If you've ever tried to perform a crevasse rescue in a two person climbing team, you know it’s a very challenging operation. The two main steps of holding the fall and then constructing an anchor are made much more difficult if you are holding most of the weight of your fallen partner. (If you do choose to travel on a glacier with just a two person team, both of you need to be completely dialed in your rescue technique and have all necessary equipment.)

For some time now, it’s been generally recommended to tie brake knots in the rope when on a glacier as a two person team. The theory is the knots will catch in the snow in the case of a fall, minimizing the distance of the fall, and also helping to hold the weight of the victim, making the rescue initiation much easier.

There are a couple of downsides to this technique, namely it makes it more difficult to prusik up the rope, and can complicate rigging a mechanical advantage hauling system. This is generally true, but with a bit of practice you can overcome both these obstacles quite easily. For the hauling, if you’re carrying enough extra rope in your pack or even a second rope for rescue, you can drop this rope to the victim and use it to carry out the rescue and pretty much ignore the loaded rope that has the knots in it. Two, if needed, it's quite easy to pass a knot through a hauling system provided you have some modern gear like a Tibloc and Traxion pulley; learn how to do it here.

But, while a popular technique in Europe, this method has not fully caught on in North America.

To get some definitive data on this issue, rigorous testing was conducted by ENSA (École Nationale de Ski et d'Alpinisme) or French National Guide School. These are some of the most expert mountain guides in the world, and they have the knowledge and engineering tools needed to come up with some solid answers.

Fortunately, they made a video (in English) that shows their testing procedure and results, which should settle the argument once and for all.

(Note that this recommendation is for TWO PERSON teams only, not three or more.)

The short version:

“Our tests validated the effectiveness of this technique, and we strongly recommend climbers use it.”

While watching the whole video is recommended, here are a few takeaways.

3:15 - “Then I did it for real, sliding for 4 or 5 meters until the knots dug into the snow and held the weight. In real life, the belayer would then be completely free to make an anchor and start the rescue procedure, without being pulled forward by the person in the crevasse, and without the rope being continually under tension.”

4:12 - Brake knots are only effective when there is deeper softer snow above the ice layer. “When the same study was done with only 30 cm of snow over ice, this was not enough snow for the knots to properly dig in, and the knots did not work at all . . . The knots simply slid on the ice and didn’t properly penetrate into the snowpack.” However . . . “if the snow cover at the lip of the crevasse is 1.5 m to 2 m deep, then the knots are pulled deeply into the snow and end up blocking the rope.”

5:06 - “With knots, at the end of a fall, there is a maximum force of only 10 or 20 kg on the belayer, which means it's very easy to hold the person who’s hanging from the rope. This obviously makes the rescue procedure a lot safer.”

5:41 - “The tests showed that 3-4 knots were easily enough to reduce the load on the belayer. There's no point in tying more knots than this, because it will just use up a lot of rope.”

6::00 - “We recommend tying one knot 3 meters from each person, and then another two knots at 2 meter intervals.”

6:01 - “Brake knots should be big and bulky, so they brake effectively when they are pulled into the snow. They should also use as little rope as possible.”

6:20 - Pay attention knot geeks, here’s a new one here I bet you don't know - the “brake knot”, a classic figure 8 on a bight with one addition to give it a larger diameter.