S/V Concordia

(~5 min), but others have been observed to last four or even six times as long (Wilson et al. 1984) LOL "The boat got knocked down, got held down, everyone was too excited to be expected to close deadlights or reduce sail (they only had a day plus warning)...nope, nobody at fault here - In fact, lets give the crew medals and sue the designer, builder and weatherman."
Sorry, you guys are simply on a tack to find the crew innocent when, in fact, in a vast majority of similar cases humans erred. Nobody said that there is no such thing as a microburst, downdraft, rogue wave or... wait a minute...uh,hmm, "a wave more than a hundred feet high". People are simply wont to exaggerate wave height, sea states and such and people that have their *** in a wringer almost killing a bunch of "Semester at Sea" students have more reason to exaggerate. Nice writing, Bistro. I have a hard time doing this without getting irritated.

 

I guess killing each other won't help in learning from the event.
The same goes to blaming anyone; crew or designer or repair crew or God himself.
If one microburst can last for 5 minutes, it seems to me to be enough to keep the boat down until the water through ports takes care of the stability. However I am still reluctant to accept the microburst theory. What would be needed to know is the weather pattern near the event. I am sure the one at helm that time could tell a lot about it. As far as I understand, microbursts occur in the middle of the storm. The other theory says that the knockdown gust would came right in the beginning. Unfortunately it seems we should wait for more informations to figure out (if ever). However I don't recall any mention of a storm already going on in the accounts.

What I cannot understand (among many other things) is the stability. The captain talks about 110 degrees, others about 70 or 90. As I understand the point of (undamaged) vanishing stability should have been even more than 110 degrees. What kept the boat there such that it would neither righten nor turn below? What the fact that it was righted when went down tells us?
In short: were there problems with undamaged stability or not?

 

I doubt there were issues with stability as it was required to be calculated.

This is the big bug-a-boo that exists in almost all regulatory stability requirements. Because it is impossible to determine stability for all possible conditions of loading AND sea surface, regulatory requirements generally specify a condition of zero righting moment and a minimum area under the righting curve...at an assumed water surface under a specified weight and wind loading condition. In actual conditions, with a running seaway, it is possible to get the rig into the water which will change these assumptions. In smaller vessels and/or racing classes it can be requred that the vessel rise from this condition by a test (such as the one used by MORA), but it is also only an approximation that may not provide the margin needed in an actual situation. It is unreasonable to test a large vessel for complete knockdown.

For the stability requirements of a US passanger (school) sailing vessel, see 46 CFR 171.055

http://ecfr.gpoaccess.gov/cgi/t/tex...v8&view=text&node=46:7.0.1.2.6.3.59.3&idno=46

Maybe somebody in Barbados could fill in thier requirements (though I think it would follow UK or SOLAS requirements).

 

Nothing too unusual in a wave 100 feet high surely?

The only trans-Atlantic surface trip I ever took was on a liner, way back when they still did that sort of thing. We spent half the trip in a serious storm that incapacitated most of the passengers and the crew had to turn off the stabilisers and bolted down the furniture in the public areas that were kept open. The enclosed promenades on the main deck were strictly off-limits to crew and passengers as the main deck, which was about 40 feet above the surface was awash frequently. In fact no passengers were allowed anywhere outside at all for three days.

I discovered that I am apparently immune to seasickness. On a bet I had the kippers at breakfast: the guy who lost the bet was driven off by the smell!

Once the storm hit there was little to do so I spent most of my time in the bar, well above the main deck. I remember hearing the screws pounding away every time their blades lifted above the water. The waves were breaking frequently with their tops being ripped off by the wind and the sea looked more white than anything else with long streaks. The log described it as a whole gale; I don't know how big a deal that is for a mariner but it had my attention. I doubt any of those waves were actually 100 feet but many were at least 50 feet trough to peak as I was definitely looking up at the peaks. I am quite sure there have been worse storms so I believe 100 feet waves.

I recall lining up to debark; a dozen or more passengers had casts, and there were hundreds of broken deck chairs piled head high and more.

 

I have just done some work on Group Finot designed and French government approved "38 ft Reve de Antilles." Given the warped priorities and incredibly flimsy skeg and rudder arrangement, while putting huge amounts of useless steel and time elsewhere, I don't have much faith in the ability of government rules and regulations to get things right.

 

It is up to the Naval Architect to get it right.

Regulatory agencies merely establish minimum required standards.

Stability requirements are based on meeting specified criteria which indicate the overall stability of the vessel (in intact or damaged condition). They are NOT meant to indicate the vessels ability to actually withstand the conditions indicated in the criteria. That is why wind speed is dictated, but no consideration is give to the accompanying seas and their affect on the vessel.

 

TSB Canada findings are released

http://www.tsb.gc.ca/eng/medias-media/majeures-major/marine/m10f0003/MI-M10F0003.asp

 

Interesting read and very complete.

Summary of major points:

- The event was NOT a microburst downdraft according to the postmortem
- Poor knowledge of autopilot / fly by wire controls and the rudder movement limits under the electronic system led to insufficient response to gust event. A manual ship's wheel with mechanical (not electronic) interconnect would have allowed more positive control (my opinion)
- Leaving companionways and vents open on the boat altered the stability curve and knockdown recovery of the boat dramatically. Displayed stability curves were significantly different in the knockdown case comparing "battened hatches" versus open door choice in use at the time
- No clear policies regarding watertight doors / companionway closings and vents in escalating wind conditions
- Crew/students were trained and equipped for expected problems, handled event well
- Somewhat onboard haphazard communications policies, storage locations and records keeping (Iridium phone, EPIRBs etc.)
- Land based management had some deficiencies in regards to policies, records maintenance and regular testing/review of procedures and equipment
- Crew familiarity with stability curves and application to wind conditions could have been better
- Brazil screwed the pooch big time in regards to internationally accepted maritime disaster response, as well as questionable naval decision making on transferring rescued people to a naval vessel (presumably for photo ops and optics)

At least that's my read. Thanks jehardiman for keeping and eye out on this. The CBC documentary was also very good - I watched it a couple times. It is available on line on the CBC site.

It was also interesting to re-read the whole thread with the factual conclusions in hand. marshmat, jehardiman, bistros (who?) and Brent Swain posts were pretty much dead on the money.

--
CutOnce

 

Setting the righting curve issues aside, it looks to me like a typical operator disconneced from the situation by technology problem. I have noticed a large increase in this type of accident, both in my work and in vessel casualities in general, over the past decade or so. With more and more of the mundane decision burden shifted to the technology (in this case, like several other casualities, steering command), often the operators become disconnected from the situation and end up "behind the curve" when they need to make a decision. The 2/O appears that he knew what he should do, but selected to use the autopilot first rather than directly intervene, effectively makeing the decision too late.

The righting area issue has been around for a while (i.e. the slope and area requirements vice a demonstratable stability requirement), and may need to get fixed, but I wonder if a good helmsman would have made this a non-issue.

 

I agree with this assessment.

I have designed and built embedded control systems for mechanical and electronic devices, as well as having done a lot of integration work gluing together functional systems from assorted components. Control systems are generally great at doing what they are designed to do, but asking them to handle situations outside their design specifications (or the imagination of the people doing the engineering) is dangerous.

Three fast hits on the port button could have been interpreted as a problem event and the system could have been designed with better options than three successive one degree rudder changes - but that is academic now. People trust electronic systems far more than they should just because they work fine in normal design conditions.

jehardiman is right from one perspective here - the Second Officer was behind the event in terms of response due to excessive faith placed in the autopilot's control. On one hand, the autopilot is probably polling multiple input sensors hundreds of times a second, but I guarantee it was de-tuned, input normalized and de-bounced to reduce excessive reaction and movement of the rudder.

From a different perspective, the Second Officer probably reacted just as fast hitting the "port" button as he would have cranked a wheel - but the response to his course alteration wasn't as direct. With a wheel and a mechanical rudder position repeater you can see exactly how fast and how much the rudder is reacting - with an autopilot, it most likely shows you what it thinks you asked it to do, not what you actually got. With a wheel, you can speed up your spin rate if things aren't moving fast enough - but you can't speed up software reaction time unless it is programmed to see how fast you hit the button, or how hard you are loading the joystick. If the electronic control of the rudder "downgraded" the Second Officer's fast reaction to slow response, the failure then points to poor education on usage of the piloting system, and poor system design that made it inferior to the manual system it was to replace. I wonder if the Second Officer ever was told he had to be further ahead of the electronic rudder control than a manual system?

The report was very carefully written to outline areas of improvement, but they also (intentionally in my opinion) did not condemn anyone for direct responsibility for the failure.

The collision of digital versus analog systems happens in many arenas. The binary nature of digital information leads to very subjective interpretation of data inputs - simulating gray areas from black and white input. As I do system design work every day these situations always help me think critically about the things I do - and how they can fail in ways I don't anticipate.

--
CutOnce