SPECIAL REPORT: Analysis of a capsizing

Posted on

sorensen_2011EDITOR’S NOTE: As the formal inquiries into the July Fourth Long Island Sound tragedy continue, Soundings asked Eric Sorensen, a regular contributor, to discuss boat design principles, especially as they relate to seaworthiness and the physics of buoyancy and stability. Sorensen is the author of “Sorensen’s Guide to Powerboats”; consults for boatbuilders, the Navy and boat owners; and was the founding director of the J.D. Power and Associates Marine Practice.

It will be some time before all the facts are known about the Fourth of July capsizing of the 34-foot 1984 Silverton Kandi Won with 27 aboard, but speculation as to the causes — overloading, a sudden weight shift, a course change, passing wakes, even mechanical failure or a through-hull letting go — continues. However, I think it would be most constructive to look at the tragedy holistically, including the physics of stability.

The No. 1 factor governing a boat’s carrying capacity is stability. To remain upright and floating, a listing vessel relies on buoyancy pushing up on one side and gravity pushing down on the other to restore the hull to equilibrium on an even keel. The force of buoyancy is created by the weight of the water displaced by the hull, and when the boat is on an even keel the center of buoyancy (CB) is on the centerline. When the boat heels (from a dynamic force such as wind, wave, wake or momentum in a turn) or lists (because of a weight shift to one side) CB shifts in the same direction as that side of the hull immerses more deeply. This offset buoyant force creates a righting arm, working with gravity to create a couple (a rotation caused by an offset force times distance) that restores the hull to equilibrium on an even keel. The lower the boat’s center of gravity with all passengers and gear on board, the more stable it will be.

An inshore planing hull such as the Silverton’s has little depth below the waterline, and it has a correspondingly high center of gravity (CG), with most of the boat’s mass well above the waterline. Because of the planing hull’s hard chines and wide, flat bottom, it feels very stable at initial angles of heel, but this is deceiving. However, the deep-draft displacement boat, with its much deeper and often ballasted hull — and commensurately lower CG — may feel more tender when rolling underfoot but is ultimately far more stable, with positive stability through perhaps 100 or 120 degrees.


A planing convertible such as the Silverton would likely capsize at 60 or 70 degrees of list, but it would also start downflooding into the engine room through-hull air intakes and capsize well before it reaches 60 degrees of heel. Once any vessel reaches its point of vanishing stability there is no remaining righting energy, and the boat will teeter between capsizing and righting, the result depending solely on the vagaries of the next wave or gust of wind.

The design of a planing boat such as the Silverton is such that, no matter where you are standing on board, your own body’s CG (roughly at your bellybutton) will be above the boat’s CG (see chart showing VCG with and without 27 passengers aboard). This is never a positive thing because a boat’s CG will move in the direction of a weight addition and away from a weight removal. And, as mentioned, a higher CG makes a boat less stable since the righting arm and the resulting restoring couple are diminished.


The cockpit on a local Silverton 34 sister ship I measured is on the high side — 17 inches off the waterline, which on this cruising-oriented boat permits a single-level deck to the cabin bulkhead and a flush deck above the gasoline inboards. With the cockpit deck — let alone the foredeck and flybridge — already slightly above the boat’s CG, each person boarding that night made it less stable still.

The stability math is pretty simple. If a 200-pound person’s CG when standing on the cockpit or pilothouse deck is 4 feet above the boat’s CG, 800 foot-pounds of energy is effectively removed from the boat’s righting arm. If the person climbs 7 feet up to the flybridge, our 200-pounder diminishes the boat’s righting energy by 2,200 foot-pounds. In other words, a person anywhere on a boat such as this, except perhaps when lying in the forward cabin, makes the boat less stable. Note that a 40- or 50-foot Navy utility boat — designed to transport dozens of sailors — has a single deck as low in the hull as possible so that it’s nearly as stable when weight is added as when unloaded.


A corollary to the addition of weight above CG in any vessel is that the removal of weight below CG produces precisely the same effects and consequences. If you take an old 2,800-pound GM 671 out of a boat and replace it with a new 1,100-pound diesel, CG rises as it moves away from this weight removal, making the boat less stable. So be careful when repowering an older diesel boat with a new lightweight engine.

The Silverton, according to online classified listings, weighs 12,500 pounds dry, which means it would displace about 14,400 pounds loaded with 260 gallons of fuel and 40 gallons of water. Based on news reports, and assuming 17 adults at 185 pounds each (the new Coast Guard weight figure for well-nourished adult Americans) and 10 children at 80 pounds each (a figure I made up), I come up with a total of 3,945 pounds of people on a 14,400-pound boat. Using these figures, having 27 people on board adds (coincidentally) some 27 percent to the weight of the boat. That’s a very large amount of weight to add topside to any 34-footer.


Depending on where the passengers were located, it seems possible that Kandi Won could have capsized when it was heading out earlier that evening if it had encountered a large enough wake, made a turn and changed speed, all at the same time. The fact that it did not capsize until the fireworks were over suggests that it was not subjected to a sufficient upsetting moment — some combination of a rise in CG (the excess passenger load) and dynamic events (the wake, turn or speed change), or perhaps a mechanical or equipment failure — until later. You can parse this any way you want to. Maybe if it had 20 or 16 or 12 people on board it would not have capsized. But if it had a more reasonable load of eight or 10 people, with two or three on the bridge and the rest down on deck, and barring some unknown factor, such as a hull breach with flooding, it almost certainly would not have capsized in my opinion.

Seaworthiness is relative

The degree of stability of any vessel in a given condition of loading is a key measure of its seaworthiness. If it lacks adequate stability, it is unseaworthy. However, what exactly is adequate stability? The degree of stability reasonably required depends on what the boat will be doing and when and where it will be going. Adequate stability in the case of an inshore cruiser such as the Silverton 34 means it can resist capsizing when encountering seas, wakes and other dynamic upsetting moments that could reasonably be encountered.

I am working with an ad hoc group of naval architects — Dave Gerr, Eric Sponberg and Steve Dalzell — to analyze the Kandi Won capsize academically (we don’t yet have an accident report), constructing a virtual model (we do not have the actual hull lines) of a boat much like the Silverton 34, with similar size, draft, displacement, freeboard, deck areas and deck heights. We have found that with 27 passengers on board, and guessing at actual passenger distribution, the resulting stability calculations appear to satisfy the requirements of the 46 CFR (Code of Federal Regulations) stability standards, which is surprising.

However, compared to having four passengers on board with full fuel and water, the 27-passenger load reduces righting arm by about 50 percent, reduces the area under the positive side of the righting arm curve by about 60 percent and reduces the ratio of positive area to negative area — a key measure of dynamic stability — by about 80 percent. If you calculate a transverse shift in CG from the passengers shifting to one side, stability is further reduced — and dramatically.

Unfortunately, there are no passenger capacity regulations for uninspected boats of this size in the United States, as there are in Europe. But in speaking with other designers and operators of similar planing boats, I find that a sort of bell curve of consensus can be developed as to a reasonable passenger load for the Silverton. That consensus is that eight to 12 people is a full load for a boat of this type in sheltered waters. Some feel that 15 would be permissible in calm waters in daylight if half of them are kids, but I think this is excessive because of the complications so many people create in the event of an emergency. Eight adults yield a passenger load of 1,480 pounds, compared to the 3,945 pounds I have speculated to be on the boat that capsized. I would allow, at most, three of those eight up on the flybridge at any given time and just one or two on the bridge if the wind increases — a beam wind alone can substantially diminish righting energy — and seas build.

An owner’s manual for the Silverton 34 from the early 1990s recommends a maximum occupancy of 10 people, or 2,227 pounds. There apparently was no manual for 1984, Kandi Won’s model year.

While stability resulting from weight distribution is much more of a concern than how much total weight will settle a hull in the water, it is helpful to fully understand how boats respond to unusual conditions. With my measurements indicating a waterline length of 27 feet, 9 inches and a maximum chine beam of 9 feet, 6 inches and a coefficient of 0.7 (to account for the narrowing hull at the waterline forward), we get a waterplane of 184 square feet for the Silverton 34. With 1 square foot of seawater 1 inch thick weighing 5.3 pounds, we have a pound-per-inch immersion figure of about 978 pounds, which means the boat would have floated some 4 inches deeper in the water than usual with 27 passengers.


This confirms our intuition about handling since a 4-inch-deeper chine immersion alone makes the rudders less effective, and with its deeper immersion the boat would have felt and handled differently, even if CG stayed at its original height. Given the deck layout and the better view from the foredeck, it seems at least possible that the Silverton was trimmed by the bow, which would have accentuated the problem with handling.

To be clear, the Silverton is by every indication a fine inshore boat. Moderately loaded and judiciously operated, it should offer safe and satisfactory service in coastal waters; it should be capable of resisting upsetting moments from a combination of wave and wake action, a sharp down-sea turn and a beam wind all at the same time. It is not an open-ocean boat, with its high profile and sail area in relation to its underwater cross section, relatively full and flat hull sections forward limiting speed in a seaway, lightweight gasoline engines, lack of compartmentation, small scuppers and so on.


Wakes and numbers

“They were out there for hours before, during and after the fireworks, and there was no incident, no problem; the vessel handled fine,” the boat owner’s attorney, James Mercante, of the Manhattan firm Rubin, Fiorella & Friedman, explained to The New York Times. “You get two or three big boats when they’re together, you get a wake becoming 6 to 8 feet. Any boat could get rolled over with a trough like that.”

With 27 people on board, I find it difficult to believe the boat could have handled well at all. It would have rolled much more deeply and more slowly, with a diminished metacenter (M) as CG moved inexorably toward M as each person stepped off the dock and onto the boat. Anyone experienced with that boat, or any boat in its class, should have noticed something was amiss, in my opinion.

Deeper rolls would have made it more prone to yawing and broaching, and with a heavier hull more deeply immersed, it would have necessarily been less responsive to the wheel and engines. With its small props and shallow gears, it is extremely unlikely it could have gotten much over 12 knots, let alone on plane. If many of those 27 passengers were up forward for a better view, the more deeply immersed bow would have made the boat less responsive to the wheel and more susceptible to broaching. In a broach, a boat rolls and changes course suddenly because of wave or wake action. This event, which becomes more likely as CG rises, could easily lead to a capsize.

Regarding Mercante’s wake hypothesis, I personally have not seen an 8-foot wake — nearly as tall as the Silverton’s flybridge — since I was an OOD on the bridge of an 8,400-ton, 86,000-hp, 32-knot destroyer. However, whatever wakes were actually present in the area that night, the Silverton, prudently loaded and seaworthy, could be expected to have taken them in stride, as did the scores of other boats present. Absent a series of extreme anomalies that likely would have capsized other boats in the area, or perhaps simultaneous (and statistically unlikely) flooding in the bilge that would have reduced stability by free surface effect, I believe the Silverton with a moderate passenger load should have been able to safely ride out any likely combination of wave, wake, course change and shift in on-board weight in those protected waters.


So what lessons are to be learned from this tragedy? The first is that there is no substitute for what the Navy calls using an abundance of caution. Deliberate action is the watchword when voyage planning. Making sound decisions is the crux of safe boating — how many people, whom to take and whom not to take, based on whether you think they would be a liability or an asset in a crisis; what route to use; what time of day and tide to leave or return; whether engine maintenance can wait until afterward. It may be a burden at times, but making good decisions is one of the great satisfactions of boating. It is up to the skipper to acquire the knowledge needed to make these sound decisions.

A good start in building a foundation for decision making would be taking Coast Guard Auxiliary or U.S. Power Squadrons courses, which teach the basics. The next level in understanding how boats work might be taking the Westlawn Institute of Marine Technology’s Elements of Technical Boat Design course. “Chapman’s Piloting and Seamanship” is another excellent resource. You can’t protect against all risk at sea, but you can mitigate it through education and being able to evaluate the suitability of your boat’s design, physical condition and outfitting for a given use.

The boat operator must be able to make sound decisions based on fluid circumstances. Running at night adds a major complication if something goes wrong, and collisions at night with objects of all descriptions are much more likely. Having more people on board not only reduces stability, but it can add enormous complexity and difficulty for the skipper if a problem develops. Managing your risk includes managing your passenger list.

There are times when you ought to be especially selective about not only how many but whom you bring along. Some trips should be made only by experienced boaters with the right instincts and skills and who are in good physical condition. Don’t invite kids or people who are more apt to be a burden in a crisis. Having a dozen 12-year-olds on board who weigh the same as five adults may have the same impact on stability, but keeping them safe will present exponentially more of a challenge in an emergency.

The safety and security of the vessel and passengers ultimately and solely are the responsibility of the skipper, who should always be thinking about the very worst thing that could happen at any moment. No matter the circumstances, the skipper of a Navy ship that runs aground or collides with another ship is almost certain to be relieved of command. There is no evasion of responsibility, and this ground rule is clearly understood when an officer accepts command of a ship. Stack the odds in your favor with experience, education, a little humility and healthy fear, and a well-found boat properly employed. And remember that when at sea, you are your brother’s keeper.


16 comments on “SPECIAL REPORT: Analysis of a capsizing

  1. Ted Guy

    I believe the boat that capsized was an aft cabin model, which would make a big difference in the calculations.

  2. Thad Kubis

    At the end of the day, the Captain or in this case the owner/operator needs to be aware of not only what your insightful and powerful article states, but they must be aware of the situation, the weather, visibility, congestion and yes the safety of all on board.
    The decision to load that many people onto such a small boat is at the heart of this tragic event and the starting point of the sinking, everything else that happened is a direct result of this loading decision. The spin has already started, the real blame in my opinion lies with the owner/operator.

  3. Sail Channel Islands

    Thanks for the excellent analysis. Though you and I believe it should have been obvious that the boat was overloaded, that was not the case. I agree that prospective boaters, particularly boat owners should educate themselves, but the majority are insufficiently trained. It’s time to start the drumbeat for compulsory education, maybe licensing of all boaters.

  4. captbh

    Thank you Mr. Sorenson for an educated, clearly written and non-prejudiced, explanation of what likely occurred. Most intelligent people believed the boat was overloaded since day 1 and would never considered the dock under those conditions. However, your expertise gives credibility to our gut feelings.

    Your Honor, I have no further questions for this witness.

  5. Dan Mielke

    The high vertical center of gravity becomes even more aggravating when it is offset to one side:

    With the same passenger loading, hull and vcg (24” above WL as a guess) and the cg moved off center 3” causing a 5 degree static list (easy to do with such a high percentage of movable weight), the boat’s maximum recovery moment is reduced by one third, 14,200 ft-lbs down to 10,700 ft-lbs. In other words the boat’s ability to handle an incident of any kind has been grossly affected. Also, people are not a free surface like a tank of water but nearly as bad; they tend to fall down to the low side or attempt to run ‘up hill’ to the high side, both of which accelerate a barrel roll condition.

  6. Marc Redshaw

    Why was the stability criteria from 46 CFR (commercial vessels) applied in the article and not the stability criteria from 33 CFR (recreational vessesl)? Because of the vessel’s size?

  7. Bill

    That is a good analysis.

    It might be interesting to actually look at the details viz. number on flybridge versus other locations.

    When you say “meets 46 CFR: could you be more specific? There are many different stability criteria there–depending on vessel type, route, size, proportion etc. I am curious about the specifics.

    Also, you picked a “similar” express cruiser. How similar?

    In my work with stability, I can say unequivocally that the devil is in the details, and the stability is very sensitive to VCG (duh). On a boat this small, a rough estimate is, well, pretty rough. If you can find one of these boats I’d be interested in inclining it.

  8. Big Bill

    Dear Eric, Thank you for taking the time to put forth the “Technical” data that you. You have confirmed a conclusion that we more technically challenged individuals came to using our years of boating experience and common sense. I believe that to put your data to the ultimate test, we should gather 27 lawyers and send them out for a ride on a 34 Silverton.

  9. bill

    to Marc Redshaw:

    46 CFR is not particularly useful for small boats. A portion of the stability criteria there apply to subchapter T (small passenger vessels less than 100 tons (US registered tons, not displacement) with fewer than 149 day / 49 overnight. While these rules are routinely applied to small ferries, they can and do come up short–really short–in adequately dealing with small vessels with a high proportion of their mass as passengers (duckboats, pontoon water taxis, small 40 foot excursion boats…all of which have had fatalities in the past decade, the latter not technically under USCG).

    Practicing naval architects (at least some of us) have noticed this problem for a long time, but there hasn’t been any concerted effort at MSC to devise a completely new set of rules for small passenger boats.

    As for 33 CFR, they technically apply only to private boats under 20 feet in length, and not multihulled. Those stability requirements are also quite difficult to calculate: they were devised through a flooded test criteria. They really don’t address dynamic stability (energy and inertia). Extrapolation of those rules wouldn’t have any scientific basis.

    The Europeans have written rules for private boats; USYRU/USSAILING have also developed standards for stability which are applicable to racing sailboats. The problem with small boats is that their stability is really a function of operator competence more than anything. Larger vessels are more at the mercy of the seas and you can account for that and design to it. Smaller vessels are at the mercy of their helmsmen and their generally overpowerful engines and rudders.

  10. redbw

    Ted: You obviously didn’t see the pictures of the boat when it finally came up from the deep. The media said it was a Silverton 34 motoryacht when they first reported the accident. They said that before they ever saw the boat. They changed their story when the boat came up. It was a 1984 34 convertible. It is the same hull that was used up through the 1988 model year.

  11. whaler23

    It was NOT an aft cabin boat, see below.

  12. Peter Eikenberry

    This is a good analysis. However, I would like to see a analysis of the boats’ stability, with this load, at various heel angles. Inclining experiments are typically done at low angles of heel, usually less than 10 degrees, more often 3-5 degrees. But I would like to know what the righting was at 10, 20, 25, 30, 35 and so on. One of the ships I was on ( a 327 foot Coast Guard Cutter) rolled 47 degrees and came back but very slowly and scared the crap out of the crew. I doubt that this small boat with that load could take even a 30 degree roll. Also, as you said, you need to know the distribution of the weight to make an accurate calculation. How many people were on the flying bridge? That will make a huge difference.

    A commenter asked why you didn’t use 33CFR for calculating passenger load. First off it applies only to small boat under 20 feet. But why? Because it is based on displacement weight, that is, the weight of water displaced when the vessel is fully immersed to the point where water begins to come in. This is the most critical factor in small boats, not stability. As boats get longer their interior volume and the volume of water displaced increases so rapidly that basing carrying capacity on displacement weight would be totally unrealistic because they can carry much more weight than would be safe to carry because of the stability issue. So somewhere between 20 feet ad 26 feet stability becomes the most important factor in load carrying capacity.

  13. Capt Philip Topps, AMS

    Mr Sorensen,
    Thank you for a well researched and concise explanation, along with the very fine diagrams by Dave Gerr, (Westlawn Institute of Marine Technology)
    It is a good thing that there are educated professionals in the marine field who can cut through the nonsense spouted by the likes of James Mercante.
    Having taken the Westlawn Elements of Yacht Design course, the detailed renderings by Dave Gerr reminded me of some of the problems presented to us, vis-a-vis VCG, VCB, metacentric height, etc, and the effects of loading of personnel/cargo on vessels, and the effect(s) on stability.
    Mercante is doing what liars, er, lawyers do. Mr Sorenson, Dave Gerr and the others mentioned are doing what marine professionals do.
    Verdict (from day one) The boat was overloaded and improperly captained. Period.

  14. Eddyboats

    27 people = overloaded=disaster
    It was a 34 convertible not an aft cabin
    Saw it on video

  15. Gregory Gaskell

    I do not understand everyone’s comment that the boat was not an aft cabin. I own a 1984 34C Silverton and the drawings depicted and descriptions in the article are spot on and never allude to an aft cabin. Am I missing something? This is an excellent analysis.

  16. dave

    length x width divide by 15 (when folks weighed 150 lbs) “Chapman”
    34 X 11 = 274 / 15 = 25 persons X .75 for increase fro 150 lbs to 200 = 19 max all aboard must be responsible to remain seated low & evenly distributed. Skippers responsibility for overall safety

Leave a Reply

Your email address will not be published. Required fields are marked *

Comments are moderated and generally will be posted if they are on-topic and not abusive. For more information, please see our Comments Policy.

Vote Today

What's your favorite month on the water?

View Results

Loading ... Loading ...

Login to Trade Only Today

Lost Password