My Internet Post on Rappel Safeties, ca. 1995

This is a longer post intended to explain some of my earlier posts discussing the use of a Prusik Knot as a rappel safety. I'm putting this out for information only, and not to try to get others to follow my advice. Each climber is responsible for their own decisions regarding the safety of the techniques the use. I have no axe to grind, I'm merely passing on my own decision not to use a Prusik safety, and the reasons I made them. The reason I'm posting such a long discussion is that I've received numerous email messages asking for more information. Here it is.

My background is 25 years of caving, taken quite seriously. I collect rappelling and descending devices (many of the devices shown in Tom Martin's Rappelling are from my collection). I was a reviewer for Al Padgett and Bruce Smith's On Rope, and have written several articles in the NSS Vertical Section's newsletter The Nylon Highway. I don't know everything, but I'm not a beginner. I don't do nearly enough climbing, (I climb 5.n, where n is too small to print) but I have done enough to know the difference between rappelling above and below ground.

I make the hopefully unnecessary apology for the gender pronouns used herein. I didn't invent the language; neither did the people I'm quoting.

My attitude is that safety is not given by any gadget, but is a property of one's attitude and experience. In Advanced Rockcraft (p66), Royal Robbins wrote "Safety in rock climbing lies almost entirely within this 'judgment' area. Little is left to chance. Equipment is a minor factor. With the best equipment in the world the man with poor judgment is in mortal danger." True words when applied to rappelling, I feel. I also believe that reliance on Prusik or mechanical safeties often causes a person to relax their vigilance. I can't prove that, I just believe it.

Even if we assume that the presence of a Prusik safety won't affect our judgment, will it work if needed? The answer appears to be "maybe," but most likely not.

Don Davison, Jr. (then Chairman, NSS Safety and Techniques Committee) discussed this in the August 1976 NSS News. On page 140 Don writes, "There were several difficulties which caused the use of the chest safety Prusik to be generally abandoned; the most significant being the requirement that the knot be released during a period of accelerating stress. The caver was asked to relinquish a 'firm' grip on the rope and relax in a panic situation. In releasing his grip, he was asked to perform a negative action, a type of behavior which only strenuous and repetitive drill can instill in the majority of individuals."

Three-rope RigDon continues: "The tremendous urge of the rappeller to grip the rappel rope (already in his hand) during the period of stress was documented by the use of Dan Meier's three-rope rig (The Tech Troglodyte, Vol. III, No. 2, Winter, 1965, pp. 31-33). [What follows is not "safe," in the sense there is some risk of injury if anything goes wrong. I don't recommend trying it, it can be as dangerous as many other aspects of climbing. Make your own choice ---gds]. To set up a three-rope rig, a rappel point is rigged on a 60-100 foot high free fall cliff so that its end hangs down only about 20 feet. A second rope, which reaches the ground, is rigged from the same anchor. An overhead belay, well to the side of the main anchors is established with a third rope that can reach the ground. In practice, the rappeller rigs his rappel device into the short rope and places his chest safety on the long rope. The belayer carefully lays out enough slack in the third rope to allow the rappeller to fall about half the total height off the drop before being caught at a measured height of about 20 feet above the ground [this is critical - gds]. Before the rappeller begins to descend, the belayer 'locks off' the belay rope at the measured distance…. Thus the rappeller is already caught and the only variable is whether the rappeller will release the Prusik or be caught by the overhead belay. Most experienced cavers were not able to release the Prusik, especially when closed eyes were required after the free fall portion of the drop had been reached [closing the eyes makes the loss of control more of a surprise - gds]. The three-rope rig was developed after the May 21, 1964 accident in Newberry-Banes Cave, Virginia (The Tech Troglodyte, Vol. III, No. 1, Fall, 1964, pp. 18-21). In this accident, a caver 'rode the Prusik down' the rappel rope for over 100 feet before the knot was relieved of human interference when the caver's head struck a ledge." [Note added in 2008: Konstatine Serafimov, a Russian Caver who survived a 1980 grabbed ascender incident in Kutuk-Soumgan Cave, used a similar setup to test rappel safety ideas in Russia. These tests led to his developing the Reflex Ascender. He sent me this drawing of how he rigged the three-rope-rig in 1981-82.]

Davison goes on to describe a "Safety Rappel Cam" which he developed and I won't go into. It is difficult to build, only works on single ropes, and never became popular. In fact, other than Don, I don't know of anyone except for me who actually built one.

In the January 1977 NSS News "Hits and near misses" (p. 18), a caver relates a story where two Prusik safeties failed to grab on a 200-foot free drop on Goldline. "All of a sudden I started falling real fast. I couldn't grab the rope below the bars to brake in time, so I grabbed the rope above me where the Prusiks were. When I instinctively grabbed the rope, the Prusiks slid along with me and I dropped 110 to 120 feet until I hit a ledge with my feet-damaging my legs. The Prusiks grabbed and I swung about 60 feet across the pit where my head impacted the wall, cracking my helmet, and finally stabilized hanging upside down from the Prusiks 30 to 40 feet off the floor…. Injuries included: Left femur broken just above the knee, head of left femur badly cracked, left ankle severely sprained, right heel fractured, deep gouge in right knee, bruised ribs, rope burned palms of both hands." Don's comments in the evaluation included "The failure of the chest Prusik to function was again caused by the human element when the victim 'instinctively grabbed the rope'; functioning was proper when relieved of human interference [but look what it took to relieve the human interference - gds]. Reliance upon and use of the safety Prusik should not be encouraged among human troglodytes."

It doesn't have to be a Prusik. In the June 1977 NSS News (p. 128), another caver used a Gibbs ascender as a chest safety. "I don't know what caused it, but the rope slipped from my right, and braking hand, and swung to the other side of my body, where I couldn't get at it. I gripped 'the rope' instinctively with my left hand which was on the Gibbs, and rode it down for the ninety feet…. Result: Broken right femur, lacerated right knee, assorted bruises, eight weeks in traction, four to six months to regain full use of the leg, possible operations to repair ligament and tendon damage." The evaluation comments "The victim had used the chest safety Gibbs, during practice. ONCE AGAIN, the very real danger inherent in using safety rappel devices which are held "open" manually is well documented. In a time of stress, the 'negative action' of relaxing the grip around the rope, be the hand directly on a Gibbs, chest safety Prusik, etc., is an unlikely probability. No rappel safety device which requires a 'negative action' to activate should be trusted…. Hopefully no more injuries will occur before cavers realize the dangers inherent in the various 'negative action rappel [safety] devices."

In the July 1977 NSS News (p. 148): "The inherent problem of human interference with the function of the 'negative action' chest Prusik safety has been well documented (NSS News, Aug. 1976, pp. 140-1; Off Belay, December 1976, pp. 14-17); but this hazard is also characteristic of any rappel safety which is held 'open' on the rope above the descending device. Some cavers, however, feel that the use of a hand held Gibbs or Jumar will eliminate the danger-not realizing that the problem is not in the device but rather the human.

"During any period of unanticipated accelerating stress, a caver who is not heavily conditioned will tend to grip the rope and, in turn, the hand held Prusik, Gibbs, or Jumar."

The Off Belay article just referred to is Ray Smutek's The Questionable Prusik Safety. It is too long to reproduce here. If you can get a copy, read it. It also provides relevant quotes from the Tech Troglodyte articles cited previously, which would be difficult to obtain at this time. From the Troglodyte quotes: "The instinct to grab onto something when falling is very strong. During the use of this rig [the three-rope], five experienced cavers tried it out. Only two let go of their chest Prusik on the first drop. The other three panicked in various degrees, freezing on the rope. If this can be taken as a valid sample…." It can't. Cavers have repeated these tests (albeit undocumented in most cases), and the results indicate that the success rate is much lower than the 40% indicated by the first five. Mr. Smutek's article goes on to discuss the possibility of the Prusik sling melting or failing - both possible, but not my main point. Mr. Smutek concludes, "The protection provided by a Prusik 'safety' is highly suspect and quite possibly and illusion. Couple this with the nuisance of accidental jamming and the danger of getting hung up and it appears to me that a Prusik 'safety' does nothing except needlessly complicate an already complex maneuver."

One alternative that seems better than the chest Prusik (but still not adequate to me) was presented by Larry Penberthy in Off Belay No. 16, pp. 10-11. Since "The trouble with the chest/Prusik system is that a beginner [anyone - gds] may lose control, start to slide rapidly, panic and then grasp the Prusik even tighter, thus preventing it from working. It is contrary to instinct to let go of the rope to gain security." So, "At an MSR 'working' field trip, we devised a new method of security for rappel The rope from above passes through a [rappel device] and then down to a security knot [Mr. Penberthy recommends either a Penberthy or a Penberthy-Pierson knot, not a Prusik - gds] attached to a webbing loop around one thigh [nowadays, attach to the harness - gds]. When descending normally, the braking (lower) hand grips the knot to prevent it from grabbing, and simultaneously applies enough friction to control the rate of descent.

"As the climber descends, the rope slides upwards through the security knot, and then through the friction device. If the climber lets go with his braking hand completely, the knot grabs and stops him. If he grips the control knot tightly in panic, the extra braking friction force stops him, assuming the friction device has a high enough friction ratio."

But there is a problem, and Mr. Penberthy recognized it: "CAUTION: The security knot must not be able to touch the [rappel device]. If it does, the security knot will not grab." My experience suggests that this disadvantage is enough to be a problem, so I do not use this technique.

Off Belay No. 30, p 37 describes a suffocation death when a chest Prusik locked off. The climber was dead within about 30 minutes. In the evaluation, the Prusik safety is mentioned and "its use is a questionable practice."

Off Belay, June 1977: An article titled "The Prusik Safety Strikes Again" tells (quoting 'Mugelnoos') of a rappeller who lost control of her rappel, then fell ten feet until her shirt tail tangled in her rappel device. At that point the Prusik safety engaged. "The girl was extremely lucky, since a Prusik safety will NOT stop a climber once he or she begins to fall freely, unless something else slows the climber to a near standstill. In this case, it appears that the friction of the rope running across her back, and probably more important, her shirt tail jamming in the brake brought her fall to a stop, permitting the Prusik to grab.

In his book "Single Rope Technique Equipment: a guide for vertical cavers" (Sydney Speleological Society Occasional Paper No. 7, 1977), Neil Montgomery writes "The value of a climbing knot [as a rappel safety] is questionable since in a panic it is likely that one will keep a tight hold on the knot and prevent its gripping." I disagree with the follow-up statement that an ascender is better, for the reasons given earlier and because that is not what they were designed for. Neil describes the "Spelean Shunt" made from a Gibbs and a carabiner. A small minority of vertical cavers use these, most of us don't. Since they work on single ropes only, I won't describe them. I personally don't like them, but there are a few vertical cavers whose expertise I respect who do, so I mentioned them.

Prusik rappel safeties are not mentioned, as far as I could find, in David Judson's (ed.) Caving Practice and Equipment (British Cave Research Association, 1991), but autolock descenders (e.g., Petzl Stop, Diablo, Dressler, SRT, GemLok, Tracson, Lewis, etc.) are. Most of these devices are too heavy for most climbers to want to carry them. In the rappelling chapter (p. 57), Dave Eliot writes "One notable disadvantage of most existing autolock designs arises from the need to release the handle for the device to lock on the rope, whereas a thoughtless or panicking caver might instinctively grip the handle tighter and only worsen the situation." Mr. Elliot says essentially the same thing in his book SRT (Troll Safety Equipment Co., 1986, p18), once again not mentioning the Prusik safety.

The use of a Petzl Shunt as a safety is mentioned in the early edition of Mike Meridith's book Vertical Caving (ca. 1979) in a single sentence on page 23. The second (revised and enlarged) edition (1986) removed this reference. These books describe the continental European approach to vertical caving.

Petzl still mentions the shunt as a rappel safety in their catalog, but includes a warning that releasing the shunt is essential for it to function properly.

Al Padgett and Bruce Smith's discussion of the Prusik Safety in On Rope (National Speleological Society, 1987) explains "It was learned through several bad accidents that if a problem occurs, instead of letting go, the natural reaction is to grab. Grabbing a Prusik allows it to slide down the rope, traveling faster every instant. If a person is actually able to come to his senses long enough to let go of the knot, the sling material may disintegrate, allowing the climber to descend even more rapidly than before. In actual usage, the Prusik safety has proven to be troublesome and dangerous." They then mention the Spelean Shunt, Safety Rappel Cam, and Petzl Shunt as options, but correctly note that "No self-belay device should interfere with rappelling technique. If it does interfere, it is counterproductive in that it exists to help control a problem, but instead tends to create one.

One last quote, from John Long, on page 155 of the second edition of How To Rock Climb: "A sliding knot backup (commonly referred to as a Prusik backup) is rarely if ever used as a normal procedure. If you don't know how to rappel, get a belay. If you are doubtful that you can make a certain rappel, don't make that rappel. Only if you are doubtful and must rappel, and no belay is possible, should you consider the Prusik backup as an option…. All told, the Prusik backup is a highly contested technique. The only thing for certain is that it can be highly problematic."

The Prusik safety may be highly contested among climbers. Among cavers, it is not. It is almost universally rejected.

----> Gary Storrick