What to do when one point in a climbing anchor breaks?

Question

Climbing anchors usually have 2-3 redundant protection points, which distribute weight. So if one point breaks, this will not be a catastrophic failure.

However, what should I do in practice if this happens? (this has never happened to me)

Hypothetical situation: I (let's call me A) have finished leading a pitch; I have built an anchor of 2 solid or 3 less-than-solid points, and I am belaying 2 persons (B and C) top-roping 10-20 m below me. Suddenly, one of the persons (B) loads the rope, and one of the anchor points is destroyed!

Assuming the anchor still works (1 or 2 points left):

• What should I do?
• What should I instruct B and C to do?
• If we cannot communicate (e.g. too much environmental noise), what to do?

Answer 1

tl;dr you shouldn't need to do anything.

The point of building redundancy is exactly for this scenario. Sometimes anchors fail. If you've built your belay correctly then loosing one anchor in a 3 point belay should have no effect on you or your partners. Just carrying on belaying. This is exactly why you must balance your anchors correctly.

If your worried about all your anchors failing then why are you belaying here at all! If your forced into this scenario (which seems very unlikely unless your leading a brand new pitch on Eiger) add more anchors (5+) and certainly don't belay 2 people at once.

What should I instruct B and C to do?

Nothing. What would be the point?! Let them carry on climbing as normal.

So if one point breaks, this will not be a catastrophic failure.

This is a little wrong, the point of redundancy is that if one piece fails, nothing happens, at all. Catastrophic or not. You should be able to continue without having to do anything, because trying to place gear or lower off on one piece of gear is dangerous. If you have any level of doubt on a piece of gear, you should be backing it up.

Answer 2

I disagree with the other answer. If you are anchored to 2 "solid" pieces and one fails, you are now in a non-redundant state and your goal should be to resolve that situation as quickly and safely as possible. If you are belaying a single climber that you can easily communicate with, try and convince them to stop climbing without weighting the rope (e.g., place a piece and clip in) while you tie them off and attempt to rebuild redundancy into the anchor. If you cannot communicate with them, maybe you can belay one handed and rebuild redundancy into the anchor with the other. The extra redundancy does not need to be beautiful, equalized and not allow for extension. The goal is to prevent death. even an extra piece of gear would be nice.

When belaying two climbers that you cannot communicate with, there really is not much you can do. The safest thing to do is to wait until you can communicate with the climbers (potentially when they reach the anchor) and then rebuild redundancy into the anchor. Hopefully you can do this before they clean the last piece of gear, since if the anchor fails, that is all that is keeping you from taking a long ride.

As an aside, simultaneously belaying two climbers is an advanced technique that limits your ability to deal with unforeseen circumstances (e.g., it is much more difficult to escape the belay when there are two climbers). Things only get worse if you cannot communicate and/or there is a potential for a fall. I would try and avoid that situation, and if it is unavoidable, I would really try and avoid belaying off a two-piece anchor in that situation.

Answer 3

While I technically agree with Liam's answer, I think there is two reasons why I would act differently: Purely for psychological reasons (personal piece of mind) and to keep redundancy consistent.

First let me constrain the scenario: I assume you haven't made a major mistake. One piece is out, so one could argue you made some kind of mistake - but that's ok, mistakes happen, having a system to catch them is the important thing, and that's what we are discussing. A major mistake would e.g. be using the same rock feature for two points and the one point failed due to that feature being loose (or loosened by the leverage of the camming device). In short: Lets assume there is no reason to doubt the two remaining points.

Now there also wasn't any reason to doubt the point that failed - still it did. So I'd try to add another point, to have a three-fold redundancy again. Now I know that argument is probably flawed if you consider the likeliness of two simultaneous failures. On the other hand there's also no reason to believe the failures are necessarily independent. Maybe it's a different type of rock than you are used to, you have a bad day (dismissing that as "you must not be climbing in that case" as often done is dangerous, this is not a binary condition), ... - essentially you just missed something that also applies to the remaining two points. In that spirit you would now adjust (lower) your trust level to those anchors and add another one.
None of these are any "hard arguments", it's very much up to the specific situation and as soon as you are at a safe point again, try to evaluate that situation and your reaction to it. Not to blame yourself, but to learn and maybe find something you can improve when you encounter this situation again.

And let me make a tangential point that will probably be a bit annoying due to being pedantic and knowingly misinterpreting words (because I am sure someone will misinterpret them), but I feel rather strongly about it, so I am doing it anyway:
You should always be worried about all your anchors failing. Even the best placed anchor can fail. There is no such think as safety, you are always managing a risk.
Now this has no immediate practically applicable implication, but I believe it's a very important mindset to reduce and accept risk successfully. Because if you are belaying two seconds on self-placed gear and are not aware, that there is a risk (maybe a slim one, but still), you are way more likely to make a mistake.

Most of this is just an opinion piece which is purely backed by having seen a few mistakes. And then when addressing the involved party, always in a neutral manner inquiring about why they did it this way (maybe there is a reason I am ignorant of - another mindset I like :) ), I got dismissed with some excuse along the line of a general "I am doing it safe, no need to think about it", which is plain dumb.