The risk is real, but not frequent, so far...
Canyoneering is an activity with inherited risks. Getting lost, drowned, flash flooded, hypothermic, heatstrocked, and body extremities’ injuries are a thing in this sport. But the accidents that attract the most attention are the ones involving ropework.
Common repeated accidents:
These seem to be the common accidents involving ropework:
Rappelling on the unsecured side of a Single Rope system.
Failure to deal safely with a rope that is short (does not reach the ground).
Failure to deal safely with uneven ends while rappelling a Double Rope system.
Losing control of rappelling speed.
Rappelling on an open system or incorrectly rigged system.
Ascending an unsecured system to recover a stuck rope.
Anchor failure - Webbing abraded and/or damaged.
Anchor failure - Natural anchor dislodged and/or failed.
Drowning due to strong current, hydraulics while on rope.
The number of incidents or accidents seems to be low per year. It is not clear if this is because the sport is low-to-medium risk (but high consequences), or just a numbers game, meaning that as more participants come into the sport, the numbers rise.
Do we learn from accidents?
Canyoneering communities break into a fever pitch every time an accident makes the headlines. The reason given to mitigate the appearance of morbid curiosity, is the desire to learn from the accident, in hopes to avoid future ones.
But the song remains the same
This seems to describe the vicious cycle every time an accident surfaces in the news and social media.
Super vage news of an accident surfaces in social media. Something like "Person falls from a cliff at location X and suffers serious injuries". Location X is a well known canyoneering destination, and the speculation begins in earnest.
A second wave of more detailed news surfaces, or word-of-mouth in the community spreads with more details, like the actual canyon, rappel height, person, gear and rigging involved.
A race to concote a cause of the accident ensues.
A biased analysis surfaces and the wrong conclusion sets in
Community moves on ready to reach the same conclusion on the same type of accident, next month or year.
Biased?
The bias seems to be rooted in the very understandable impulse of not blaming the victim of the accident, and shifting the cause elsewhere, usually the gear, the beta, unexpected conditions, etc. Some examples of how accident analysis in social-media tends to reach biased conclusions.
Cause of the accident was not a failure to deal with a short rope. The cause was not updated beta indicating a new anchor further back.
Cause of the accident was not "failure to control rappel speed". The cause was a 'dangerous' rappel device not appropriate for long rappels. Or…
Cause of the accident was not "failure to control rappel speed". The cause was a not using a self-belay prusik. Or…
Cause of the accident was not "failure to correctly install a self-belay prusik". The cause was a 'nor providing a bottom belay’. Or…
Cause of the accident was not "failure to provide a bottom belay". The cause was bottom belays do not work.
The cause of the accident was not an incorrectly rigged releasable block. Rigging releasable blocks is complicated and dangerous.
Another interesting aspect of reports and discussions, is that sometimes they stress that the accident was not due to the lack of victim’s expertise and/or experience.
And so on…
Hopefully with these examples you can discern a pattern.
Moving forward
Communities and organizations have made efforts to formalize a repository or database of accident reports with details and analysis. Most notably iCAD. If you want to learn from accidents, read reports in iCad. Unfortunately, reading reports without the ability to refute them or provide personal hot-takes is not as engaging or entertaining as doing it in social media.