I won’t defend the “tone” of some people’s posts but the debate is something we are not going to get anywhere else. For those of us who command and those of us who aspire to command, these discussions about casualties are helpful to form our own game plans and learn from others mistakes. It’s my single favorite aspect of this forum.
6 Likes
Using APRA or AIS for avoiding vessel in close proximity is a technical error. As is over-reliance on GPS or ECDIS in restricted waters.
Over-reliance on weather routing software is essentially the same error. On the El Faro the captain failed to cross-check weather information. I have not claimed here that it was a cause or even a factor. Just pointing out that it happened and it’s poor practice.
It’s been enough time since the tragedy that I don’t think there are too many grieving families perusing these forums. And if there are and they see this discussion, they would understand. They don’t want this sort of thing happening again, I’m certain of that.
1 Like
You make good points. The Hindenburg is an appropriate example. Losing the engine without the hurricane - or other proximate danger - would not have been fatal.
That said… why do we build multimillion dollar commercial ships carrying scarily dangerous, incalculably polluting and/or enormously costly payloads to operate in situations where there is high traffic density, risk of life, and millions in property damage… with only one engine?
The immediate answer is - it would cost more. Jack Devanney in the book, The Tankship Tromedy, addresses the issue much more competently than I ever could. I found his case for requiring two engines on commercial ships and the economics of it to be compelling. As someone who captained single screw vessels and twin screw vessels before becoming a pilot in a busy port and getting wide experience on a great variety of ships I can say without any doubt that twin screw vessels are safer to maneuver (as long as they’re equipped with twin rudders. A single rudder between two propellers is a disaster). Since reading Devanney’s book I have come to look at maritime disasters with that issue in mind… would a twin screw ship have suffered the same fate? In the case of El Faro, probably not. It doesn’t take full power to heave to head up to the storm. No power puts you in the trough within minutes.
You talk about the relationship between safety and “a sufficiently hardened target”. Maybe it’s time to take the next step. The shipowners will make a little less money, but not much and the ships will be infinitely safer.
Concordia/Stena did just that with their P-MAX tankers. Twin engines, rudders and machinery for extra safety, shallower draft on a wider hull, and with more cargo volume. How successful they’ve been I can’t say but they are impressive ships. But for all I know you’ve likely piloted them so you probably know better than I.
Problem is, if only one company does it then they are operating at a competitive disadvantage. Even if it’s only a slight one. It has been more than 10 years since I read Devanney’s book (available as a .pdf online, btw) and my recall of the details is not there. I remember he painted his tankers white to better spot leaks and rust. Crews weren’t particularly happy about that, but he liked it as a management tool. For pilots it was a big advantage in the days before tractor tugs to have a ship that could slow down and back without running off to port or starboard and scaring the crap out of everyone. Twin screws does that.
Could they get a break from the P&I club to even things up a bit?
It appears to me that there was over-reliance on the BVS system which data was at least 12 hours old. I’ve used a few different weather programs (last being SPOS) and they were excellent programs, but these programs were based on weather data collected by NWS, NOAA, & National Hurricane Center in the US and several overseas (UK, Norwegian, Danish, etc) weather services. Hurricane track predictions are based on input from several sources with the most likely track being the most popularly predicted track by the several services. The software programs were only as reliable as their input (garbage in / garbage out). It is essential that the data inputted is accurate and CURRENT. As there is a lag time between the weather services reports and the receipt of the weather “file” it is critical that weather routing decisions are based on the most accurate and current information. That sounds pretty logical. My company contracted to have a SPOS file delivered to my ship every 12 hrs, BUT I always had the option to request it more frequently (which I did at times) based on anticipated heavy weather. As good as all of this is, It doesn’t replace the good seamanship practice of checking “other” sources. Those being what you can receive as the most current weather reports off your Sat C / Navtex / GMDSS. Additionally, those reports are the basis for the data inputted into your weather software “file” that you may receive several hours later! The telex weather that you would receive years ago (before all of the software programs) you would then plot, constructing a weather chart & your ships position, verify the weather (wind & seas) as being what you should experience based on your position from a hurricane or low (or high for that matter), consult Bowditch if you must, to satisfy yourself that your routing is having it’s desired effect. Not checking the most up to date weather information is just poor seamanship, plain & simple! No where have I heard or read the suggestion of the El Faro using a weather routing service. Now I know she was on the “coastwise” trade, but perhaps that should be considered particularly if the ship managers aren’t going to use a lot of oversight regarding heavy weather avoidance routing. Many shipping companies use weather routing services, or as I liked to call it “Fuel Savings Services” Weather routing in my opinion is more about saving fuel, but avoiding heavy weather is a key component to saving fuel, so it does increase safety. Regarding the loss of propulsion, I disagree that this can’t be anticipated. If you’re experiencing heavy seas, machinery just doesn’t operate the same, period. Fuel purifiers and strainers / Lube Oil strainers get clogged and you’re going to have problems. In heavy seas, you must be prepared for the worst case scenario to happen, loss of propulsion, loss of aux power (which in turn causes loss of main engine, fuel pumps, ballast/bilge pumps, etc.). Emergency generator (if unaffected by the heavy weather) can only keep your steering, Nav & E lights on. Then you have old/rusty d-rings & securing buttons securing cargo (which hopefully prevents cargo from busting loose), water tight doors, that may be not so watertight, and finally huge sideports, all this with the ship rolling 15-45 degrees, it’s not a pretty picture. It’s very tragic and should have never happened. With all of the shipping companies concerned with “costs”, perhaps a weather routing service (contracted strictly to save fuel $$$) would have helped save the El Faro & crew. May they rest in peace!
I’m not that big a fan of the Tankship Tromedy. I have read parts of it (had issues finding the free pdf from the old link to refresh myself on it) and the premise and there are certainly features of the governance over ships where he’s spot on. Despite this, his ‘impending doom’ assertions seem alarmist and (13 years after publication) largely unsupported. The world is supposed to be awash with failed tankers spilling oil due to unholy bargains between owners, operators and governments and class societies. Yet, the world is shipping more oil products than ever and significant spills from tankers are continuing downward trends. The bigger the package the worse the possible impact, sure, but as a trend… well, here’s some data.
ITOPF data and graphs:
https://www.itopf.org/fileadmin/processed/e/4/csm_seaborne18l_4ec9df16fd.jpg
So why—if his observations seem accurate— doesn’t the data seem to support it? Variables not factored is one reason. If the data is correct, the world is getting more cautious. Diligence more diligent. Insurers, port state control and attention to high risks ships like tankers more stringent perhaps despite the alleged collusion of owners and flag state and class? Possibly increased standards in international rules, surveyor and flag state oversight or perhaps also regional planning for response to loss of propulsion events involving tankers due to increased regulation… any number of reasons present themselves and all of them have impacts. Not to mention - spilled oil is spilled product, less profit.
I especially like the idea that port state control impacts have impact as a possible factor and this has a relation in my mind to the EL FARO and the old age Jones Act domestic fleet. International trade vessels have to deal with routine, fickle, inconsistent yet constant port state control as they trade between multiple countries, in addition to class surveys. This subjects vessels to increasing (unsympathetic) independent oversight in between survey windows. Compare to the old Jones Act fleet. Not that many trade international. So they only get CG and class surveys - both mostly scheduled at owners convenience.
1 Like
The equilibrium point with regards to cost/safety for each individual company in a particular sector is going to be roughly within the same range. If the cost of an incident increases significantly then the actors in that sector, regulators, insurers, owners, all face incentives to lower the risks and will be willing to accept increased costs to avoid an incident.
The costs of a tanker incident have been made much higher than is the case with a Jones Act dry cargo ship.
Even faced with increased incentives to lower risks the tanker sector remains mostly single screw. Which tells us the market does not consider twin screw cost effective in increasing safety.
If I may contribute my practice as a recreational mariner?
I have the luxury of being able to not put to sea. So I really want to get the weather right as far as possible as I don’t want to be caught out, nor miss a lovely sail. When planning a weather dependant event, I begin looking at the GFS forecast as soon as it is available for the specified date, usually about 14 days in advance… Obviously a 14 day forecast is always going to be suspect. But I look at it the next day (13 days) and again (12 days) etc all the way down the calendar. The object is to see if the pattern is staying the same or changing every day.
There are times when a weather system establishes and it is stable for days/weeks. Then others when the pattern is extremely variable day by day. (I am based in the UK where we have weather not climate). I will also look at the rain radar and compare it with the forecast. Obviously hurricanes are extremely unpredictable. But even so if the updating pattern is staying steady, one can build more confidence in the forecast. If every update is different… I know I can’t rely on the forecast being accurate until afterwards:fearful:
If I wanted to improve my technique, I would copy the methods used by ocean sailing racers who do mad things like the Vendee. They win and loose races on the quality of their forecasting and weather routing.
I just finished a heavy weather avoidance course at MITAGS for the Indian Ocean. They have 2-day seminars for the Atlantic, Pacific and Indian Ocean separately. One thing that is emphasized is that different models have different “spins” on what the weather will do. The commercial weather services such as BVS or SPOS may well put their own criteria into their model which may change the data somewhat. They do not appear to do this when dealing with tropical systems, but there are multiple models at play there. as well. There is frequently a graphic put out by NHC/JTWC that discusses the models and shows where/how they diverge. Having been in OPC for a video conference between OPC and NHC regarding a hurricane, it was a subjective call between the forecasters as to which model was relied upon. Granted, the call was based on objective evidence such as prior position, satellite imagery and ship obs, but it ultimately came down to their opinions/consensus.
3 Likes
Yes, that is a prudent behavior.
However, this will not work with a hurricane in the neighborhood.
This thinking was part of El Faro’s fatal behavior: The hurricane will shortly return to the initially forecasted path… it did not.
Joaquin’s path, a former extra-tropical depression, was difficult to predict; only one service, the ECMWF, had it nearly right over days.
The Navtex/Sat C was always right with the real position of Joaquin, near the transmission time.
A simple comparison of the actual Navtex position, with the outdated, forecasted position of BVS, for the same time, would have helped a lot…
This is correct, in the U.S. Applied Weather Technology (AWT) uses NHC (National Hurricane Center) Tropical Cyclone (TC) current position and NHC TC forecast positions and tracks to use in their BVS software. AWT does not do Tropical Cyclone forecasts, they strictly use NHC.
AWT does however use their own models to generate the wind and sea height contours which appear as various colors on the computer based on position and forecast data they receive from NHC.
There are not various tropical cyclone forecast models in play here, only one, the NHC.
That and comparing actual weather conditions to the expected conditions.
To head into the wind to reduce list the captain steered a course around 050 degrees true or about NE. A NE’ly wind would put Joaquin at about where the red symbol is, the actual position of the TC.
If Joaquin was at position of the blue symbol winds at the ship’s position would have been SW’ly. If Joaquin was where the captain thought it was the wind would have been blowing about 180 degrees different, or opposite from actual wind direction.
2 Likes
Please follow this link (NHC Track and Intensity Models) to an explanation on the NHC website of the models they use.
Yes, I should have worded it differently. From the point of view of the position and forecast track of Joaquin received on the El Faro a discussion of which models were used for the forecast is irrelevant.
That is with regards to BVS vs NHC.
The green El Faro track shows the course change from SSE to ESE (116°) at 01:30, after:
23:02 - BVS’ mail to captain, to download the last forecast
23:05 - 3M calls the captain for new instructions
23:16 – Navtex shows Joaquin’s center at bearing 111° from El Faro, proceeding at 220°
01:20 - 2M calls the captain to explain the Navtex forecast with the planned course change.
She proposes to go South and to escape through the Crooked Island Channel
If the captain had downloaded the BVS forecast from 23:02, he would have seen that BVS forecasted, for the first time, Joaquin going South of San Salvador Islands.
Times from my timeline >>> https://forum.gcaptain.com/t/el-faro-an-overview/47034?u=urs
1 Like
At 30/1747 the captain download the BVS file which was NHC Advisory #11
On 30/2305 the third mate rcvd NHC Advisory #13 on the SAT-C and called the capt.
The 23:02 BVS email contained NHC Advisory #12 but was not opened until 01/0445
Here is the NHC forecast for 01/1200Z in each case.
This is from the Tote report:
In other words, when a given Advisory and BVS download were available on the vessel (which became available roughly at the same times each day – approximately 0500, 1100, 1700, and 2300), the Advisory was the more current forecast and the BVS presentation was based on the Advisory from six hours earlier.
The ECMWF model got Joaquin’s track correct because it got the intensity correct. As the model intensified the storm, it reached higher into the atmosphere and therefore had more atmospheric layers steering the system. This is how the ECMWF got the track correct. The NHC discussion did include the large uncertainty they had in the forecast track. You can now get all hurricane discussions by ftp, something that was not available to the El Faro. I hope that more mariners read these discussions so they are aware of how confident or uncertain the NHC is about their forecast.
The location given by the Navtex is only the NHC ‘best guess’ on where the center of the storm is based on all available information. It can be adjusted by several miles as they go through additional data to see how well they did on the forecast track and intensity. If your margin of error is only a few miles, I think you’d be in trouble already, but you might not want to take any position or radius of maximum winds as gospel, it’s only based on available information and sometimes data points are few and far between.
Of course it would have been helpful if the El Faro had a working anemometer. They did pay close attention to the one weather instrument they had, the barometer.
2 Likes