Chris Vreeland said:
what scenarios were they thinking of that could cause a a high fall-factor fall on cows tails?
I think they included factor 2 falls on cow's tails simply because they are theoretically possible, not because someone would likely experience them. They did relatively few factor 2 tests.
in nearly any scenario, (if you're doing it right) you're below, or at best level with the anchor, when crossing a re-belay, so falling onto that anchor would never create too much of a fall factor.
That is correct. A fall from the level of the rebelay could theoretically be a factor 1 fall, although more likely one would pendulum more than fall free, so even then the fall factor would likely be less. The only time I can see climbing above the anchor to which one is attached is during an uphill traverse where you have not yet clipped into the next anchor up the chain. But even then a pendulum type of fall seems more likely than free fall.
The worst-case scenario seems like it would be if you were hanging from a cows tail attached to a re-belay and the re-belay anchor failed, and you fell the length of whatever slack there was in the bight of rope above the re-belay.
Although one might fall a longer distance under this scenario, it would not necessarily be a higher factor fall because you would be some distance below the upper anchor which ends up holding the fall. As Chris pointed out, there would be stretch in the rope and tensioning of the knots above, which would help reduce the load on the cow's tail. This has actually happened. I have read at least two reports where a knob bearing a rebelay has broken off while being crossed. In one case the caver was unharmed but was left hanging on the rope with 40 kg of rock hanging from his harness!
I know a lot of people use figure 8 knots at re-belays, but one should consider that a figure 8 deforms when side-loaded, as it would be if an intermittent anchor failed. There are "those who say" (I'm one of "them") that a butterfly knot is better to tie into re-belays, as it does not deform when side-loaded.
According to Alan Warild's book "Vertical" (2007 edition available free at <http://cavediggers.com/vertical/>; look at Chapter 3, Knots, p. 45: Mid-rope knots) the figure 8 is a better knot than the butterfly even under sideways loading. I had also been taught that the butterfly was better under side loading, but apparently it isn't as strong and the figure 8 does just fine. I do still use the butterfly on horizontal traverses where the rope goes directly out to the side of the knot in both directions, mainly because it is easier to get the rope tight to minimize slack. But at rebelays we never use a butterfly. Mark Minton