I started this article evaluating the Backtow™ feature of the TeamO lifejackets, but quickly realized that first I needed to figure out the tether tension a crew member who falls overboard will experience while being dragged, before the remaining crew can slow the boat.
This was doubly worth my time because many of the chapters in this Crew Overboard Prevention and Recovery Online Book will benefit from a better and more granular understanding of these loads, rather than the simple single-number estimate that I used when writing them.
Also, despite googling like a fiend, I was not able to find any studies, or even sensible guesses about what tether loads will be when being dragged. Seems the industry and safety regulators have their heads firmly up…err…in the sand about the whole issue.
Scary when you consider there are thousands of sailors out there clipping to sidedeck jacklines, or hard points close to the edge of the deck, with standard 2m (6 foot) tethers, on the assumption that if they fall in their buddies will haul them out and all will be well.
And, of course, the key question about TeamO jackets, which everyone seems to be ignoring, is will the COB be in good enough shape after hitting the water to activate Backtow?
So let’s dig in, and after that I will finish my thoughts on Backtow.
Brace yourself. I will share a lot of detail so that the technically oriented readers, particularly you engineers, can correct any errors I have made in either my assumptions or calculations. Think of this as a working paper exploring a complex subject.
So if you don’t feel like reading all of this, it’s fine to skim, focusing on the text in coloured boxes, tables, and graphs, and then read the example scenarios where I really get to the meat of it. You also don’t need to worry about the footnotes, unless interested, which make up a lot of the word count.
Hi All,
Many of the comments so far are, despite my suggestion not to, drawing conclusions from the above article. Fine if that’s what you want to do, but I will not be engaging with those comments since this discussion is about my methodology, not conclusions.
We will be discussing conclusions in a later chapter so comments of that type here will be lost and forgotten quickly, and I will almost certainly delete them when I clean up.
So to avoid duplication and wasting your time (and potentially mine) I would suggest keeping your powder dry for the next chapter.
before commenting, happy New Year!
Great work, thanks!
Whatever the numbers are in the end, they are sobering. Seems that going over board needs to be avoided at all cost. So strong-points inboard and short tethers, as you recommended (if I recall properly).
Also, when going overboard, not having a tether might be preferable, provided there is a proper life-jacket, beacon and a crew that can get the casualty out. As horrid as that is, it is still better than being squeezed to death by (only) 120 kg/kp or being dragged behind with the head under water. Both feel like re-enactments of particularly gruesome medieval execution methods.
Hi George,
Thanks for the kind words. Yes, the issue of whether or not it’s safer to not be tethered at all, rather than being dragged, is an interesting one, and dependant on a lot of factors. I will tackle that in the next chapter, once everyone has had a chance to weigh in on these numbers.
Wow, that took an extra piece of toast and three cups of tea to digest, great stuff. A question though please – what about ocean wave strike?
A large breaking wave crashing over the deck from windward bow to leeward quarter with say, a white water speed of 25 knots? Strikes the COB and sends them into the water at a relative speed of just under negative 20 knots, ie. backwards? And worse, still perhaps accelerating (backwards) when they enter the water?
A glance at your tables above John, suggests in this COB event (perhaps not an unexpected one offshore), such revised numbers wouldn’t be survivable, or could they be?
Hi Rob,
Wow, a lot of variables on that one. I’m pretty sure most scenarios would result in higher loads than those above with the governing variable being whether or not the tether comes tight as they are washed off, or slides aft. If it comes tight, and they go under the lifeline, then let’s just say not good . If there is plenty of slack then they will decelerate very quickly once in the water, so maybe not so bad, but of course that depends on the wave too. That was long way of saying, I don’t know.
Hi. What a series of articles you have gotten yourself into here John! Very interesting, and hats off! 🙂
I am not really qualified to challenge any of your calculations, but I think the numbers you get are sobering enough, even if they would prove to be wrong. Still, I might fear it can easily get even higher. To name a couple things to worsen the situation:
Most likely you will fall overboard when the conditions are rough, and there are significant chances of the boat rolling and pitching, and given you’re falling overboard at a bad spot (the bow for example, where it is hard to have a short enough tether), the jerk from the pitching boat can be significant… and repeated every wave… and a hull crashing into your head every time…
I also find it probable that you will be somewhat deep underwater, and positioned in a non-optimal angle, with a substantially higher projected surface area for the exact time when the acceleration is the highest when the tether is tensioned.
I am medically trained (helicopter rescuer and anesthetic+ICU nurse), but not really qualified to estimate the maximum force a body could handle. However, what I can say, is that when doing chest compressions (CPR), it is quite common to cause fractures to the thorax (costae and sternum). The older the patient, the more brittle and easier it will crack. These fractures will make it much harder for the patient to breathe if they wake up, and the more fractures, the more difficult. Add to this: The chest harness that a crew use, is quite bad at distributing the load, so the chances of some fractures are very much present (and you may well break some bones as you hit the deck on the way overboard). The harness will keep squeezing your thorax, and the more fractures, the worse it will be (think exponential increase in worsening). Broken ribs and trauma to the chest always involve a very real chance of a collapsed lung (pneumo/hemo-thorax) that require advanced medical care. Then to add some excitement to this mix, the victim is now somewhat submerged and having lots of water splashing in the face…
That’s a pretty horrifying picture, but even if you don’t break any bones (though I find it quite likely that a couple ribs will break), the tight chest harness will make it hard to breathe. So difficult, that combined with the face being splashed with water, I would say the crew have a matter of minutes to slow the boat down before you suffocate.
This is not even considering the risk of spinal injuries, dislocated shoulder(s), etc.
I think the point here is that the forces are more than high enough to injure yourself enough for it to be an issue, and thinking that you can self-rescue is just deluding yourself. Another point to make here, there are a number of videos online showing people jumping overboard to test it, but they are always prepared! From personal experience in climbing, I can tell that it is a massive difference in taking a fall you’re prepared for, and one that comes unexpected because you lose balance or slip.
The climbing world used to have chest harnesses back before my time, but it was abandoned. For a reason. The potential for injury is just so high. And those harnesses are actually a lot nicer to hang from than a lifejacket. I think a somewhat reasonable way to get a feel for this would be to don your lifejacket, then put on a HEAVY backpack or a friend to sit on your shoulders (think the same weight as the static load John calculated), and try to hang from a halyard. Then try to breathe. I have done this with “just” a 20 kg backpack, and I didn’t want to do that for a long time.
I am not sure about any legalities about linking this here, but I am going to risk it. It shows some easy-to-read calculations about fall forces and how it feels. It is for climbers, but still, giving some fairly realistic numbers to it. I would say the last one is the most applicable to sailors. This is falling on a dynamic (elastic) rope, but it is also falling without water breaking the fall. It also assumes the use of a climbers sitting harness, which is an order of magnitude more comfortable and healthy to take falls with than a lifejacket. Here is the link. Look at the short video clips, and imagine that is yourself attached with your lifejacket instead of a climbing harness. The videos look sobering, but if you look at the numbers, and the fall length etc, you will see it is not too far off from what a potential fall on a boat could look like.
This comment got long, perhaps without adding anything specifics or value to this. So a little summary:
You can very likely injure yourself from the forces John have calculated here with a chest harness/lifejacket. These injuries can be so severe that even if you are brought back onboard alive, you may very well be in need of urgent advanced healthcare to survive. Even if you’re not injured, it will cause significant difficulty breathing. I consider myself quite fit and young (30’s), and I would not take a bet that I would be unharmed from jumping overboard at 6+knots while tethered to my lifejacket, and I would certainly not think I would be anywhere near capable of self rescue at those speeds. I would love to prove myself wrong if I ever fall overboard since I mostly sail solo, but I doubt it.
Bottom line, stay onboard.
Kindly, Arne
Hi Arne,
Thanks for the thoughts on the chance of injury from one who knows. Just what I was hoping for to inform the next chapter.
Hi John,
Really interesting work, and in the main I think your assumptions are in line with reasonableness. A few points did occur to me as I read through, though I don’t think they would change the outcome much in an absolute worst case:
It all reinforces what you’ve been writing for years – the tether should keep you on the boat, rather than keep you attached once you fall off.
John, would you share your spreadsheet for easier checking of your calculations? I am also interested in doing some sensitivity analysis on it.
Also, one thing that made me think on this long and hard is your summary statement:
”So I’m pretty sure the real-world peak shock tether tension will be somewhat lower than the sum of the drag tension and acceleration tension.”.
That might be the case if you assume that the biggest load comes from the boat pulling you after the intital yank on the tether has happened already, and getting a more accurate answer would probably require building and solving a differential equation describing the effects of the interdependent values of increasing speed, increasing drag and their effect on the forces. Not my strong suit as a non-mechanically oriented engineer.
However, the bigger concern I have is with the very fact that water being so dense and fairly viscous, it will restrict the initial acceleration when the COB is more or less fully immersed, and almost fully static right after falling and decelerating (assuming the worst case of a long slack), just when all slack runs out.
To illustrate why this is potentially a bigger concern than what is the peak load when drag goes to max with speed, think about what would happen if instead of water, you were immersed in, say, semi-solid, thick concrete?
The snatch on the harness would cut you into pieces – not because of a sudden acceleration but because your acceleration is restricted!
Oppose this with being snatched when immersed in air only: in this case you would only have to worry about the load force of F = mass x g, required to overcome the inertia caused solely by your body mass.
The human body can easily sustain several g’s, and with a proper harness, a substantial number of Newtons – when allowed to accelerate freely.
The case of a human body being suddenly accelerated from zero to six knots through water by a 20 000 lbs boat (meaning the acceleration is near instant), lies somehwere between the two extremes of air vs. concrete (of course concrete being an unrealistic extreme for making a point here).
What the human body cannot tolerate, is being crushed or ultimately being torn apart (which, btw, is exactly the very same thing that happens when acceleration is too high, if being tied to a harness, even accelerating through air, due to ones own inertia).
The initial load when accelerating begins, as slack runs out, before the inertia of the water surrounding you gives way to an increase in speed, is likely to injure you if stretch is too low – perhaps to a much greater extent than the later peak load when drag is increased with speed. Stretch is key here, as you rightly pointed out in the article, but drag may not be, as much as (initial) inertia.
Why so?
Well, opposite to what you inferred by saying “the person is being accelerated through water, which is dense, so the tether tension will be way higher because of the drag.”, I think that in a worst case scenario (=little stretch) the biggest problem is not drag, but inertia. This is illustrated by thinking about why you will die when jumping from a sufficient height to a pool of water, but be just fine jumping onto a gigantic cushion from the same height: water does not give way to your body infinitely fast, therefore deceleration goes up.
Modelling this is easier if you think of being yanked real suddenly by a tight tether analogously to falling vertically onto a the same tether – in both cases your speed tends to change very suddenly, with only the stretch governing the ultimate acceleration (i.e. the change in speed).
If you fall with 10m/s onto a tether with zero give, it will kill you, as deceleration goes to near infinity – just as being yanked horizontally from zero m/s to 10 m/s very suddenly. It really does not matter per se if you fall or are being yanked through air or water – only the acceleration/deceleration counts, as that alone will determine if the harness will cause slight discomfort, hurt you, crush you or cut you into pieces.
Being yanked in water might, however, add to the injuries in this case because of the very fact that the inertia of the water will restrict the acceleration for a very short moment while the tether pulls, before the force on the tether gets bigger due to drag of said water. So, potentially, a double jeopardy, if you will.
This is not to say that water drag will not increase tension (force) on the tether dramatically, or that it would not make survival and rescue unlikely: it will. Just that the biggest shock load might happen before your body moves an inch, due to inertia.
As an attempt to try to make a very long story short, I drew a picture of what the force curve might look like.
I want to point out that getting the real force curve would require exactly the kind of analysis you suggested (=a tedious, expensive and error prone effort), and this is just guess work trying to illustrate that there might be an even bigger initial shock load in the very first milliseconds, than what will follow when drag goes up.
Sort of like modelling the initial yank when inhibited by water’s inertia as an arrest similar to falling vertically onto a tether, followed by an increasing pull on the tether as speed & drag goes up.
Your thoughts and those of the mechanical engineers are very welcome 🙂
This is engineering-wise way beyond what I have knowledge about, but the body would not act as a solid and static piece when being accelerated/decelerated. There would be compression of tissue, bending, deformation etc, which, given the very short timespan and relatively low speeds (compared to say a car crash), would contribute substantially to decreasing the shock load on the tether and anchor point.
The human body is able to withstand quite large forces, but the chest harness/lifejacket we use, is a very poor way to transfer those forces on our bodies.
I also find it VERY interesting to see the numbers John comes up with, but from a medical point of view (without much data in the form of numbers, other than personal experience from various patients I’ve handled at work, and from taking falls etc in a harness myself), I would say the forces here are big enough to make it a substantial risk of injury.
That being said, I still think all the operational risks far outweigh the “medical” risk, especially for a shorthanded crew offshore. You don’t fall overboard on a calm day (or if you do, you just climb up and feel refreshed or embarrassed, depending on temperature and number of spectators).
Hi Vesa,
Good points, but I think Luc has solved all that for us, with a very easy way to think about acceleration in water. To solve for this we add the mass of the water that the COB’s body must accelerate with it. See my answer and Luc’s comment: https://www.morganscloud.com/2025/01/09/tether-tension-on-a-dragging-crew-overboard/#comment-312918
As to inertia, I’m reasonable sure that’s being taken into account. Given we know acceleration and mass (although we must guess the water being accelerated) force will be easily derived using F=m.a but I don’t think the answer will be any different than using my rather klugy, but still valid, conversion from acceleration to g and then multiplying mass.
The standout here is not having long tethers, as short as is permissible to stop a tethered crew going into the water is the best outcome. From the recent Sydney Hobart and the recovered skipper of one the yachts caught in the storm was he was not wearing any tether and washed straight overboard. Lucky to be recovered and only because of the level of safety each crew has on their person in that race.
The person that went overboard from the sailing yacht Porco Rosso in the 2024 Sydney to Hobart race was Luke Watkins. Although he has been identified as the captain and skipper in various articles in the media, the owner and skipper of the boat according to the Sydney to Hobart website is Paul McCartney. He was interviewed as well. Both stated that indeed Luke was tethered, but swept under and pinned beneath the boat. He was holding his breath but realized he quickly had to make the choice to stay clipped and risk drowning or unclip and risk being apart from the boat, he made the decision to unclip.
Hi Betty,
Thanks for the fill on that. Very useful to know and another risk to take into account.
Thank you very much for this thorough investigation into an important but under-researched subject! I too have been enjoying bouncing ideas off of ChatGPT – the best thing is that it never gets bored with me… Something that comes to mind is that I think the initial deceleration of the MOB’s body by the water will tend to rotate it to align with the direction of movement, and I am pretty sure this will be to a head-first orientation, based on the relationship between the center of mass and the center of lateral resistance, although inflation of the life vest, if quick enough, may change this. If this is the case, I would expect most of the subsequent forces to be imparted via the crotch strap. This actually makes me look very differently at any designs with a single and narrow crotch strap…
Hi Alwin,
Sure, whether or not there is a crotch strap, and further whether or not it breaks is significant. I will explore that further in the next chapter.
Interesting, sobering, and well written article well worth the full read. The Practical Boat video was also worth the full view. Thanks John.
I think this one of the times we probably have an inherent advantage with the beam and the landing area of our trampolines on our Dragonfly trimaran such that we can typically rig our jack lines placement set up to the center hull and the tether length to where if you fall and the tether goes tight, you have a +90% chance of still being on board either on the center hull or on a trampoline with only a bruised ego. We do the same with the tether anchor points in the cockpit. That’s pretty important as it’s quite easy and comfortable for our boat to “cruise” at 8-12 knots and I shudder to think about the loads if you did hit the water and the tether goes tight. I worry as much about our dog. He has his own tether and I have to sure he cannot go overboard at speed as I think the load from the immediate drag would either kill or severely injure him (he’s a miniature poodle, 12 pounds).
I believe seatbelt research might inform the questions you are trying to answer. If you can compare the deceleration forces in front-end collisions with COB, while knowing the types of injuries the former causes you might be able to predict the outcomes of the latter. Most importantly, injuries might end up being trauma which occurs inside the chest like aortic root tears not to mention the whip-lash injuries to the unprotected cervical spine. Bruce
Another very important and well-argued topic, John.
I think the most important thing we can do, besides NOT falling overboard, is to train the crew for immediate action to crash stop the boat, taking the load off the person in the water.
Hi,
Thanks for the interesting article.
One comment on the engineering side:
The mass accelerated when the body hits water is not only the body itself, but also some mass of water. This is a well known phenomenon in ship motion analysis and is called the “added mass”. Physically, it is the mass of neighboring water that is accelerated when a object is accelerated in a fluid. This is happening when the jackline is tensioning, accelerating the COB and its neighboring water to boat speed. This added mass depend on many factors, such as the shape of the object, the frequency of the motion, and the type of motion (heave, surge, pitch, etc..). To have an order of magnitude, the added mass can vary from 10% for hull like shape in forward acceleration to more than 100% of the body mass for hulls been accelerated sideways (sway). It means that in the worst case, the mass that need to be accelerated is twice the body mass. Consequently, the dynamic force in the line is also doubled.
In conclusion, I am in the process to have all lifelines on the deck to avoid as much as possible to fall in the water! The work is more complex than I thought. But it worth it as it is for our safety!
Have a nice week-end
Yours
Luc
Hi Luc,
You just caused a total head smack moment here as AAC headquarters! I just spent an hour reading up on “added mass” and now get it. Of course what I don’t know is what factor I should use to calculate the amount of water being entrained. That said, I have done a bit of reading and it seems that 50% of the person’s mass would be a decent approximation, resulting in all shock tensions being half as much again higher. What do you think?
The other thing is that I’m pretty sure is that this total overcomes my worry that summing drag tension and acceleration tension yields too high a number.
In other words, if we sum these two difficult to quantify factors, they are likely to cancel each other out, or close enough for our purposes?
Thanks again.
Hi John,
Yes, 50% is a good guess.
For the summation of forces, it is more complicated, as the forces do not take place at the same time. How they add up is highly dependent on the stiffness linking the COB to the boat. I have sent you by mail some numerical simulation on how the stiffness impact the total force profile in the line.
Hereafter, a plot of the total force versus time in a line between the COB and the boat with a relatively low stiffness link (1000 N/m). Boat speed 6 kts.
Hope it helps.
Yours
Luc
And this one with a stiffer link (5’000 N/m)
Hi Luc,
That’s great, particularly that you have answered my worry about adding static and shock loads together as being too simplistic and also added in the added mass, which I completely neglected to do. Of course, as you say, the wild card here is still stiffness (stretch) in the system, but my simplified model allows me to play with that number and so get some ranges. Also it’s good that the two factors you have examined tend to at least some extent cancel each other out. So that your high stiffness number is higher than mine and your lower one lower.
So I think we are now “good enough” to draw some conclusions, although these will always be qualified by being ranges of possible outcomes and I will also make clear, once again, how much difference the stiffness in the system makes.
Reading with interest and fear.
I suggest that we are most likely to be thrown/fall off when things are worst (sea state, sudden acceleration, boat falling off a wave etc) rather than when on a benign reach at 6 knots.
That leaves us with terrifying numbers and high probability of severe injury.
It’s great that you are tackling the assumptions that people are wearing a harness and fall off in benign conditions.
Hi Dave,
You may be right, but on the other hand I think that the actual effect on tether tension of wave state will be less than might first be imagined because the actual water in a wave is not traveling that fast and in a circular motion, and at different rate in the trough and crest. Of course wave state will have a lot of effect on survivability, but that’s the next article.
Great work. I’ve been through this a number of times, looking at different cases, and I think you are about as close as you can be. The methodology makes sense and I have run through very similar numbers. I too, have been towed at up to 8 knots in testing (I used a full body harness at higher speeds–I don’t like injuries). I did NOT do the sort of impact testing you describe. It makes me wish I had, with a fully dressed 180-pound dummy. We know that most MOBs don’t result in broken ribs, so the forces can’t be too high (studies I have read suggests that broken spine and ribs starts at 500-700 pounds with a chest harness). On the other hand, tethers have been broken when deck sweeping waves were involved, there was too much slack, and the tether (not modern) was clipped to a hard point (no jackline stretch), suggesting forces well over 2000 pounds are possible. But that is a completely different case. So there is a range and you have captured the case you studied well.
I’m looking forward to chapter 2. Please consider the effect of harness impact on the body. The armed forces, OSHA, and UIAA have studied this. Also, as I read of MOB cases, the ones with the lowest recovery rates seem to involve injury before going overboard; for example, they were struck by the boom, they hit their head on the combing, or some other trauma. They range from unconscious to semi-conscious, and although the crew often observes that they were swimming, they were probably less effective and less well off than they appeared. This also relates to the MOB drowning while towed. Were they injured before they hit the water? Were they stunned by the harness impact or by hitting the hull? Did the initial harness impact so squeeze their lungs that they could not begin breathing normally or control their breathing, resulting in swallowing water? I can see getting the wind knocked out of you, to some extent, would be likely. I suggest this because the forces alone do not explain serious injury and I never had any difficulty breathing being towed, because I was not stunned and enter the water smoothly. I could never figure out how to study these combines effects safely. I think most drownings of swimmers are from combined causes.
Hi Drew,
Well that’s a relief! I knew that you had probably spent more time on this than anyone, and that adds to your climbing experience, so I figured if anyone was going to blow me out of the water (pun intended) it would be you. So great to hear I’m close enough, which is, I think, all anyone can aspire to here. Still I will wait to see what some of the other engineers around here come up with before I get too confident.
Also great to have your views on incapacitation. Assuming maiming starts at 500-700 lbs seems to correlate with the little real world data we have. But as you say, ability to breath and function will compromised well below that, so at the moment I think I will go with 300 lbs as the static tether load that will make it very difficult to do anything much for ourselves, particularly since we will also be getting firehosed and banged around. Does that make sense to you?
Thank you John for this highly interesting and relevant article. It is legitimate to try to understand the forces at play during a COB and the potential effects of such a dramatic trauma on the human body. And attempts to put numbers on these forces are as relevant as comparisons with other comparable situations in the field of climbing.
But obviously these approaches have their limits and can only help us *imagine* the consequences of a COB. There are probably other ways, less scientific and more ampirical, of knowing these consequences. For example:
– Repeat the PBO experiment, but attaching a scale to the tether, to see the exact load in all scenarios: short or long tether, leeward and windward side, below or above life lines, etc.
– Read accident reports involving COBs and see what were the injuries or casualties.
I would end my comment with a tricky question: Which way of dying would you chose :
a) falling overboard tethered and being violently towed and crushed and broken to pieces and finally drown
b) same as “a”, but instead of drowning, being brought back on board and die after hours or days of agonising for being too far from medical assistance
c) falling overboard untethered and not being rescued and finally drown ?
Personnaly, I would do everything possible not to fall overboard in the first place. This can only be done by using a short tether attached to the center of the boat. And I can’t wait reading John’s next articles on the subject !
Hi Pierre,
Sure, the ultimate answer is on the water testing with strain gauges that measure peak load. That will be tricky though because most of the meaningful scenarios will require having the gauge as part of the system and so flailing around. Or to put it another way, it would tempting to just attach the strain gauge to a hard point, but that would only test the bad scenario above, which is also the least likely…actually, it would not even test that one properly since it would miss the slide back part.
So yes, an on the water study would be great, but it’s a lot harder to make it meaningful than might first be supposed. Also, we would not even know that without the study I did above, so that shows the value of first coming up with a theory based hypothesis and then testing it, not the other way around.
Anyway, if anyone wants to take on water testing I’m happy to advise.
As to accident reports, I have been an avid reader for years. Short version (from memory): being dragged usually, although by no means always, results in a fatality.
Great analysis John. Wouldn’t it be fairly straightforward (and probably less time consuming) to test the assumptions and calculations with an empirical test or two?
For example, one might perhaps add a tension measurement device to the tether, have a crash test dummy or similar don the vest, and toss said dummy off a moving boat? I think one could probably debate the various assumed inputs, and calculations ad nauseam, but until we have some empirical evidence to back up the results, we won’t know whether or not the theoretical results are in the correct ballpark relative to reality? Gut feel as a sanity check is a good thing, but as you regularly point out, it’s sometimes wildly wrong.
Just a thought.
Hi Quinton,
See my answer to Pierre: https://www.morganscloud.com/2025/01/09/tether-tension-on-a-dragging-crew-overboard/#comment-312927
And note that nowhere have I suggested that gut feel is useful. In fact the whole reason I put so much time and effort into this is because I think there is way too much gut feel being applied to this issue. Rather than gut feel I have come up with a calculation methodology and am now subjecting it to review, particularly from the engineers. So far we have one meaningful correction from two different sources.
Thanks for your response John. I think I’m less convinced than you are that an empirical test would be difficult to make useful. In my mind (TM 🙂 ) I would tether the dummy precisely as a crew member would be (e.g. at an appropriate location to a jackline), and toss the dummy overboard precisely as a COB might (e.g. under/over a lifeline, etc). I gather that tension gauges designed for these sorts of things are more than up to a bit of rough handling. I would undertake to do the experiments for you (because I think they’re well worth doing) if I had access to a motor boat (for good speed control etc) , test dummy and tension gauge. But ‘unfortunately’ I’m anchored off a remote Polynesian island right now 🙂
Hi Quinton,
Keep in mind that you need to measure shock load plotted over very short period. This was the best load cell I found, but its not clear it’s waterproof: https://coastropes.ca/products/enforcer
It will need to be on the dummy end of the tether otherwise it will interfere with the run out as the jackline comes under load and deflects.
Also a motor boat will not do since we need lifelines to test many of the scenarios as well as jacklines. And then each scenario needs testing at least 4 different speeds.
So that’s four speeds and say 6 scenarios, so that’s 24 tests.
There is also a risk of damaging the boat. For example the over the lifeline test with a realistic dummy has a very real risk of bending stations and/or the pulpit or pushpit. Keep in mind that the impact load on both will be in the order of three times the load on the tether because of geometry of a tight wire being deflected. Think over a ton.
And then we need a realistic dummy. They do exist, but not cheap. Think US$500 to US3000 for a really realistic one: https://emrn.ca/simulaids-reg-i-a-f-f-rescue-randy-with-additional-reinforcement-145-lb
Enjoy Polnesia
You make very good points John. Regarding breaking things, I thought about that too, and figured that starting off with a few slow speeds would allow one to plot the left (low speed/force) part of the curves first. Ideally one would stop the test once the curve trajectory had been determined, and before progressing to higher speeds and breaking anything. The aim would be mostly to validate the theoretical numbers as being in the correct ballpark, rather than gather comprehensive empirical data.
Great article! And discussion!
Hi Quinton,
Given the complexities of the stretch in the system, and that the dummy behaves differently in the water as speed increases, I don’t think extrapolating from lower speeds would be that useful. It’s at higher speeds that the dangers exist, and so that’s where we would need to test to be meaningful, in my view. Anyway, thanks to the engineers, I think we now have good enough numbers for our purposes. After all even the results of real world test at speed would need to be qualifies with large ranges since there are so many variables that could be different from the test in an accident.
There’s a bit of chaos to entertain based on how the person is aligned when they hit the water. If they are directly against the hull, they are interrupting a laminar flow as well. I’m curious to know at what point in the acceleration you assess the life jacket would begin to inflate. My experience suggests that a hydrostatic trigger might be unlikely to activate in this scenario.
Hi Jonathan,
I think laminar flow will not be consequential since it’s a very thin layer (think a few mm) so will not act on the COB to any measurable degree.
On the other hand, whether and when the life jacket inflates is significant to drag, but not acceleration (discounting added water (mass)). I did think about this quite a bit and probably the most significant effect would be to reduce peak load if it did not inflate until after the acceleration is over. That said, if the COB went in quite far forward the jacket will have time to inflate as the clip slides aft on the jackline before the acceleration starts. Bottom line, it might make a difference, but it might not, therefore we should assume worst case which is that it fully inflates before acceleration starts.
A long way of saying I thought about it at length and then discounted it.
Only talking of the static side of things and in a completely qualitative way, we often swim off the back of the boat.
When the wind is boring-low, swimming from the back with a floating line is a great way to cool off.
I start to get nervous when the boat reaches 3 knots and I do my best to never have someone back there over 3.5. Pulling the line with an adult wearing only a swimsuit at 4 knots becomes not-fun very quickly.
Foulies and an inflated PFD must increase drag significantly. I think your first static calculation is closer to the real thing.
I think Drew was a bit nuts to test over 8 knots 😉
Hi Mattieu,
Great data point, thank you!
Hi John! Thanks for the email linking us to this discussion. WOW! I had no idea. I simply accepted the idea of Jack Lines and tethers as an accepted maritime concept. ‘From the days of yore’……I had no idea about the numbers on loads you posted. Thanks …. Like I wrote before, when delivering boats I would always tell crew “DON’T FALL OFF” …. seriously…. I have always had doubts that we would be able to get the boat turned around in time to still see the person who had fallen off in the swell and recover them. Not to even mention darkness or diminished visibility from weather.
The alternative of being tethered to the boat, given the reality of the loads on the person in the water/tether tension, are equally daunting.
MY wife, admiral, boss…. she and I agreed awhile ago the only viable method of protecting ourselves was to embrace/trust…. ELECTRONIX! I know that these gadgets also have their naysayers/potential issues. BUT what are the choices? I know this chapter/article is a work in progress. I’m sure you are onto it. But just in case: we are using strobes and MOB transmitters from ACR (linked to the MFD chart plotter). We have practiced their application using the dinghy. Believe me when I write this: practice IS important.
If anyone has a better idea please let me know.
Hi Devon,
Absolutely on the MOM1: https://www.morganscloud.com/2018/03/30/person-overboard-recovery-making-the-most-of-ais-beacons/
That said, if you want to reduce drag risk to near zero when staying clipped on we have a complete sep by step guide: https://www.morganscloud.com/category/safety/book-crew-overboard/
I think some useful information comes from the studies of crowd crush incidents. They use vest with pressure monitors integrated. The tolerable force (when you get anxious) against a 100mm flat bar and your ribs is 140lb. Less for women or the older man. I guess you can apply that to the straps of a harness. There is a phenomena called crush asphyxia: 250 lbs for 4 to 6 minutes. Drops quickly as the force increases. Ribs fracture at 20% chest compression which is 675lb. Flail chest (which is multiple rib fractures) at 950lb. These are the things that will kill you assuming no injuries before you entered the water.
The other concern in the initial phase would be on straps around the upper airway and jaw with the potential to strangle. Doesn’t require much force for this to happen.
By the way, the Sydney to Hobart skipper who went overboard was swept off by a breaking wave on the deck. He was reported as being pinned under water and released his own tether so he did not drown. Took 2 kilometres to get the boat turned around. The two people who died both sustained head injuries, one from the boom and one from landing on a winch
Hi Danny,
That’s great information, thank you. Do you have a source online you can link me to? It would be great to link to that from the next chapter.
https://www.gkstill.com/CV/Modelling/Pressure.html
Fascinating stuff. It appears to be time for some experiments and real-world data. Load cell and a life-like mannequin?
Hi Brian,
Good idea, but see this thread, it’s not simple: https://www.morganscloud.com/2025/01/09/tether-tension-on-a-dragging-crew-overboard/#comment-312927
Hi John
I read your article on COB. Well, I read all
of it but only understood some of it. Engineering is not my specialty. But as I read I wondered … are we becoming too analytical
about the sailing journeys that we love so much? My comment doesn’t directly relate to this article and as a liveaboard ocean sailor I genuinely appreciate your content. Most of it anyway. But we go sailing because we love the experience in every respect. It’s an inherently high risk activity (ref Sydney Hobart) and we do all we can to mitigate these risks and therefore ensure our safety but if we go too far down the analytical rabbit hole are we losing touch with the ‘sailing moment’? I know that we all see these things from a different perspective but when I read how you applied every waking moment (I’m paraphrasing you) I thought to myself, it seems so far removed from being in the sailing ‘moment’. Do we risk losing touch with that moment? It’s a rhetorical question and intended for reflection and consideration.
These thoughts aside, thank you for your tireless contribution to our collective knowledge base. I do genuinely appreciate your flow of information.
Best
Tim
Hi Tim,
Great observation, and in many ways I totally agree. After all I spend over three decades offshore voyaging (and racing before that) and it was the joy of sailing that drove that, not the joy of analysis. And even today I spend a lot of time just sailing our J/109 or rowing our whitehall, just for the joy of it.
But I guess that begs the question: why can’t we have both, the joy of sailing and the satisfaction of learning new things through analysis? I for one love learning new things. It’s what gets me up in the morning.
The other thing to think about is that if we at AAC don’t delve into these issues, who will? Practical Sailor used to, but has recently gone off in a different direction, and the mags, other than maybe Practical Boat Owner, have long become shills for the advertisers.
The point being that without the cold hard light of analysis, vendors and boat builders can say anything they want with impunity and we just need to look around at the level of general BS to see where that’s heading.
Heck we live in a world where a guy who has sailed down to the Bahamas and back is now a guru of all things offshore sailing simply because he is good on YouTube. Some of what he says is fine, but a lot is just plain wrong…because he does not do any analysis, I suspect.
In this case, I went down this road because a lifejacket vendor is claiming that their product makes being dragged not a problem. Is that true? Don’t know, but the above is the first step in finding out, and several members asked me to take a look.
And calling BS without good analysis would just make us as bad as all the rest.
Anyway, I share your concern on balance, and to show that, here’s an article that demonstrates that: https://www.morganscloud.com/2018/10/26/11-things-we-do-to-stay-rational-about-safety/
Thanks John – as usual your perspective is on point, and fair and balanced. This is why I enjoy your content. You ask questions and you challenge. Much appreciated. I, for one, though a lifelong sailor in my own right, have learned much from your material and the generous way you share.
Great analysis John – and very difficult to do. Maybe I can chime in with my waterskiing and barefoot water skiing experience (I was Dutch champion for a few years). Below is what I experienced …
Water stops you virtually immediately. I know that from falling at speeds from 25 kph to 60 kph. In barefoot water skiing specifically if you’d ‘trip’ at 60 kph you’d fall forwards and your head would dig into the water. It would stop right away and your body would fly over until your feet would dig in, after which the process would repeat (cartwheeling head feet head feet) until you’d come to a stop (not causing injuries other than a stiff neck). So your assumption of some residual speed after falling in seems to me to be enthusiastic – I would probably assume a dead stop.
In barefooting there was a thing called a ‘tumble turn start’. The skier is at rest in the water facing the boat and is then dragged through the water at increasing speed. Pretty soon the skier would start to plane on his belly (head first) after which the tumble turn was made so the feet would face the boat (turn on your back and pull the handle to your thighs, so now below the center of mass. Physics turns you around) after which your feet could be planted in the water and you could stand up. Thing is, the skier starts planing because at the start the point of force is at his shoulders so his body will inevitably very soon go horizontal, dramatically reducing the forces. His shoulders would accelerate in a straight line but his body would make a gentle arc. SO were I to fall overboard being tethered at the (say) shoulders forces would likely be as you calculated or less. HOWEVER we are tethered at our bowels, pretty much our center of mass. Assuming falling in straight (feet or head first) the first yank by the boat will be horrendous as your entire body (with sailing clothing!) is under water and will try to prevent you from being accelerated – not just your shoulders. It will take a while for the body to assume a horizontal position like in the PBO video. A Note: PBO doesn’t show the fall, only the dragging.
So I guess the static forces may not be as bad as calculated but the yanking force, given the attachment location on the body, makes me think you’d need to up your calculations a lot.
As for the dragging: that is a real problem. I might try a ‘tumble turn’ but I am pretty convinced the boat will be in the way and the tether is in the wrong position to manhandle it properly.
If in waterskiing I failed a start I could (and definitely would) let go of the handle. However, hanging on a tether attached to my center of gravity I have no doubt at all one will drown in seconds – no deep breath before falling in, total surprise, panic, disorientation, boat banging into you, being yanked all the time, water in your airways … very little chance of survival.
Keep up the good work John, keeps us all thinking and preparing.
Hi Rob,
Great information, thank you.
One clarification, I agree with you about the COB slowing to zero in most cases. I only assumed two knots in the second case because the tether might not go slack in that scenario. But overall I agree that when making critical decisions we should assume that acceleration starts from zero. I will elaborate on that in the next part.
I also spent a lot of time thinking about the position of the tow point and I agree that if we are indeed towed by a point close to our centre of gravity the static tension will be way higher. From the article:
However, I think that in practice the tow point will be way higher on the body than that in many, perhaps most, cases because the crotch strap will not be that tight and if the loads really come on hard, will break.
More on that in part 2.
Hi John,
The assumptions you are using in the static case seem reasonable but us mechanical guys are definitely not guaranteed to be fluids experts. For example, I had 1 fluid dynamics class which was very interesting but left me realizing how little we were scratching the surface, fluids is a subject that is not as laws of physics driven as mechanical or electrical are, there are tons of tricky constants. The other thing is that the forces seem reasonable and having Drew confirm that is great.
On the dynamic side, it is super hard to figure out for 2 reasons. The first is that people are not perfectly rigid, in fact, they are not even close to it nor is how the harness attaches to the body. The other is that it is easiest to calculate worst case where the person slides in at the exact angle they will tow at, the tether is straight, etc. and in reality, this happens extremely rarely. I like to think of these problems from a probability standpoint. Let’s pretend that people are in fact perfectly rigid, then it would be nearly certain that a person falling in just the right way would be killed by the tether load even at moderate speeds. On the other hand, the vast majority of the time, they would go in and the peak loads would be a lot lower as the time over which the energy input happened would be greater due to the person spinning around, the tether straightening, etc. and it may well be survivable at lower speeds. If we took the absolute worst case and highest speed, it could be discouraging but if we take the max load at a 90% case and a normal quick speed, we may well find a solution that works and while not perfect does really move the needle on safety. By the way, in this type of calculation my experience is that it is usually easy to overestimate the loads due to it being hard to account for all of the places that energy is going.
To put a stake in the sand of a way too worse case calculation for what tether tension looks like for simply accelerating a person to 6 knots, I threw together a quick dynamic model. Hopefully the numbers are right, I have given it a quick check but not a rigorous one. Note that this assumes a perfectly rigid person, +50% for added mass, the perfectly bad geometry, etc. so is not at all realistic. As can be seen, there is quite a large velocity overshoot. Your max tether tension will be when the COB velocity equals boat velocity and since this is much higher than the drag, you will have a significant speed overshoot. In real life, with all of the other things deflecting the velocity and tether tension overshoot will be a lot less but still present. If I kept this model going, you would see that it would just oscillate as I haven’t added appropriate damping. I hope no one takes these numbers at face value, it simply gives an idea of what a much worse than realistic case would be and the general shape of the curves.
I saw your justification for using 5% tether stretch dynamically but this seems high to me on a polyester tether. Yes, the stitching does deflect some but years ago when I looked at this, I think I remember finding that it was made up for by the length over which the webbing is doubled. I would expect 5% stretch to correspond to a load of more like 1500lb/680kg based on a typical curve. Of course, if you want to start including the harness, by geometry that gives much more stretch so it just depends what bucket you put that in. I wonder if anyone has data on how stretch is impacted by age (UV, wear, cycling, etc.), I get the impression that lines and webbing get stiffer with age but haven’t seen data on it.
In addition to the Team O review, I am looking forward to seeing your jackline setup on the J109, I know our setup could use some improvement.
Eric
I realized it would be interesting to look at the unrealistically bad dynamic case with a DCR tether but all of the other assumptions identical including lacking damping again. As can be seen, there is still an overshoot but everything takes a lot longer leading to lower tether forces.
One thought for people wanting to play around with more realistic forces, you can trick a climbing fall calculator into giving you a reasonable idea of what is going on for this case of falling in the water by setting the distance from the last anchor and length of rope correctly. To reach a speed of 6 knots in freefall, you have to fall 0.49m.
Eric
I realized that I lost the axis title for the right axis. It is COB speed through the water measured in m/s (6 knots = 3.1 m/s).
Eric
Hi Eric,
Thanks very much for looking at that. I agree that the danger here is that we fixate on worst cases and in fact in many cases it won’t be that bad. In fact we even have a data point to prove that: a member who went over on a tether and was brought back aboard in relatively good shape and without injury from tether loads. Unfortunately he is not sure how fast the boat was going (5-7 knots) and of course that makes a huge difference.
On the other hand, I think I also have enough confidence to highlight a quite narrow boat speed and low stretch range where permant injury is a real risk that we should be aware of.
More on all that in the conclusions.