The Offshore Voyaging Reference Site

“Cheeki Rafiki” Loss Update

Many of you will remember that I wrote an in-depth opinion piece on the tragic loss of four lives due to the sinking of Cheeki Rafiki and what I see as the total failure of the enquiry report to highlight the real lesson: existing yacht construction rules, particularly as they relate to the keel to hull joint, are nowhere near stringent enough.

Sadly, it seems that nothing is being done about that, though there are other developments, as the director of the sail training firm that managed the boat is being held criminally responsible.

I suppose this is a good development in that it may encourage others to be more careful.

On the other hand, I still feel that the industry and regulators have missed the most important point: if the boat had been properly built no one would have died regardless of what short cuts the management company may, or may not, have taken.

And since the boat was built to applicable standards for offshore work*, surely improving those standards is where we should be looking.

Thanks to Colin Speedie for bringing this to my attention.

*I quote from the report:

ISO 12215-9 defines Design Category A as: ‘category of craft considered suitable to operate in seas with significant wave heights above 4m and wind speeds in excess of Beaufort Force 8, but excluding abnormal conditions such as hurricanes.

With reference to the hindcast weather data for 0300 on 16 May 2014, it is concluded that Cheeki Rafiki was operating within its design category criteria at the time of its loss.

Further Reading (Free)


This is a complex issue, please read my original in-depth opinion before commenting.


Please share using the links below if you feel that we just have to get a grip of this inadequate keel to hull joint issue.

Inline Feedbacks
View all comments
Dick Stevenson

Hi John,
To some large degree, I feel that having categories that give different levels of scantlings for “offshore” vs “coastal” cruising propagates a myth that is dangerous. A keel falling off near-shore might be less likely to end up in death than offshore, but likely only marginally so. Keels (offshore or coastal designs) should be able to hit a rock at the boat’s cruising speed without integrity compromising damage (and surveyors/repair people need to be more accomplished at diagnosis/repair). Although I will relinquish the point that boats and crew are often beat up worse offshore, I am clear that some of the Alchemy’s (and her crew) most challenging times have been near shore coastal cruising. The myth, to my mind, is that when you go out in a boat locally, that somehow your boat does not need to be as well designed/made nor does the skipper need to be as experienced. In general, this has clear validity, but obscures the fact that when out in a boat, events can take a difficult and deadly turn in ways that most of our activities do not dish out. Being out in a boat has some relatively unique aspects. Mother nature just has a lot more influence on a boater and there are just not always easy and safe alternatives (like pulling over to the side of the road if out driving). In this, I believe that the airplane analogy in your prior essay is relevant: long flight or short, high altitude or low, you do not want any compromises in structural integrity.
Just some random thoughts, not sure where they go.
My best, Dick Stevenson, s/v Alchemy

Marc Dacey

Good point, Dick. In-shore, one can find both harsher conditions and a 50% reduction of options (or worse, in a bay) than is the case offshore. I’ve seen cases of production boats (but not those built prior to circa 1980, interestingly) in which the tabbing on bulkheads has failed or the hulls have oilcanned in harsh conditions, which should not be the case anywhere, never mind Lake Ontario. I find the aircraft analogy apt and the glossing over of the physics versus dollars argument is a serious shortcoming of an otherwise incisive report.

If this is what “built to code” is, then we have issues with the code.

Rob Gill

Hi John,
I am interested in what merchant shipping law is deemed to have been broken, when the yacht apparently compiled with Cat 1 and the build standards as required. It will be interesting to find out though as I can’t help feeling this may be a beat-up – the regulators can’t be wrong surely!
However, as I read it Cheeki Rafiki had a mixed usage profile with periods of bare-boat charters and offshore skippered charters. Due to this mixed use, “failure of keel or rudder due to previous unreported grounding” should have been noted as having HIGH risk and HIGH impact in the organisation and vessel’s risk register. I do wonder if Storm Force Coaching had this covered with appropriate strategies for mitigation or minimisation?
Based on 15 years of chartering both in NZ and whilst working overseas (including the UK), I wouldn’t take any vessel that had been in bare-boat charter across the Atlantic, no matter how well equipped. We experienced a litany of failures in charter yacht portholes, rigs, sails, anchors and other equipment failures too numerous to list. The simple fact is that most charterers do not report major incidents for fear of losing their deposit. Scarier still, even if charterers do report a major incident, then this is often treated as “business as usual” for a charter boat. We came out of the Lefkas canal (Greece) under instruction to follow our flotilla leader in line astern and to our (and his) surprise, hitting a rock and coming to a dead stop from six knots – the boat was never even dive inspected. The leader’s comment – “happens all the time Rob, don’t worry about it”.

For a business like Storm Force Coaching, the only effective risk minimisation strategy I can think of would be to remove charterer deposits and mandate full incident reporting encouraged by compulsory replacement insurance cover with no excess fees. The insurance monies then being rigorously applied to inspection and repair. The only risk mitigation strategy I can think of is no mixed use (bare-boat charter / offshore charter) for any vessel – no matter how much the keel-to-hull build regulations are improved in the future.
Kind regards,

Alastair Currie

The UK Merchant Shipping Act is the law that is being referred to by Criminal Prosecution Service. The yacht is defined as a small ship and there is a responsibility placed on owners and / or agents to send a ship to see that is fit for purpose. If there is evidence that the yacht was not fit for purpose, or that reasonable information was available that that could be determined, then there might be a case against the owner / agent to prosecuted / defended against.

Dick Stevenson

Hi Rob,
That was a very interesting analysis.
Thanks, Dick


Hi John & others. For exactly the same considerations, which you are describing here, we have changed our mind last year, when we were shopping for a boat for our syndicate: initially, we fell in love with Bavaria 42 Match; but even the structural re-inforcement of the keel attachment, which was mandated on this type after a tragic keel-separation event, did not convince us about its safety. Finally, we purchased a Salona 37, because it has the keel attached to a steel structure, which distributes the load into the fiberglass shell.
I very much agree with your suggestion to change the code. But before that happens, I hope that the mere existence of this thread and the investigation reports would inspire the manufacturers to make improvements; or else they will face negligence claims. Compliance with the code alone is not enough for the shipyards to demonstrate that they have acted with due care. If Salona (and I understand that also a few other ship builders) have adopted a re-inforcement/load distribution steel structure, the economics must work somehow – why would not the others do it, too? If the sailing magazines will not write about it (for obvious reasons), websites like this can – and do – spread the message. And eventually, the customers would vote with their feet and this would make the builders change their way. 10 years ago, this would not have worked, because news did not travel as fast. It is different now, with webistes like this one. Who would like to buy a boat, which presents even a non-negligible risk of killing you and your crew due to structural failure? You don’t need to sail across the Atlantic to see this happening. The tragic Bavaria Match event occurred in the Adriatic. On the other hand, I may be wrong about buyers and reputation concerns changing the approach of the ship builders – after all, the Bavaria event is more than 10 years old now and they. And less-than-ideal keel designs are still being manufactured in many places. I wonder how many more sailors must be killed before the codes change. It is always so sad to see regulation where common sense should be doing its job in the first place.


While the stability/integrity discussion is paramount and necessary, I wonder why no one questions the placement of eventually life saving equipment. Not only on Cheeki Rafiki the liferaft has been placed in a position (below a cockpit bench) that had required substantial preparation, something that was not possible due to the immediate characteristics of the catastrophe.
At least when going offshore, better always, the liferaft should be deployable even from a capsized boat, and this is only possible IMHO when mounted externally, e.g. either on the transom or at a suitable position at the deck rail.

Ernest Vogelsinger

Hi John,

its not that I wanted to argue against the keel issue being by far the most primary concern, I completely agree with your thoughts. It is just that accessing a stowed-away life raft is in fact, even when we’re not capsizing, a very time consuming effort which might be the sometimes only few seconds between life and death.

I fully agree that it is more than doubtful that the Cheeki Rafiki tragedy might have come out less harsh with a functional/accessible raft, given the environmental circumstances at that time and location.

Dick Stevenson

Hi Ernest,
You bring up a dilemma that is not always easy to find discover a good answer for. Your espousal of life rafts externally mounted has a great deal to argue for it. That said, the vast majority of rail and deck mounts I have observed are what I consider near shore installations: installations which would not survive a bad fall off a wave or other deck sweeping events that occur on ocean passages. (A raft just makes for a huge amount of exposed surface area that most rails are not designed for when hit by a wave or a fall off a wave. Raft deck mounts are often just screwed to the deck, not through bolted with backing plates and where the cored deck has been reefed and filled with epoxy. The cradle is often flimsy as well.) Further, they are subject to the elements (many re-packs find water intrusion/damage) and to theft. I believe, for offshore boats, that most anything on deck not in use to run the boat (jerry cans, kayaks, solar on lifelines, etc.) are problematic. There are good solutions, but they often demand considerable foresight at the design level; a dedicated compartment with access from the stern swim platform for example.
We went with a raft which was custom packed in a soft container to fit our cockpit sail locker: lift the lid, then lift it straight out, all done in the cockpit, likely the most secure place on deck in an emergency. That covers the vast majority of raft usages, fortunately a rare event in most cruiser’s lives. In the even less likely event of the boat being inverted, it is necessary to dive down and open the locker. The raft should fall out. Be sure the raft inflating tether is attached to the boat so it inflates as it falls or it will just keep going. Needless to say, I have not practiced the last scenario.
Certainly some compromises embedded in the above, but is what we have come up with.
My best, Dick Stevenson, s/v Alchemy

Alastair Currie

Life rafts float in the deflated condition. If the upside down locker is underwater the life raft will not fall out. This is why they work with hydrostatic releases. I do not know how easy it would be to force a life raft down into the water and under an upside down coaming.

Dick Stevenson

Hi Alistair,
You bring up an interesting point. I suspect my custom configured and then shrink wrapped raft would actually sink deflated as it seems very dense and heavy for its volume. I know I am quite careful the few times I have transported it in a dinghy. It certainly would float when it got 10 meters or so out when the tether activated the inflation mechanisms. In the end, I think that, even if it had a bit of buoyancy (which I doubt), I would get it out and this compromise still beats the alternatives. Not something easy to test, though.
I do not see many private recreational vessels with hydrostatic releases, although I have paid scant attention to this detail. How do they stand up to the kind of solid water that is possible on a small sailboat offshore?
Thanks for your thoughts, Dick

Alastair Currie

My own liferaft, like many are now vacuum packed, it is more or less the standard method. Mine is in a solid container on my coach roof secured to thin but wide teak runners on the coachroof. There are 4 x straps that secure it to the runners which meet at a common point on top of the canister. The hydrostatic release is configured to activate between 1.5 to 4 m of head. A blade shears the hydrostatic release lanyard which separates the securing straps but not the painter which is secured to the yacht. As the yacht sinks, the canister floats up pulling out the line which eventually activates the inflation mechanism. I also have a pelican hook that allows manual release of the raft although I am changing that for Hammer’s special fast release hook as releasing the securing ring on pelican hook under tension can be hard. Both the painter and the hydrostatic releases have weak links in them so the liferaft in the manual or automatic release mode will break free once inflated if it is still connected to the yacht.

I have always fitted one as it is low cost assurance, about 14 beers at current UK prices, the only way to measure the worth of low cost safety devices 🙂

I don’t like the cradles as they are quite high and spindly. On my own boat big seas crash over the bow because she has a high and flared bow. I thought about mounting it on the aft cabin coach roof but that is poop territory, defo solid green potential. Everything is a compromise, is mine at risk of being washed away, yes of course, but it’s probably less of a risk in other areas apart from down below of course.

I think, the majority of emergency situations that require evacuation and are the most probable would allow a controlled abandonment to be organised. I only have one after nearly being run down in fog by a large coaster who was also well inshore. I remember thinking that boat would just have been cut in two if it had hit us. It just slipped out the fog, 50′ or so in front, slipped back in again, very silently. No radar or AIS back then, still don’t have them fitted, too many beers 🙂 Seriously, my instruments are due for complete replacement, so all that is on the cards.

I have no connection with CMHammer except as a customer.

Marc Dacey

Thanks for this information, and for the seamanlike units of measurement.

Eric Klem

I agree that the inquiry missed the point. We get so focused on safety gear which is designed to save us once a series of bad events has occurred that we often don’t put enough emphasis on preventing those series of events. This is often true of both the people who get in trouble but also of the press and inquiries afterwards. I did search and rescue in my early twenties and we would always complain when the press report would focus on something like the person wasn’t wearing a lifejacket totally missing the point that the person was trying to paddle 35F class IV whitewater in a rec boat with zero whitewater experience. We eventually complained enough that they changed the press releases somewhat but it really got no better as they would say that the person was experienced totally missing the point that 100 days on flatwater doesn’t do you one bit of good on whitewater. We really need to measure people and gear against the expected conditions and then understand that not everyone shares our views of acceptable risk.

I have seen a lot of discussion recently around whether boats need to be designed similarly for coastal and offshore sailing. I believe that this is a flawed argument as it would suggest that all daysailers are unsafe. The trick is in how we define coastal. If you are an experienced person daysailing in an open boat on a day with a good forecast on a forgiving coast, encountering heavy weather is simply not a major concern. However, if you are a delivery captain sailing along a bold coast sticking to a schedule, you may well encounter conditions similar to a storm offshore. Therefore, it is crucial that the skipper and crew have the experience and skill to determine what level a boat must be prepared to and have the discipline to stick to it. It is perfectly possible to encounter just as dangerous conditions or even worse near shore but the question is whether you go out in conditions where this is possible.

One of the hardest things in engineering is figuring out appropriate boundary conditions (loads, constraints, etc) for your modeling. In certain circumstances, this is fairly easy. For example, in an engine, the loads are all determined by how much fuel and air you put in and since you limit those, you can easily calculate a maximum cylinder pressure and design around that. Anything that interacts with humans or nature has loads that are much harder to predict. Grounding are more straightforward to calculate loads for than many things on boats but even they have lots of variables beyond the obvious one of speed such as where on the keel it hit, how solid the object was, was there any wave action causing a vertical component, etc. I would think that it would be a good idea to design keels to sustain repeated cruising speed groundings on rock but there will still be occasional failures from people who surf their boats onto rocks sideways and other things. Keels have the advantage that they are allowed to be heavy but there are many parts of the boat where you can’t do this and can cause other problems if you try to build for all eventualities. For example, if we designed a rig to handle all loads potentially thrown at it including severe knockdowns, we would get such a short and heavy rig that the boat would end up being a terrible boat and probably less safe. In the case of keels, I think you can design them strong enough to take any reasonable punishment but this is not necessarily true of all items on a boat.


Marc Dacey

John, for me the key point is not simply that the keels can fail for reasons of design and/or materials or construction choices, but “that damage is impossible to diagnose or repair properly”. I realize that a sort of “monocoque” construction, in which all the elements of the vessel (or car, or airplane) are tied together to create a unitary strength, has great advantages in streamlining and weight savings, and may be very appropriate in the rarefied world of sail racing, but cruising boats, even fast ones, must be “field and operator repairable” to be of use. That’s worth half a knot and maybe more to be able to access the keel/hull attachment points for inspection and, if necessary, remediation. How horrible we would find a situation where that particular area was an inaccessible “black box” and yet salt water was trickling in… from somewhere… beneath the glued-down framing and the skies to windward were darkening. At times, and despite all the good things that have come from it, I think the predominant model of “what shows up on race boats will be standard on cruisers a decade later” is a notion insufficiently examined by people who actually buy boats.

Greg Beron

John, after reading both articles and the associated comments, I would like to ask for another follow-up article focusing on what the average sailor can look for to determine exactly how a keel is attached to a boat. I’m not an engineer or a naval architect but it seems clear to me, for anyone planning to sail offshore, this kind of “matrix” construction is best avoided. But I only have a general idea of how to recognize what I want to avoid.


Not sure if it is the right place to post, but this happened in 2014. Not questionning the safety of three young childrens on a J-40 in the roarings forties… my question is how is is possible that a boat can be dismasted so easily by simply rolling in the water?


Hi John you’re right its a open 40, my mistake. I wonder though if someone has reviewed the myth, once the boat turns over the mast goes over. I’m checking the scantling of the rigging on my boat and i wonder if one day i capsize, if the rig will stay on or will go down. Does someone has ever studied this subject ?

Marek Nowicki

Hi John. This is the link to our friends boat repairs :

Making the long story short: after about 2 years of US and Mexico sailing they experienced an inexplicable case of keel bolts nuts not holding the specs. They have to tighten them twice within less than 2 years (once in the US and once in Mexico). This is what they found after they returned to Seattle via Hawaii…it is scary. IMHO this is VERY instructional case.

Marc Dacey

A cautionary tale. I wish I could say I was surprised at “Steve’s survey missed this as did mine.” I am not. One of the reasons we’ve taken so long to finish the refit is because I’ve had to learn how to spot problems before they become crises because I’ve had elementary things missed by so-called “experts”. Keels parting on modern designs may be part of this issue.

Dean Conti

Many words have been written about this disaster. I will not try to add to the large body of facts and speculation. Rather I would like to point out that valuable lessons appear to have not been learned. Last fall two friends were aboard a chartered Beneteau First 40 in the 2021 Fastnet Race. They were forced to abandon the race quite early due to severe flooding in the keel area. I watched the race start and while the conditions were challenging, I saw nothing that should cause a properly maintained boat to have a keel start to fail. I have been racing my boats for 35 years (including open ocean to Hawaii) and have seen my share of rough water. If I thought the conditions they were experiencing could cause a boat to fail, I would quit boating. My opinion is the charter fleet owners have not received the memo and are not adequately monitoring the condition of their fleet. Chartering a boat for Fastnet, Middle Sea or Hobart was on my bucket list but I’m not so sure now.