[caribdis]

Peter Swanson is a journalist of the old school – and a sailing nut. After 20 years as an investigative reporter and editor for newspapers in his native New England, he worked for specialist yachting publications before launching his own newsletter on Substack, Loose Cannon, with the aim of publishing stories “the boating public cannot find elsewhere”. On 30 August, he had an exclusive. Swanson had got his hands on a document which, although rather boring on the face of it, would cause consternation in sections of the boatbuilding world and blow the debate over how Bayesian foundered wide open. A vessel’s stability booklet is a sort of operating manual for the captain or master. Across 89 pages of mostly tables and graphs, it set out the physical limits to which the vessel could be subjected before certain important things happened.

“The stability document landed in my lap,” Swanson told The Telegraph. “The motivation of the guy who leaked it? He’s probably a friend of the captain of Bayesian, I don’t know. Nobody had any faith in the Italian justice system not completely screwing the captain and the crew. You could see that coming out with Costantino’s statements.”
The Bayesian's Stability Information booklet, provided to Peter Swanson by an anonymous source
The Bayesian's Stability Information booklet, provided to Peter Swanson by an anonymous source

Among the fraternity of naval architects and other maritime professionals who make up much of Loose Cannon’s readership, two criteria stood out. One was what is known as the angle of vanishing stability (AVS), also referred to as the limit of positive stability. Simply put, that is the angle of heel – lean or tilt, in landlubber parlance – at which a vessel may capsize, rather than naturally righting itself. For Bayesian, AVS comprised two figures at maximum load: 72 degrees with the keel retracted, 92.3 degrees with the keel down. The other key criterion was the downflooding angle (DFA), the angle of heel at which water might begin entering the vessel: 42.7 degrees at maximum load with the keel up, 43 degrees with the keel down.
Serious trouble

Stephen Edwards captained Bayesian between 2015 and 2020 and knew the Lynches well. He stressed that the yacht was “sound and seaworthy by design”. But he added: “The downflooding angle is much more important… in the scenario we are talking about [in contrast to AVS]. This is the angle of heel at which water will start to enter the vessel, usually through the engine room or accommodation ventilation ducts. Once this starts, the vessel is in serious trouble as stability is quickly reduced or lost due to the flooding.”

Ventilation ducts? This is the point at which the tussle over Bayesian’s essential purpose comes to the fore. It was built to look and, yes, operate as a proper sailing yacht. But in reality its function most of the time was as a motor vessel supporting, in effect, a boutique luxury hotel. Its twin Caterpillar C32 diesel engines required significant ventilation to run safely. The yacht also had a sophisticated heating, ventilation and air conditioning system (HVAC) to keep its guests at the peak of comfort. Captain Edwards has no direct knowledge of what took place on board that night, but in a statement shortly after the vessel went down, he strongly implied that at least some of these vents, located in the main shell of the hull, might have been open. It was a warm night off northern Sicily, after all. “Unless the vent dampers are closed, which with HVAC systems and generator running they would not be as they need to be open for that, the vessel will start to flood rapidly if heeled more than the downflooding angle,” he said.
The angle at which a vessel may capsize

Swanson and his friends were astounded. Both metrics, but particularly the downflooding angle, seemed surprisingly low for a sailing vessel. Phil Friedman ran a shipbuilding yard for years which turned out 50-metre yachts. He said the superyacht’s AVS appeared to be considerably lower than in smaller offshore sailing yachts, which can heel to 120 degrees or more before capsizing irretrievably. Meanwhile, a downflooding angle of around 45 degrees (assuming vents open) was, he said, “surely insufficient for a sailing vessel that regularly runs offshore in open, unprotected waters”. This aspect of the design was, while not technically unstable, in his view “surely not a good idea”.
‘Size matters to a superyacht owner’

Why was the downflooding angle so low? The answer, a number of experts have argued, is to be found in Bayesian’s standout feature: its mast. Perini Navi built a series of similar vessels also using designer Ron Holland, but at the time of order, Bayesian, then Salute, was the only yacht in the sequence not to be built as a ketch, i.e. with two masts. This was at the request of the original owner, who wanted a statement, a single focal point of towering elegance for his boat. Perini Navi’s answer was 72 metres of aluminium, the tallest mast of its type in the world and the second tallest overall. It was also extremely heavy, creating a relatively high centre of gravity which had to be counterbalanced by added ballast in the keel box – an extra 30 tonnes, according to Edwards, part of more than 200 tonnes of ballast on board overall, including the keel itself.

The combined weight meant Bayesian sat low in the water, pushing its vent openings close to sea level, even before any heeling took place. For Chris Freer, a veteran yachtsman and boatbuilder, the problem was one of ambition – too much of it. “I’m sorry to say that size matters to a superyacht owner, and naval architects are seduced into providing solutions,” he said.
The crucial placement of air vents

None of this, however, explains what caused the yacht to heel so dramatically in the first place. Here we must return to the perplexing sight confronting Börner and his crew aboard Sir Robert Baden Powell in the minutes approaching 4am on 19 August – that shroud of furious rain engulfing their close neighbour, the storm within a storm. In the week after the sinking, Italian authorities said they believed the culprit to be a “downburst”. Also known as a microburst, this is a meteorological phenomenon whereby air that has risen to a great height suddenly plummets towards the surface, usually in conjunction with rain. It can hit the ground or sea at 100 miles per hour or more – tornado speed – before barrelling outwards horizontally. Often downbursts last only a handful of minutes. They are also highly localised, with the column of downward air sometimes just a few hundred metres across. Meteorologists in the US were becoming increasingly aware of them throughout the 1970s, implicating them in a number of air disasters in the west of the country, usually over hot ground. But awareness of their danger at sea began to take hold mainly from 1986, following the Pride of Baltimore disaster, where a replica 19th-century clipper was blown straight over by a freak microstorm and sunk, claiming four lives.
The most contentious question

All sailing-boat mariners understand that violent winds can be dangerous. However, Bayesian’s sails were down. Here we come to arguably the most contentious question of the whole mystery, an idea that is being staunchly, indeed legally, resisted by the manufacturer. Even without sails, could that extraordinarily eye-catching mast have generated enough surface area, along with the rest of the vessel above the waterline, for the gale to have leaned the yacht to its downflooding angle? Guillermo Gefaell, a Spanish naval architect and consultant, believes so. He conducted computer modelling to calculate the effect on stability of the wind Bayesian might have faced that night. He included in his calculations not just the mast but the six pairs of spreaders, the horizontal bars connecting the mast to the rigging, as well as features such as the satellite navigation domes – anything generating what is known as “windage”. Along with a colleague called Juan Manuell, Gefaell concluded that a gust of around 60 knots or more might have been enough to push Bayesian to an angle whereby water would begin flooding in through the vents.
How tall was the Bayesian?

Peter Swanson aggregated the modelling on his blog. “The mast was too high,” he told The Telegraph. “It was the original sin of this disaster. It’s a sin of arrogance, pride, however you want to put it.”

The Italian Sea Group strongly disagrees with this. Indeed, after The New York Times published an article last month which referenced Gefaell’s model, the company announced it would sue the paper. This month, TISG insisted again that the crew should have closed “all openings” as soon as the weather deteriorated; they should have summoned all passengers to the muster point above deck and sounded a siren; started the engines and “thrown the chain”. The firm denies any structural weakness in Bayesian’s design and stresses that the yacht complied with MCA (Maritime and Coastguard Agency) stability criteria. The company has stated that a 60-knot crosswind on the night would have heeled the vessel to just 18 degrees, “and therefore in adequate conditions cannot lead to the immersion of any downflooding point”. It said at 80 knots of wind the heel angle would be 28 degrees. TISG also said that vents in the hull were “in any case protected by remotely controlled watertight closures”.

“As repeated several times by all the experts, it [the sinking] happened because a massive quantity of water entered,” a spokesman told The Telegraph. He added: “With reference to the issue of the angles and inclinations of the ship, I do not believe that any newspaper is responsible for technically establishing what happened.”
The violence of the storm

Was the keel up, thereby reducing vital stability? Again, we don’t know for sure, but most commentators believe that it would have been, as that was normal practice in sheltered water. There are rumours that, when deployed, it made a repetitive thudding sound when Bayesian was at anchor, which might have disturbed the guests.

Bill Prince, a naval architect, stressed the violence of the un-forecast storm as a principal cause. “It takes an act of god, and maybe a little bit of hubris by the crew and guests, to make this happen,” he said. Prince added, however, that while Bayesian had “massive ballast”, its ratio of ballast to gross weight was relatively low for a sailing vessel. “It’s on this continuum of not quite being a sail boat,” he alleged. “It has a mast and sails but it doesn’t have the level of reserve stability.”

Roger Long, yet another US east-coast veteran boat builder, went further. He alleged that if the wind was strong enough to push Bayesian to the point at which it started to flood, just a “6.8 per cent increase in velocity would be enough to lay the vessel right over on her side”. His point implies that downflooding was less important to the capsize of Bayesian – that the culprit might have been wind strength alone. Ultimately, he says, Bayesian was committed to capsizing at just 50 degrees.
‘Like a major earthquake’

So much remains unknown. There are multiple investigations: the Italian judicial proceedings, a probe by the UK’s Marine Accident Investigation Branch, which dispatched four experts to Sicily in the days after the sinking. In Ipswich, near the Lynch’s Suffolk home, inquests have been opened into his and Hannah’s deaths, as well as those of Jonathan and Judy Bloomer. Lynch’s cause of death has been given as drowning. But in the case of Hannah, the talented young woman who loved Shelley, the Romantic poet who died at sea, the investigation continues. The inquests will formally resume on 15 April.

They are unlikely to piece together definitively what took place below decks in those dark, disorientating minutes before and after the vessel sunk. Hannah, we know, died alone in her cabin. Concealed between two floating mattresses, her body was the last the divers found. Five others – her father, Jonathan and Judy Bloomer, and Chris and Neda Movillo – were discovered together in a separate compartment. Girolamo Bentivoglio Fiandra, of the fire service, said they may have been seeking air pockets on the higher side of the ship, or they gathered together for comfort. The body of 58-year-old Recaldo Thomas, the chef, was found outside the vessel.

James Cutfield has returned to his home in Majorca, where supporters have launched a fundraising drive. In common with his colleagues, he remains silent. Italian justice moves slowly, and until then some will continue to question, as Giovanni Costantino does, why the crew did not react differently. Why, in the 15-plus minutes between Bayesian starting to blow side to side and the location beacon being lost, could the crew not have made the vessel fast, weighed anchor, used the engines to point into the wind – if indeed they did not.
Giovanni Costantino, founder and chief executive of The Italian Sea Group
Giovanni Costantino, founder and chief executive of The Italian Sea Group Credit: LaPresse / Alamy Stock Photo

Jan Miles, captain of the replacement Pride of Baltimore II, is a man who understands all too well the horrifying power of a downburst, having lost friends in the 1986 disaster. He cautioned against second-guessing. The initial phase of such an event is, he says, like that of a major earthquake, prompting an unavoidable moment of confusion and inaction. But after that: “They were reacting, clearing the deck,” he said. “They weren’t standing there twiddling their thumbs.” Ultimately: “Did they tip the boat over or did the wind tip it over?”
A greater chance to escape?

A finding that any hatches in the hull were left open, as Costantino has alleged, would be truly damning. But at least one photograph, taken shortly before the disaster that night, plus underwater footage, appears to back up the crew’s case that the main hatch was closed. For his part, former captain Stephen Edwards believes it to be very unlikely. Examination of Bayesian’s water-tight compartments will also be a factor. The hull was divided into five subcompartments and designed to stay afloat if one flooded. Some were entirely self-contained, but others were only water-tight with the right doors kept shut. Does a flooded engine room explain the sudden blackout and the death of the location transmitter?

Bacares’ evidence, when it comes, could prove critical, both because – we think – she was on deck at a relatively early stage, but also because of the moral weight any words condemning or absolving the crew would carry. In the meantime, she inherits the financial horror of her husband’s legal row with Hewlett-Packard, with the company confirming after the sinking that it would continue to pursue damages of up to $4 billion.

Questions for the industry persist. Bayesian appears to have been designed and built within regulatory parameters. Its design had been approved not just by the MCA but also the American Bureau of Shipping. Legally, it was stable. Yet a number of people who have spent a lifetime thinking about these things have been shocked to discover, as they see it, the boat’s vulnerability to heeling and downflooding. It is not simply an issue of the vessel’s ultimate demise. If – and of course we don’t know – the downburst had succeeded in knocking Bayesian flat without immediate largescale flooding, it would likely have stayed on the surface for longer, affording those still below a greater chance to escape, rather than, as some believe, the matter of seconds they had. Why, experts have asked, could the mast not have been built of carbon fibre, which might have been 30 to 40 per cent lighter? Why could the engine and HVAC vents not have been piped above deck? “It may not be as elegant in appearance or as in keeping with the panoramic windowing, but it would be significantly safer,” said Friedman.
The terrifying power of a weather 'downburst'

Moreover, sudden downbursts, the potential for un-forecast tornado-strength wind arriving from multiple directions, have been known about for decades. They are rare, yet they have been implicated in numerous air disasters and at least one sinking of a major sailing ship. They are made more likely by a warmed-up surface, and the Mediterranean has never been warmer. Should the megayacht industry have these more in mind before designers put pen to paper?
Storm of speculation