The Offshore Voyaging Reference Site

“Cheeki Rafiki” Tragedy, Time For Changes

Overturned hull of the Cheeki Rafiki

I think most everyone in the offshore sailing community is aware of the tragic loss of the Beneteau First 40.7 Cheeki Rafiki.

What We Know

Much internet ink has been expended speculating about what may, or may not, have happened to the boat and her crew. But two things we know for sure at this point are:

  • The keel is missing from the wreck.
  • The liferaft is in its original stowage position.

That would seem to indicate that disaster struck very quickly and in a way that prevented this experienced crew from taking steps to save themselves.

Beyond that, we can all speculate for months, as I’m sure will happen in the media and on the forums, but none of that will do much, or maybe anything, to prevent a recurrence. This post is my attempt to look at what we can do.

Cheeki Rafiki is Different

I always take a keen interest when a yacht is lost, and often write about it, because I believe that studying casualties at sea is one of the most important things we can do in our quest to sail offshore safely. But the Cheeki Rafiki loss is different. Not just because four sailors are missing but because, unlike with most losses at sea, I think that there is little anyone, including the crew, could have done to avoid this tragedy, given the current state of offshore sailing.

Don’t get me wrong, I’m not saying that a keel failure is acceptable, and I expect that the appropriate authorities will investigate and try to determine what happened, but what I am saying is this accident may have been inevitable, just as I believe that the tragic loss of sailors’ lives in the 1979 Fastnet race was inevitable, given the state of offshore sailing at that time.

I firmly believe that Cheeki Rafiki should be a wake up call for the offshore sailing community, just as the ’79 Fastnet was. And, once again, we should honour the memory of those who lost their lives by making some fundamental changes.

Keels Are Different

What fundamental changes you ask? Before we get into that, let me quote the late Don Jordan, designer of the Jordan Series Drogue and an aircraft engineer:

In the design of aircraft, certain machinery and equipment is in a separate category, “safety of flight”. These items must be absolutely reliable and must be capable of enduring the worst environment that the aircraft may encounter.

I would argue that the ballasted keel of a sailboat is the marine equivalent of a “safety of flight” item—an item that simply must not fall off for the expected service life of the boat. A sudden keel failure, particularly in heavy weather far from land, well, that’s the marine equivalent of the wings falling off an aircraft in flight—there is pretty much nothing that airmanship or seamanship can do to avert a tragedy.

Yes, I know. I have heard the suggestions that the crew of Cheeki Rafiki would have survived if:

  • The liferaft was stowed in a more accessible place.
  • The liferaft had had an automatic release.
  • Ditto the EPIRB.

And on it goes. To me all of that is like saying “don’t worry about the wing structure, let’s just make sure all the crew have parachutes and the escape hatches work”. No, a sudden inversion due to keel loss is such a catastrophic event, particularly in heavy weather far from help, that even with the best gear in the world survival is doubtful.

And yes, I know that several Open 60 sailors have survived sudden keel loss, but at least two have not; besides which, the crew of Cheeki Rafiki didn’t sign up for that kind of risk.

By the way, it is tempting to say that many more parts of a sailboat should be in the “simply must not fail” category but, if we did that, boats would become even more ridiculously expensive than they already are. And anyway, we can deal with potential failure in, say, rigs with frequent inspection and parts that fatigue with use (stays, for example) with regular replacement—not so keels.

And, after all, if the mast falls down the boat still floats. And even if the hull is ruptured, and the water pours in, there is a reasonable chance of getting into the liferaft. Neither are trivial but they are, in most cases, survivable.

What Changed?

First 36.7 keel 1

Before we look at what needs to be done, we need to understand why we have a keel problem. After all, up until a couple of decades ago, sudden keel loss on offshore sailing yachts was extremely rare.

What changed? I’m sure there are a lot of factors here, but I think a big cause is the move to keels that concentrate most of their weight at the bottom and couple that with a very small attachment area at the hull.

Don’t get me wrong, I’m not saying that every boat with a keel like this is dangerous, but there is no question that keels like this are much more difficult to design and build strong enough.

The even more worrying aspect of this is that many boats with these intrinsically difficult to engineer keels are now aging. And while their keel attachments may have been adequately strong when the boats were new, the ravages of fatigue and/or multiple groundings may have made a large number of boats out there at risk for sudden keel loss.

What We Need To Do

In this kind of situation it is always tempting to say “they need to do something”. The faceless “they” comprising boat builders, designers and regulators. But the bottom line is that it’s simply impractical to expect mass production boat builders to increase the price of their products, or cut their profit margins, to fix a problem like this without changing the market conditions that caused the problem in the first place.

Or to put it another way, we won’t get strong keels until we-the-market start insisting on strong keels. Just as we-the-market started insisting on more stable boats after the ‘79 Fastnet. And yes, I know that regulators and race organizers have a part to play here, but it is still up to us to support them.

New Boats

So what do we need to demand of the industry that supplies the boats we go to sea in, as it relates to keels?

Grounding Proofing

We must insist that the keels are engineered to withstand a grounding at hull speed without structural damage.

Fatigue Proofing

Materials fatigue, in which a structure that was plenty strong enough when it was built weakens over time due to cycle loading and then fails, has, over the years, caused many disasters, most notably in the aviation industry. But it does not have to be that way.

You see, engineers know, to a high degree of accuracy, how much strength common construction materials lose over millions of load cycles. And so it is perfectly possible to design keel to hull joints so that, even after decades of hard sailing, there is more than enough strength remaining to do the job. We must insist that we will only buy boats that are engineered this way.

By the way, fatigue explains a lot about why we have the failures we do on sailboats, so I strongly suggest you read Matt’s excellent article on the subject, and my follow up on how you can use that knowledge to make your boat more reliable and safer.


If enough of us question enough boat sales people about these criteria and insist on a well-reasoned answer, instead of fixating on the latest whiz-bang electronic or electric gadget, these improvements can and will happen, just as more stable boats (both static and dynamic) have resulted from the market’s insistence on them.

So what about the cost of all this? I’m no engineer, but I understand, from those that are, that incorporating the two criteria above would not be expensive in relation to the cost of the typical offshore boat, probably less than 1%—surely it’s worth it. And if every builder is pressured to improve keels, there are no competitive pressures to prevent any builder from doing so.

I also think that regulatory bodies should add the above requirements the offshore classification.


But what about the existing fleet? Well that’s a hard one, probably harder than solving the problem in new boats. But if we do indeed, as I believe, have a ticking time bomb in the form of thousands of boats with keels that, while probably adequate when new, are not adequate after a couple of groundings and a few million load cycles, we must grasp the nettle.

We must demand that those we hire to survey second hand boats being sold and boats that have suffered a grounding, carefully look for evidence of grounding in the former case and render an opinion in writing on the damage in both cases.

No, they won’t get it right every time, and nor should we expect them to, but they must be duly diligent in searching. Yes, this is not easy, but it’s not impossible either. There are tell-tale signs of keel problems to come—dings on the keel filled with putty and carefully painted over, evidence of patching around the hull to keel joint, furnishing and fittings that no longer fit properly—and surveyors should be looking for them, particularly on deep keel boats with small keel attachment areas.

My thinking is that every survey should include a section on examination of keel attachment with special attention to possible grounding damage and any evidence of flexing. In cases where there is any doubt, I would suggest this process should include suspending the boat off the ground and subjecting the keel to a known sideways force and then measuring deflection. And if there is still doubt the surveyor should withhold his or her report until the keel has been removed for full inspection.

Yes, these are big and expensive changes, but no more than the introduction and use of moisture meters and insistence that hulls be peeled and recoated was when customers realized that there was a major problem with water invading fibreglass laminates.

An Halls Spars engineer inspects a carbon fibre mast using an ultrasound machine.

Further, just as most of us started to insist that we would only patronize surveyors with moisture meters, we should be rewarding surveyors with more business who look for and adopt new and innovative ways to detect keel to hull joint problems.

The carbon fibre mast industry already uses relatively inexpensive hand held ultrasound machines to look for laminate damage in carbon masts that have been hit by lightning and I’m informed that it is likely that the same machines, in skilled hands, could detect laminate damage in keel to hull joints. We must insist on the adoption of this technology, or something like it.


This is my best attempt at learning something from the Cheeki Rafiki tragedy. I’m sure it’s not perfect, but one thing I do know, we in the offshore sailing community must not just return to business as usual or just shrug and say “they should do something”.  It is up to us to make something happen by changing the market forces that drive boat design and construction, and start looking at existing boats with a much more critical eye.

I would like to close by offering our sincere condolences to the family, friends and loved ones of the crew of Cheeki Rafiki.


I would be the first to admit that I’m not an expert in yacht construction and may have things wrong in this post. If so, I would be grateful to the engineers and naval architects in the audience for corrections, and particularly for alternative solutions. Please leave a comment.

If you are a lay person with suggestions relating to this post, I’m all ears, but please keep it positive. Just saying my ideas won’t work, without suggesting something better, does not help.

One other point on the comments. I will delete comments that contain wild accusations about blame or cause.

Inline Feedbacks
View all comments
Dick Stevenson

John, Nicely put. Just think how the sailing world would change if only 20% of the boat show guests asked the dealers etc “And what damage occurs when this vessel grounds at hull speed?” and followed up on that question.
Dick Stevenson, s/v Alchemy

Rob Withers

Fatigue failure in yacht keels is a big issue. When Hooligan V sank off Salcombe a few years ago, the MAIB report (worth reading: highlighted a number of problems:
– the keel was designed with minimum tolerances
– The keel was not built as designed (the diagram showing how it was built sends shivers down my spine – they might as well have stuck it on with glue)
– Extra weight was then added to the keel to improve performance (!)

The result was catastrophic failure of the keel in fairly benign conditions with 1 fatality.

Partly as a result, my father-in-law – who has been a structural engineer for 50 years and has a deeper understanding of these issues than anybody else I’ve come across – volunteered his services to the group redrafting the European Recreational craft Directive design guidelines – with a particular interest in getting them to improve the design of keels. I must refrain from speaking on his behalf, because I will inevitably be less accurate than he would like, but after much committee work (which he hates) the guidelines for keels have been improved – but not enough.

The main reasons, I gather, are 2-fold:
– the long term load variations on keels are not actually well understood. Keels are designed for a safety factor with a STATIC load (typically 4 times the load when the boat is at 90degs with keel flat, I think). The load variations – which causes fatigue – is not understood, or factored into the design.
– fatigue failure is crucially dependent on the details of design and fabrication. Small changes to joins, welds or corners can make a huge impact to fatigue life. It is possible to have well-designed keel that is weaker than the design spec which will not suffer fatigue failure – just as it is possible to build a keel that exceeds the design spec for static loads but will fail sooner because the details make it prone to fatigue. The factors affecting fatigue were, in the end, deemed too complex to include in the RCD guidelines, so they have some catch-all about designs should reduce the risk of fatigue by complying with the relevant ISO standard.

Going back to John’s post – ensuring surveyors have a better understanding of keels and their potential modes of failure is important, but if the keel bolts, covered in several inches of GRP, are suffering fatigue, it is going to be very difficult to spot from the outside.


John, I am inclined to say that you are being overly optimistic. Most people buying cruise boats, buy a used boat not a new boat so they are stuck with the design engineering in place at the time and how the builder interpreted those specifications and standards. Moreover, there is boat-to-boat variation within builder/manufacturer. In those days ISO (et al) were pretty much non existent. Net, you can do little about the existing stock of used boats. And, can you imagine the cost of retro—fitting a keel, a rudder, poorly laid up and cured polyester hulls.

Well, you could survey them, but how many surveyors are qualified to do proper structural surveys. I suspect very few. And, read some of the comments about Chapman’s survey school — your confidence in Chapman-trained surveyors will take a hit. Would supplemental training in the form of a 2-hour seminar really work? Read the disclaimer on a survey report — no surveyor in his/her right mind would stand behind the results of his/her work.

Where does that leave us?

Emilios P.

Hello All,
May I also add that the market gets the products that advertising/marketing has convinced it that it wants…
Most modern mass produced boats resemble go-fast racing designs to fulfill the fantasy of the less experienced buyer that he is buying a Ferrari at the price of a KIA… then he drives his KIA as if it were a Ferrari and inevitably gets in trouble… even the curvy/sleek/aerodynamic cabin tops of said boats are lethal when wet…
LONG LIVE the A40…
OFF TOPIC : “our” A40 is also, and by the original “declaration of intention”, a one design, so maybe she could be advertised/marketed as such ? A REAL offshore one design cruiser/racer for the real offshore…


I wish you were right about market driven but ask you to consider the automobile world: Virtually all of the safety and environmental changes were fought tooth and nail by the industry when first introduced. Government or quasi government (e.g., Lloyds, ABYC) have to take the first action and require the change. Ironically once in place the manufacturers flip and extoll the virtues of their ‘innovations’ (and then the go overboard — radar integrated braking, etc).

Years ago over a beer I was chatting with Bob Perry about the Valiant 47/50 which is as fine a sea boat you can find on the market and he lamented that he did not put enough beam into the design — that doing so would have made a more appealing boat to the market. I interpreted his comment as one about sales volume and NOT design per se. As someone else has commented, most buyers look at the amenities and not the sea keeping and structural aspects of the boat.


That’s a great positive outlook. Stronger keels and stricter offshore rating criteria is a start.

My positive suggestion is that off shore cruising boats should be based on commercial duty sailing vessels. Brute strength, stability, tracking, heavy weather comfort, load carrying, simplicity of systems, low crew loading and low maintenance must somehow become chic.

The root of the problem is the perception that certain performance features are universally good for all sailing craft. We as consumers eat it up. The naysayers are dismissed as eccentrics. But what is our definition of performance?

Formula One race cars have high tech light weight chassis, motors that have high very high power and efficiency per displacement, and over steer so they rip through a slalom. But the best motor homes are built on heavy duty, load carrying commercial chassis made to carry heavy loads for hundreds of thousands of miles with the lowest maintenance.

Why are cruising boats based on light weight low wetted surface hulls, high aspect rig and keel, with “highly responsive” helms? It’s an industry wide perception problem.

RDE (Richard Elder)

Without being too sarcastic, could I point out that there has never been a single catamaran that sank because the keel fell off. The Chris White 42 in the south pacific that inverted a year or so is the only relatively modern cruising design catamaran that I’ve aware of flipping, including thousands of condomarans designed with sailing as a very low priority.

Ask anyone in the bar about their objection to catamarans, and I’ll wager you will hear that they flip over. It would be interesting to have the actual statistics for boat loss and fatalities of flipped catamarans/vs monohull lost keels!

Everything in design is a trade off, but when we hang a ballast keel on a hull we must accept that it may sink. And the keel not falling off is an absolute necessity, regardless of how its shape effects performance or racing ratings!

Jean-François EEMAN

Hey Richard,

John might blame us for being of the subject and he is of course right. But
sometimes demysthification is important.

Time after time I read your comments and almost always they are, in my eyes very, to the point…
I’m a big multihull fan. I have crossed the Atlantic on a catamaran of 50 ft, I have sailed shorthanded ORMA 60 trimarans, I own a 30ft trimaran…
So no negative a priori but we have to dispel the myth that cats/multis do not flip. They do.
You are right to say when you flip a multi, you don’t sink and you can “always” get out through the compulosry hatch in the hulls…

By the way, see Evans’ comment : When Christophe’s TS52 flipped over he was himself on board with a very very experienced crew.
I just happened like that : Calm conditions, One gust, coming from the hills one late/wrong reaction and hop…
A few weeks before it happened, I delivered Pampero from Lorient to Belgium (600 miles) and I can tell you, at some moments, I had the impression, the gut feeling “the thing” can flip over. (I know it is so easy to say now, but we told Christophe when we stepped off the boat)

Evan Gatehouse

I agree that it’s the unknown loads that are the real problem. Airplane wing loadings are much better understood. And I suspect that most naval architects do not understand fatigue well, if at all, myself included. It’s not covered in depth at most engineering schools.

Getting off topic, but it’s also important to dispel myths: Even though I’m a cat owner and don’t really worry about capsizing there have been lots of cruising cats that have flipped:

– a Chris White in one of the Great Lakes (Michigan I think)
– Chris White 57 “Anna” off Tonga
– a Catana in the med
– a TS52 designed by Christophe Barreau here:
– a Fountaine Pajot 35 in the Gulf of Mexico (same owner capsized this boat twice)!
– A Voyage 440 off the coast of Oregon – bad storm
– A PDQ32 in the Bahamas (running through a breaking surf inlet)
– A PDQ32 off California coast

RDE (Richard Elder)

Hi Evan,
Thanks for filling in the gaps in my information. I’m sure there are more if we had the full records. (Although I’d certainly not call a PDQ a “reasonably modern” catamaran or by any stretch of the imagination an ocean going one—– with its narrow beam its more like an unballasted wide monohull.

Now that you mention it I do recall the loss of the Voyage 440. I’ve been ashore on that coast a number of times during winter storms and they can be unsurvivable by any small vessel. 85 knots of wind, thousands of miles of fetch, and a shelving lee shore.

Peter Dawson-Ball

It is not just how a keel reacts to fatigue that needs to be addressed. “They” also need to look at what happens if/when a keel encounters a submerged container, or hits a berg or is t-boned by another vessel.

Steffen Jacobsen

Dear John,
I think you have written a very good and sober comment on this tragedy. Since I don’t have any knowledge of the particulars, I will refrain from explanations, but It’s clearly not a matter of life raft position! Many blue water sailors don’t even carry one, but they rest assured that their keels and rudders just don’t break off. I live in Denmark, a member of the EEC, the home of Bavarias, Jeanneaus, Beneteaus and so forth. I look at them daily in our marina, all laden up with twin steerings, very short and vulnerable chain plates, undersized shrouds, wifes and children and dogs and no knowledge of the kind of boat, they’re actually going to sea in. Thay may be categorized in class A by the authorities for Open Ocean, but theyre clearly not constructed for the task. As a reader of Yachting Monthly in many years, the magazine is full of stories of French, German and English cheap production boats that just break up and fo to the bottom. Their keels are extrmely deep beacause custumers like speed, their hulls are triangular with huge sterns and the keel chords are ridiculously narrow.


Steffen Jacobsen

Deb Doyle

Not really sure what to say. The days when all vessels were heavy full keeled vessels are gone and not coming back. Keel design improvements I agree need to be evaluated for more then speed. Whether it was grounding fatigue I have no idea. Possibly poor metal in bolts. We will not know without samples. Your argument certainly holds merit for conversation. As most of us do that respect the ocean we try to mitigate the risks that we know we might confront us. Yacht racing gives more weight to speed than safety I fear. That’s where the committees (made up of racers) need to be the voice of reason.


Looking at the first photo and reading Rob Withers’ comments, Cheeki Rafiki keel loss reminds me of the famous trick used to be able to tear a complete set of playing cards.
If you handle a full set of playing cards using your both of your hands, without cheating somewhat, and you try to tear off this stack, you can’t succeed, because while tearing one card is easy, tearing a stack of 52 cards with your hands is not possible, unless you push the small sides of the stack to create small intervals between each cards. Using that trick, it is quite possible to tear of the complete stack with bare hands, because cards will start “breaking” one after another instead of all at the same time. Using more technical words: first card will exceed its flexion limit, and will start breaking, while most other cards are still far, or very far from that limit, which means that the total force needed to start tearing off the stack will be far, or very far, from 52 times the force needed to tear off one card.

On Cheeki Rafiki wrecked hull photo, we see 5 keel bolt’s holes (possibly only 4..), with 3 of them in good order, meaning corresponding keel bolts have broken, and 2 of them torn away, meaning corresponding bolts have been torn away from the hull without breaking (the hull has broken in the way of those 2 keel bolts, which are probably still good and attached to the keel).

I understand that the hull delamination we see on the photo started before the accident and that the 2 corresponding keel bolts were in fact not contributing very much to keel attachment, because corresponding keel bolts were deflecting much more than the 3 others, because of hull delamination (same trick as stack of card’s tearing…). In that case, the effort was carried only, or nearly only, by 3 bolts, and half of that effort by the rearmost bolt. If you consider that scantling has been calculated for 4 times static load at 90 degrees with 5 bolts, here the last bolt was taking half the total load, that is 2.5 times what it was supposed to handle, which means that because of hull delamination it was sized to support only 1.6 times static load at 90° instead of 4 times static load at 90° as requested. I understand that, at time of accident, dynamic load exceeded 1.6 times static load at 90°.

This looks like an accident waiting to happen without any obvious necessity of grounding, collision or whatever. If the wreck is recovered it might be interesting to find out whether hull delamination is normal or not (fabrication problem….), plus, I understand that this problem would probably not have happened with a thick reinforcing metal sheet covering the complete keel between keels nuts and hull, or with a factor of 10 instead of 4 between keel bolt scantling and static load at 90°

RDE (Richard Elder)

Hi Laurent,
This series of photos shows the keel bolt pattern and back up plate support for the First 40.7 design, along with another nearly identical failure.

Consider for a moment how this keel is attached:
—Narrow base both fore & aft and laterally. Since i assume it is an iron keel, it could have had a wide T shaped top inserted into a hull recess and wide bolt spacing at no sacrifice in cost or performance.

— The boat has a full liner with longitudinal stringers and ribs fabricated in a mold as a one piece layup—probably just a weak chopper gun part. This liner is then bonded into the hull using Plexus if the bean counters will pay for it, or resin and chopped fiber otherwise. The mating surfaces are inside to inside (relative to the mold face) and thus never will be a precise fit. The keel is bolted on, likely much later in the production process. This means that the keel bolt load is transferred from the (fairly small) washers through a stack of material that has a middle layer consisting of brittle resin or flexible bonding compound. If this is the way the boat is put together I’m amazed that all the keels don’t fall off!


A sailboat was lost in the Mauritius Durban race and when the upturned hull was found, 3 months later there was no keel. The inquiry’s findings were not made public! Great way to learn from mistakes. It seems to me that the small contact area of keel to hull is a major part of the problem. Would not a keel like that of the Flying Fifteen dinghy go a long way to helping this issue? ie; the keel has a flange at it’s attachment to the hull giving a far larger bolt-on area.


This disturbed me. The image of the boat with the keel off just cut through me, It’s got me all upset again. I’m still not over the rescue. I completely agree with you John and I just find it unforgivable. The Columbia 32, 2 accidents in 5 boats made. The first was rudder failure, a sailor died. The second in the Bermuda 1-2, the keel fell off. I know firsthand that Jan, the owner of Solid Air communicated to the designer and builder of his Columbia 32 that he felt a vibration in the hull prior to the race. They assured him it was OK. As an engineer of gravity switches for nuclear weapons, Jan has more than a clue. His concern was that vibration break everything, electrical or mechanical, everything. Had his keel failure been more rapid, he would have died as well. There would be absolutely no time. So what about the designer and builder now? where’s the accountability? They get to say oops, I missed that one. I’m so pissed off about this. They’re putting the rescuers at risk as well. I understand we are responsible for our choices but I can’t just blame the sailor. The Nader bolt is put into all car’s because the doors would open at impact. A solution was mandated to protect the public. I don’t think this is rocket science. This is old knowledge. The industry is trading safety and proven design for speed, profit and fame. Life rafts have mandatory inspection why not mandated load testing on keels? It doesn’t matter that this happened offshore, it’s unrecoverable. Lack of righting moment belonging to Dinghies. How about suspending Beneteau’s right to build and pay restitution. The public needs protection. Keels need a “Nader bolt”. Good article John. Keep the discussion alive.

László Szkrckic

The changes you suggest may improve the situation in the long-term, either when new boats are built or old ones are sold, but they do nothing to ameliorate the short-term risk to the existing fleet from keel failure.

There is one thing that could have saved the crew of the Cheeki Rafiki: a high index of suspicion. The keel failure was adequately foreshadowed by a serious leak of undetermined origin. If this had caused among the crew, and most particularly in the captain, a sufficient apprehension of impending doom, conservative measures may have prevented total keel failure, and certainly could have averted any loss of life.

I recently witnessed the young crew of another racing boat ground their keel hard six times in rapid succession (I counted). They were tied to a dock, over a rocky bottom in very thin water, when a sport fish went by at top speed. It was moving enough water to be reminiscent of the Fukushima disaster.

Their attitude after the incident was cavalier, to say the least. Where I would have insisted upon a keel inspection and a call to the local police, they shrugged it off as no big deal.

It is this attitude of casual invulnerability that leads to tragedy at sea and in the air. The only antidote is the kind of experience that cannot be compressed in time; you must survive enough years at sea to become the kind of pessimist who anticipates disaster in the strike of a keel or a mysterious leak.

In the well-regulated world of aviation, a wing strike (or even a hard landing) inevitably leads to an inspection. The experience of a century is codified into a set of rules thus, “if A happens, do B” so that young pilots may live to become old ones.

Offshore sailing is, comparatively, the Wild West.

Ian Tyler

A very interesting and well thought out explanation. Those photos of the de-lamination and failed bolts are horrific. Is this the natural difference between a voyaging and a cruiser/racing boat? Where unfortunately if you race on both sides of an ocean you have to voyage in between.
If any readers here feel like contributing, there is a charity collection in memory of the crew at:

Eric Klem

Hi John,

I think that you have come to a good logical conclusion on this. When possible, it is always best to solve the root of the problem rather than adding safety equipment to mitigate the bad effects when problems happen. In this case, it is absolutely possible to virtually eliminate these types of issues.

In engineering, you should always first come up with a set of requirements before actually designing anything. If these requirements are not nailed down before the design starts, the design tends to lack focus. I actually like to do it as user requirements in layman’s terms (in the case of a boat, where would it be sailed, in what conditions, by how many crew, etc) and then engineering specifications (length, beam, displacement, etc). I am certainly not a naval architect so I can’t speak to the exact design process for them but I understand that we go through many of the same things.

The user requirement should be largely created by the marketing department. In order to avoid overdefining things, this requirement must indicate which factors are absolute and which ones can be tweaked (does it absolutely have to cost less than X even if the keel strength requirement is not met or can the price go up by a bit to meet the specification for keel strength). One of the most challenging things is then translating the user requirement into the engineering spec. If the marketing guy says that the boat should be capable of being sailed in force 10 conditions, determining your keel loading in these conditions is no simple task. This is because each wave is different, the crew could be using different tactics, the boat could have a different history, etc. Realistically, you need to say that the boat should be able to use X tactic in these types of conditions and survive Y% of the waves without major structural damage even with Z deterioration in original strength from corrosion and fatigue (you have to have some percentage of cases where you allow a failure to account for the tail end of distribution of compounding factors or you will overbuild everything). A major problem with doing a good job on these specifications is that everyone wants to rush into the design and these actually take a significant amount of work to get right. In some cases, rules of thumb work really well and in others, they do not.

Once you have the spec, evaluating your design against it is relatively straightforward and accurate. Making sure that it is executed correctly is a whole different problem. Most of the design plans that I have seen for boats do not fully define everything, rather they provide a guideline. To fully define a boat, I would expect it would take several hundred pages of drawings and I think one great thing about Erik’s approach on the A40 is that he is going to go to this level.

I think that this was the really long winded way of saying that it can be really difficult to figure out how strong something needs to be in the first place, especially when dealing with the forces of nature.


Eric Klem

One thing that I forgot to add is that the user requirement is often never made available to the actual user which is really dangerous. In some cases it is but for boats, it is my opinion that it is not. Without this information, the user does not know what the designer’s intentions were.



Also a note for thought Sunsail Europe is selling off 4 Beneteau First 40’s for less than 90,000 euros ea. Proclaiming them as a racer but family friendly.

Marc Dacey

And so they are if you keep them fair-weather and coastal, as most do most of the time. Which is, I think, both the problem and the solution in one observation.

R. Todd Smith

This is a great tragedy and while I don’t know much about the sailors involved and I don’t ever want to presume to second guess the planning and actions of a crew I am not part of, with their experience level it raises some thoughtful questions in my mind:

1) I have sailed that model of boat, and while I probably don’t have anywhere near the sea time or technical experience of many commentators here, I found the basic boat substantially inadequate for offshore sailing in its basic handling and strength of build. That is just an observation, not a critique. Don’t get me wrong, I like the boat for the application I used it for, which was two weeks of wonderful island hopping around in the BVIs. Isn’t that why these are popular charter boats, that they are designed for that environment: inexpensive and used mostly in calm water for day sailing with copious living quarters and conveniences? Mine even had air conditioning! The wrong equipment in the wrong use is a problem for even the most experienced operator in any endeavour.

I guess people need to sail in what they have, but in the Spring Atlantic gales, I am not surprised that these types of boat are readily thrashed.

I do see these boats making passages successfully from the US East Coast to the Caribbean, but it seems a little like Russian Roulette to me, regardless of the preparation and crew experience.

2) Irrespective of the vessel, in terms of this incident, I see a glaring example of a simple missed safety practice that seems to occur over and over again in the past several years: No personal EPIRBs. (At least none that I am aware of here, please correct me if otherwise.) This is a $350 backup insurance policy at this point. Why would any offshore sailor be without one today? There is so much vessel traffic in the Atlantic today and such incredible response capabilities by various governments on both sides of the pond (as this incident well demonstrates) that if you lose your vessel or are swept from it, you still may have a fighting chance, at least for a time, if you wear your personal EPIRB, and your PFD of course, without lapse during the passage.

Even the best made boat can still sink and there are a number of enabling events that can be imagined in bad weather. I am a big fan of John’s credo of imagining that there is a 500 ft cliff on the other side of the lifelines. Staying in the boat is always Job #1. Still, I think that offshore crews -particularly during bad weather- should be as prepared as possible with options for dealing with loss or separation from the vessel due to a sinking or catastrophic gear failure.

We plan similarly for these catastrophic equipment failures in other extreme activities like mountaineering and deep technical diving.

Hopefully, these types of open discussions and accident analysis will prove useful in enabling others to be better prepared for catastrophic equipment failures since “they” aren’t likely to do much about current vessel construction and design practices.



Mark Morwood

One clarification – they did have PLB’s. The search started and the hull was located based on brief signals from two PLBs. No signal was ever seen from the boat’s EPIRB.

Marc Dacey

John, a good if sad post acknowledging a reality of modern boat construction, of which I have been both generally (bean-counter) and specifically (garbage and cheap and obvious shortcuts and defects I’ve seen first-hand at boat shows) for a few years now.

I’ve just had an email exchange with a good friend of mine who has his RYA Offshore cert and has been a professional mariner and has owned everything from Trintellas to Bavarias. He commented in an email yesterday that he was delivering a Beneteau down the east coast of Spain and that he was “coming around to your views about Beneteaus (and similar boats) not being oceanic. You should have seen the hull move and the interior cabinets flex in the swell. It was scary, the whole thing was bending all over the place. Even (the skipper) noticed it, he said the hull in the bow cabin where he was trying to sleep, in vain, was flexing by over one inch…The local yard owner, who set up the boat on behalf of Beneteau, said these boats were not designed for motoring head-on in rough weather. Another chap for the yard also said that keel-bolts are vulnerable to breaking under such loads. See Cheeki Raffiki.” I replied that “I personally think the market is now saturated with “not-good-enough” boats and we are seeing the results coming apart and giving SAR resources plenty of real-life exercise.”

I like “performance cruisers” as much as the next sailor, but it’s a very exacting set of metrics to meet in my view, and we are now seeing some of these boats reach middle age. I can’t help but dread that more of these sort of catastrophic failures will continue to occur…and I think the analogy with the dire IOR-style vessels of the 1970s…which were unforgiving after what seemed to be an arbitrary point of bad seastate and weather…is pretty solid. That said, the lightly built, high freeboard condo-style isn’t going anywhere…it’s too much a marketer’s wet dream.

Simon Fraser

I wonder if the solution is already in front of us.
Insurance companies are capable of examining matters of structural integrity and design from a business/ financial perspective, and providing a pragmatism that may be missing in more emotional analysis.
Insurance companies already make decisions that certain boats, or types of boats, shall not be insurable for any/ some/all voyages, and this helps to prevent some tragedies. If their analysis can be extended to include the information of the type discussed so far in this post, then many boats that do not pass their new criteria will not be sailed, or bought or built, as they are uninsurable.
Good insurance companies are of course informed by engineers, naval architects, and surveyors in these decisions, and may indeed be an appropriate broker of common sense into the lives of people who did not even know that they needed it.

Marc Dacey

A good illustration of this in action might be the reinsurance market, which are the firms that insure the insurers so that a big claim (a flooded city or a vast oil spill) does not take down several middle-sized insurers. While the general public and politicians are still babbling about the degree to which climate change is occurring, reinsurance firms commissioned studies as far back as the 1970s…you bet they believe it’s happening, because they are the first to track claims for climate-related losses. So, to look at this another way, are the boats failing because they are weak, or because the weather is getting stronger and/or less predictable? Or is it a combinational thing? Risk can’t be usually put down to one factor, such as “Firsts are built with weak keels” (although they may be…) but is often where several factors collide to make the failure.


Well said and informative. Keels are something I definitely will pay much more attention to in future sailboat evaluation.


Brett Anderson


I think your post is on the mark and, as an airline pilot for the last 25 years with over 15,000 hours in the air, I liked your analogy to safety of flight items in aviation. One thing that you didn’t mention though is that those safety of flight items are not engineered to never fail but rather are subjected to a rigorous inspection, preventative maintenance, and overhaul/replacement schedule which is adhered to religiously. Every commercial airliner goes through a major overhaul process in which the entire plane is practically stripped to the stringers after X amount of hours/years, in addition to much more frequent inspections and other preventative maintenance.

For sailboats, a rigorous pre-purchase survey is not enough, because many years can pass between those, years in which corrosion and fatigue take their toll. In order to insure the integrity of keel bolts there would have to be inspections on a regular basis, including occasionally removing the actual keel and possibly replacing the bolts at a set service interval. Of course this all takes money. It doesn’t matter how well engineered something is, at some point it may give way, because there is no way for an engineer to predict the exact loads it may be under over its lifetime. Only inspections and replacement at regular intervals can adequately protect against failure.

Ed Finn

I lost the keel -on my C&C 35. Struck a rock in an abandoned harbour in Newfoundland. Actually the keel partially separated from the hull, and the boat took on water badly. Fortunately we put her aground in shallow water before it sank out from under us. So you can imagine I have more then a passing interest in separated keels.
Look at the photo above, perhaps that’s all the evidence we will ever have.
The thickness of the keel – measured from Starboard to Port seems very narrow perhaps 6 to 8 inches at the widest point. Versus along the centerline it measures about 5 ft or 60 inches.
A strip of the hull is missing on the starboard .
Had it become loose because of an impact in line with the long axis , and broken completely off over a period of hours, then the hull would have more damage in line with the keel bolts.
Their would be more damage on the centerline, and keel bolt holes that were worn oblong.
So it seems this keel was impacted from the side, laterally, perpendicular to the length of the boat, fell over to the side relatively quickly, and peeled off the strip on the bottom over time.
The keel /hull joint is much weaker from side to side, then along the centerline of the boat. A relatively benign impact from say a whale , porpoise, or a partially submerged log, would cause the damage shown.
This might explain the rest of the evidence, the reported taking on water and the quick catastrophic failure, not allowing the crew to launch the liferaft…
Unfortunately many racer/cruisers built in the last 20 years, have this same design, and it looks like a serious and common flaw to me.
This issue – a narrow hull keel joint should be addressed by designers and insurance providers.

David Mitchell

I agree with Ed. Definitely looks like a side impact but the sailors did not report that they had struck something. Most likely this damage is from the lateral force exerted on the keel from being pushed hard (raced) combined with a lot of weight low in the keel.


Could be due to an insufficient scantling of the hull, or hull reinforcement in the keel area, or to a defect in composite materials, or composite materials setup


Hi John,
What a wonderful job you have done with this article. Thank you for such sobering information.
Is there anyone here at AAC that knows if Beneteau advertise that their boats are made for offshore passages? I don’t follow their advertising but I do not recall any adds that say our boats are made for bluewater. I’ve seen pictures of them in tropical waters in ads making you think they are made to cross oceans. But no ads like Mason used to have in sailing magazines which advertised, “Best boat for under 500,000 dollars to cross oceans in.”

While in the Canary Islands waiting a weather window this year I was sitting on a Beneteau talking with its owner. We were talking boats and safety for blue water passages. In part of the conversation about his boat he mentioned that there is a woman who owns the Beneteau company and once when she was asked by the press if she thought her boats were made for blue water she said, “never, they are not intended for that. ” I have no idea if that story is true but it brings us back to what do production boat companies think about their boats sailing across Oceans? Do they let the sailing rag magazines say they are blue water boats in their foolish boat of the year articles. Do the companies just keep quiet and let uniformed people do as they please with their product.

pat synge

Interesting discussion.

As a working marine surveyor the bulk of whose work involves inspecting and reporting on ‘cruising’ yachts I will pick up on John’s comments on priorities in this area.

I agree with him entirely when he says:

“We must demand that those we hire to survey second hand boats being sold and boats that have suffered a grounding, carefully look for evidence of grounding in the former case and render an opinion in writing on the damage in both cases.

No, they won’t get it right every time, and nor should we expect them to, but they must be duly diligent in searching. Yes, this is not easy, but it’s not impossible either. There are tell-tale signs of keel problems to come–dings on the keel filled with putty and carefully painted over, evidence of patching around the hull to deck joint, furnishing and fittings that no longer fit properly–and surveyors should be looking for them, particularly on deep keel boats with small keel attachment areas.

My thinking is that every survey should include a section on examination of keel attachment with special attention to possible grounding damage and any evidence of flexing. In cases where there is any doubt, I would suggest this process should include suspending the boat off the ground and subjecting the keel to a known sideways force and then measuring deflection.”

But disagree with him when he says “And if there is still doubt the surveyor should withhold his or her report until the keel has been removed for full inspection.”

It is my responsibility to describe the condition of the vessel as far as I can ascertain it and present my findings to my client. When I have concerns these should be clearly and unequivocally expressed in my report and appropriate recommendations given. That is all. It is up to my client (and their potential insurers) to then make their own decisions. They may choose to engage me to do further examination, engage in consultation or take advice elsewhere. It is not my role to “withold my report until the keel has been removed for full inspection.”.

And where lies my responsibility in commenting on design factors that I think essentially unfit for purpose?

Is it my role to tell clients that a professionally designed and built production yacht is an accident waiting to happen despite finding no defect? I personally really dislike the way many modern boats are put together and agree with the comment above about “Russian roulette” but when the relevant classification organisation deems it suitable for offshore use is it my role as a surveyor to question this when I can find no defect?

As for John’s comments about the use of moisture meters I will make the point that at least 95% of the hulls that I inspect that have just been removed from the water show extremely high readings and, without keeping them out of the water for at least 48 hours, moisture meter readings are simply not very relevant (though they are essential when undertaking repairs). In practice, percussion testing (while far from perfect) is the single most useful diagnostic tool for finding soft or delaminating FRP in a ‘lift and hold’ scenario.

Ah, the beauty of metal boats. Even when relatively badly built most steel boats have immensely strongly attached keels – not to mention chain plate arrangements.

Depending on a bolted on appendage to keep your boat from capsizing is always going to pose problems and I very much look forward to seeing Erik’s solution for the Adventure 42.

Are you absolutely sure it would be prohibitively more expensive to build the hull in aluminium using the best of modern production techniques given the savings that could be made in the area of the keel, chain plates, mooring bitts, mast steps etc etc? Have you really explored this or has the perceived market demand for a round bilge hull perhaps been a factor?

Alan Teale

Dear John and All, As someone else has already commented, the yacht concerned has a European Recreational Craft Directive categorisation of “A Ocean”. In my view, this is misleading at best. The ECD is a piece of trading standards legislation that I have heard said was initially designed to make it more difficult for non-European builders to break into the very lucrative primary market of the European builders. Much of it is the product of poorly informed committees, and much of it does nothing for safety. Further, I suggest that in the case of yachts of the type we are concerned with here, it is contributory to loss of life. A few years ago a Beneteau Oceanis (sic) of some length or other was knocked down or rolled in Biscay in winter. A crew member was lost. The MAIB enquiry concluded a vessel of its kind should not have been in Biscay in winter. It was an ECD Category A (Ocean) design. I am afraid that with the Cheeki Rafiki we are in some ways here again. Why?
Concerning another aspect of the loss we are discussing, I am not sure I would have gone to sea with a crew of four and a 12-man liferaft. Our 4-man weighs 42kgs and we stay in training to be able to throw it around. I am not sure where Rafiki’s raft was stowed, but if it was below I am not entirely surprised it was never deployed. And if it had been deployed, would the crew have been able to right it or hold it down in a really strong gust from the wrong direction? Alan

Marc Dacey

I agree with the implication that the Category A (Ocean) cert is misleading. I have seen any number of dock queens and boat show specials with construction details I found appalling in something allegedly fit for ocean conditions, and every year there are reports of any number of circa 40 foot production boats having to turn back or be abandoned…sometimes in the first few hundred NM… with detabbing bulkheads, severe “oilcanning”, leaking hatch frames from hull flexing, and bent rudder shafts/water ingress.

I also agree that it’s better to have four bodies in a 4 or 6 (my preference) liferaft than trying to get something possible over 100 kg. deployed in rough conditions, but that’s probably a separate discussion.

Denis Bone

Dear All,
This is not intended to be a negative or derogatory comment but I must say that the more I read of this excellent and thought provoking string, the happier I am that my boat, a forty four year old Fisher Freeward 30, has fully encapsulated ballast and a long keel. She may not be as fast or as responsive as modern boats with bolted on deep ballasted keels but she is not drastically slow, certainly fast enough for my relaxed style of cruising. I have, through my own error, had a direct collision with Norway, a very unforgiving place, which put her very much aground on a rock without any significant damage other than about a foot of epoxy scraped off at the point of impact and a very red face plus a lesson learned, I hope!
Racing and cruising are two different things, perhaps for cruising we should focus a bit more on safety and simplicity at the concept stage rather than allow our desire for speed to drive us to enter the extremely complicated and apparently uncertain area of bolted on keels that is the subject of this string?


In keeping with your “safety of flight” comment, would we knowingly fly in an airplane that had its wings bolted onto the skin of the fuselage? We have known for more than a century that a carry through structure connecting the wing spars is essential for take-offs to routinely equal landings. Obviously there are exceptions to such a broad statement, but most are in museums or junkyards.

We have a keel stepped mast rather than a deck stepped one for this very reason.

Looking at the photo it’s pretty clear the keel snapped off laterally and peeled away the skin from the hull. Why would we not consider “deck stepped keels?” Modern high modulus. low weight materials would allow us to carry one or two keel-embedded “spars” from inside the keel through the cabin and secured to the deck. The weight penalty could be negligible with sound engineering, and the strength increase immense.



My design engineering work came to an end decades ago, so I bow to Erik, but one thought to consider is traditionally attached fin keels will see some amount of flexion at the keel-hull margins. Clearly it is possible to render these effects minimal by spreading the loads widely as you say, (or adding a lot of mass to the joint) but in the end it is still a “T” joint with attendant stress and fatigue issues.

With a “spar” approach the keel-hull interaction is lateral compression rather than flexion. Webbed ring fittings attached to the hull through which such spars would pass would see only compression through 360 degrees of loading. This ring fitting would also spread the load but without flexion. At this point, keel bolts become secondary stabilizers rather than primary load carriers. The fitting on the underside of the deck would essentially be a pinned cantilever clamp.

Just the thoughts of someone who has hit the hard at hull speed thanks to a barge losing concrete debris in a channel and not reporting it.


Just a few weeks ago I was replacing badly rusted mild steel backing plates on stainless steel keel bolts – mild steel backing plates that had been installed by the builder! Fortunately the owner is conscientious and was willing to pay to have the work done as soon as it was discovered and before disaster struck.

Ed Finn

Agreed on bringing the keel up inside the boat
Deck stepped keels. Radical thinking, but a good idea!