Water Knot: Hidden Risks

"The water knot can be unsafe with long tails.  

The water knot can be unsafe pretensioned.  

The water knot is unsafe in most applications."

The water knot remains popular in the US for specialized climbing purposes. Several federation members of UIAA promote alternatives as safer however - especially DAV, the world’s largest and most influential climbing club. Let's explore if this bend is as safe as believed.

We can begin with issue #58 of Bergundsteigen, a safety magazine edited by DAV and SAC. This article titles it a death knot, describing an accident on Hohes Wand, in which a woman became paralyzed after the knot pulled loose. The writer cites Pit Schubert, past president of UIAA Safety Commission, who warned about this knot in his books and describes 15 accidents involving release of a water knot.

A Norwegian climbing magazine Klatring has a similar article, stating: “German Alpine Club (DAV) has documented that failing water knots can be linked to at least a dozen fatalities. Therefore, it is now strongly discouraged.” This is seemingly confirmed in a DAV chapter manual, which states “Warning... no longer recommended!” for the water knot. An April 2007 DAV newsletter also cautions against it.

Locally, in a recent fatal accident review, OSHA fined Exum Guides $7500 for negligence in using the water knot and omitting inspections. It appears the water knot caused the fatality.

Known Issues

1. Cyclic failure  This is the well-known issue highlighted by Tom Moyer in 1999. Most water knot failures in North America are ascribed to this mechanism. Moyer concluded a 3”+ tail forestalls the failure. 

2. Snag failure  This mode was covered by Schubert in his European books, but is less well known (example):

In a lab test (video), the snag seems to recur even with long tails and the knot pretensioned to 50kg. Long tails and proper dressing do not mitigate this failure. Snags include rock extrusions, trees and bolts.

Benefits of Water Knot

1. Inspectable - the bend is easy to inspect at one glance.

2. USA standard - this is the most typical way to tie 9/16” webbing.

3. Releasable - The bend is easy to untie after loading with body weight.

Known Accidents

AAC reports seem to speculatively ascribe water knot failures to the cyclic issue. Both mechanisms leave a half-tied runner however, so the snag may occur more often than realized. 

For example, in this fatality the remaining tail was over 36 inches long - if tied symmetrically, this suggests the original runner was tied with very long tails. The report notes the remaining half was snagged behind a rock, and the anchor was prepared just one day earlier. Perhaps a snag failure could better explain this accident.

Examples from AAC:

1. “When I leaned back to lower off, the water knot on the sling popped open.” (link)

2. ““..witness statements suggest that he was using a 98 inch section of 9/16 [inch] blue tubular webbing tied in a loop with a water knot to tether himself to the anchor and that the knot came untied…. Among the clues investigators used in their probe were burn-like marks on the tether webbing, possibly indicating stress from the friction of a knot pulling loose.” (Falk death)

3. “In this case, the knot (presumably a water knot) securing the webbing came undone.” (link)

4. “The knot that parted was a water-knot, or overhand follow-through. The knotted sling that came undone had held body weight before.” (link)

Alternatives

Double Fishermen’s Bend

“Do not use the flat overhand for tying slings or webbing. Use a water knot for slings that may be retied and a double fisherman’s for permanent knots.“ - American Alpine Club

One safer alternative for webbing is the double fisherman’s (DF) bend. Some indicate it looks cryptic in webbing though, and is difficult for a climbing partner to inspect:

Flat Overhand Bend = Not Safe?

Another alternative in use is the flat overhand bend (EDK). However, Tom Moyer cautioned against using the EDK in webbing, since it seemed more prone to capsize. More recently, canyoneering guru Rich Carlson also recommends against it.

Accessory Cord

Another solution is forgo webbing altogether for accessory cord (en564), tied with DF. To most climbers today, cord and sewn slings (en566) are more familiar than loose webbing (tape, en565).

*Historically, tubular and flat webbing were inexpensive and used for homebrew harnesses and draws. It was ideal in the 1970s when a hardware store was the only local gear source for many. Tape was also rated to US Milspec prior to EN standards. Today, tubular webbing is mostly used for disposable rappel anchors, and flat webbing is not used in climbing. Certified tape is identified by tracer lines:

Certified cord is more popular today, and nearly as affordable as tubular webbing of equivalent strength. Cord is easy to inspect in a double fisherman’s bend. Best of all, cord’s sheath protects the functional core, so it survives abrasion better than webbing. Additionally, core yarns are shielded from UV, extending outdoor lifetime.

That said, tubular webbing is softer and wider, offering more friction. This gives more security slinging a marginal anchor, like a minor rock knob or chickenhead. Those features remain risky for rappels though, so this advantage of webbing applies more in an emergency escape than regular use.

Redundancy

Using a redundant sling likely prevents water knot accidents.

Euro Background

For water knot concerns of UIAA members, also consider the overall culture of climbing. North American climbers have perhaps more tolerance for adventure than Euro clubs. As Sam Newton reports, this is somewhat by necessity, since accidents often prompt a new law. As he notes for CAI:

“They express a great deal of anxiety... around risk management and the regulatory environment… The impact of the public and regulatory response to such fatalities is very real. In several cases, some areas and activities have become forbidden or closed to the public”

In some locales, climbing accidents are even considered a crime. As Newton reports, 

“...in the case of any mountaineering fatality in Italy, the public prosecutor is legally obligated to ‘open a file’ on the incident…. It has been the case that a surviving member of a climbing partnership, on a completely private and independent trip, has been held criminally accountable following a fatal accident.”

For many European routes also, descent anchors are well-established with bolts and other fixed gear. Rappelling from krummholz or threaded rock features is perhaps less common, so there is less need for a water knot.

Finally note the size of DAV, which may result in more draconian safety rules. The largest chapters (80,000 active members each) are larger and lead more events than all mountaineering clubs of the USA combined. DAV hosts about 300 trips per day on average, and deals with 1-2 fatalities per month in classes and trips. This enormous scale (1.1 million members) may focus more attention on unlikely accidents.

Experts on Water knots

“Experience has now shown that the commonly used Water Knot is too insecure (it unties itself with remarkable ease) for use with tubular webbing.” -Yvon Chouinard

“But I believe in embracing the risks inherent in the activity, the ones that are, essentially, determined by nature. When climbers have a choice of what they will do, then I think that choosing a dangerous option rather than a safe option is wrongheaded, particularly when the choice imposes risks on others.

Seriously, isn't this a no-brainer? One knot (water knot) has been proven, in tests and in the field, to come undone under cyclic loading. Another knot (DBF), just as easy to tie, is stronger (which doesn't really matter) and completely stable (which is the point here). Why on earth would you choose the unstable option?” -Rich Goldstone

“I use a water-knot for slings I plan to untie later, and a single or double-fisherman's for slings I don't plan to untie - like slings that I leave at an anchor.” - Tom Moyer

“I still use this 'bend' for forming tape slings however I make sure I have nice long tails, the knot is set and tight, and that the knot can not rub or catch on anything.”  -Richard Delaney

Literature

• National Cave Rescue Commission (NCRC) requirements for its classes is a hands width or 4" of tail 

• Freedom 9 recommends 3” of tail minimum.

Analysis

The water knot is less secure for unattended anchors, where it may snag. This danger is not mitigated by longer tails or pre-tensioning the knot. The need to guard against known cyclic failure also makes the knot less safe.

The water knot is also used successfully and is well accepted. In modern North American climbing, it is receding into a niche role for expendable anchors. Here it is used with tubular webbing plus a rappel ring, and seems acceptably safe in that role.

For other applications, it seems better to opt for cord as a safer alternative. Cord is also preferable for threaded rock anchors and trees. Perhaps at minimum, the water knot should only be used with 3”+ tails, with the knot positioned carefully to not rub or catch on anything.

Suggested Conclusions:

✅ Use 3”+ tails, tighten, and ensure that the water knot won’t rub or catch anything.
✅ Opt for accessory cord over loose tape.
✅ Only use tape for expendable rappel anchors, and only when they demand high friction.