Testing a method of failure with a Bowline-on-the-Bight knot

A recent video clip has come to light on UKCaving Forum from the French Caving School which identifies a method of failure for the Bowline-on-the-Bight knot. The School’s film shows a failure that could lead to serious injury or even death when clipped into only one arm of the Y-hang.
I have not seen or heard of any occasions where this failure has happened in the U.K. and have certainly not seen it with my own eyes until today’s tests. It is rare for such a statement to come in regards to such a well used technique so I had to investigate.
We made use of the excellent training facility at Pindale Farm in the Peak District and the not so excellent weather.

Background:
The Bowline-on-the-Bight or ‘BotB’ knot is widely used and taught as a preferred knot for most SRT rigging applications involving Y-hangs and SRT. The knot has certain shock absorbing and self equalising properties that make it a good choice. These traits form one side of a double edged sword, the other side being the knot is easy to undo and therefore easier to loosen in use. It is this property of the Bowline knot family that can lead to slipping failures like the one we are looking into here.
For anyone who does not know the BotB knot, I will not explain it here as there is a wealth of literature and information on them online. Spend 5 mins Googling or pick up a copy of a UK caving manual.

The Failure:
The problem occurs when a caver is clipped into one arm of the Y-hang formed by a BotB. The two arms are formed from different parts of the knot, one coming from the pitch rope up into the knot, twisting round and emerging to form a loop and the other from the traverse line or on occasion a stopper knot.
Being clipped solely into the arm formed from the pitch rope can cause the rope to pull through and effectively untie the knot when loaded by the caver. If on a straight pitch with no other attachment the caver could fall to the floor. Even with a re-belay or end of rope knot present to stop the rope being pulled totally up the slip can still occur. I ruined a length of rope today, wearing through the sheath on a slip of less that 2 metres.

Observations:
We managed to work through a set of differing conditions and factors to arrive at a scenario where we could generate an almost 100% failure rate. We used the same rope throughout the tests, although the knot was tied in different sections to prevent too much damage to the same area.
The rope was a 5 year old length of 10.0mm semi-static rope widely used by cavers. We also tested a length of smaller diameter rope from another brand with similar results. The rope was stiff but could still form and hold a knot well.
We varied the knot for each test in the following ways:

  • Dressed correctly or not
  • Tightened or not
  • Evenly loaded or not
  • Clipped into left or right arm
  • Clipped into both arms
  • 10 & 9mm rope tested
  • Wet and dry

In the outdoor world we use a accident analogy called the ‘Lemons’. If you imagine a slot machine is your work and the tumblers and their ‘fruits’ are the chance of events happening, the lemons are the poor practice, bad judgement or bad events.
As a lemon appears on a tumbler it is usually outweighed by a good fruit, i.e. a bad thing/poor practice is safeguarded by the other good practices along with it and the chance of an accident is low.
When all the lemons come up on a line together we have a series of poor decisions or events that collectively form a chain of events and choices which dramatically increase the chance of having an accident.
It is this analogy that I think fits perfectly with this method of failure in the BotB.

When the following conditions were met we experienced a near 100% failure rate.

  • Wet rope (very high rate of failure on dry too)
  • Knot not fully tightened
  • Y-hang arms not equally loaded – specifically the one formed by the pitch rope
  • Caver only being clipped to the loop formed by the pitch rope
  • A dynamic fall similar to someone slipping on the lip of a pitch
  • Knot recently tied – i.e. first person down after rigging.

It is worth noting that we did have other failures with different test conditions but on a far smaller percentage of tests. The loose arm being the common factor in all of our testing.

Bowline-on-the-Bight test – YouTube

Best Practice:
We teach what is known as best practice. This ideal of coaching changes and we are constantly evolving our advice. This is one of the reasons I went and did these tests today, to ensure that what I did was still the best it could be.
Current best practice advice for SRT, rigging and progression where relevant to this article is as follows:

  • Rope used should be of good condition and supple enough to hold a knot well
  • The rigging should be tight and all knots dressed correctly and tightened down before use
  • Y-hangs should be loaded equally
  • The caver should always have their cowstails clipped into both arms of Y-hangs (you’ll notice we use larger krabs than most for this reason)

Conclusions:
We experienced no test where a failure occurred when all of the best practice conditions above were met.
Failures did occur in other tests, the chance was lower unless all of the ‘Lemon’ factors were involved.
I suggest that anyone with genuine concerns seek advice and, of course, under no circumstances should you use a technique that is unfamiliar to you. Get trained, get experienced, get informed.