlie Police
• Lindsay Main from Aspiring Enterprises
• Grant Prattley from SARINZ
• Art Fortini

References
Avalanche handbook; D McClung and P Schaerer, 1993, The Mountaineers/Seattle
Foundation Engineering Handbook; H F Winterkorn and HY Fang, 1975, Van Nostrand
Reinhold Company
Fortini presentation “On the Use of Pickets and Flukes as Snow Anchors” given by Art
Fortini of the Sierra Madre Search & Rescue Team to the International Technical Rescue
Symposium, Denver, CO, November 2002

5/07/2005 Page 20 of 20 Snow Anchors