Well well, look what thevYT algorithm dragged in:
4 Likes
From this perspective, not picking sides but the tug operator was much more composed than the pilot.
1 Like
I came around Pt. Wilson in Puget Sound one time on the Aleutian Freighter down by head. Weād come all the way down outside because we couldnāt steer good enough to come down the IP. Tried to steady up at the end of the turn - rudder hard to port and ship took off for a wild ride to stbd and back. Not at all in control for a while. Lucky I had enough room to finally catch it.
In this case the captain of the tug/barge is in control of his vessel but the ship pilot is not. Iād say given the situation he was in he did a good job communicating to the tug. The warnings to get everyone up on the tug is some level-headed thinking.
1 Like
I thought about who was in control and who wasnāt. Have experience with bank shear, and not fun. But I was in the wrong place that contributed to itā¦ Down by the head is problematic, that was not the case here. I guess we can agree to disagree. Pilot put himself in that position being wide on the slight turn. The warnings were level headed thinking, what got them there is another thing. Seeing that coming from the tug guys perspective and knowing he is gonna get the worst of it, I thought he did as well as he could and kept his composure. Lucky no injuries or worse. A more recent event in the Suez regarding bank effect is perhaps fresh on some peoples minds.
Most likely GR down by the head.
I stand corrected on the draft, down by the head is not desirable as you and I both know. Still in the wrong spot. Handled ok outbound from her departure in a very curvy channel upriver. What changed? Speed ordered by the pilot.
By the looks of it, the Voyager complied with the pilotās passing request as much as he could but there was no way out and no happy outcome possible. In that section of the channel, if Voyager had been able to comply with the GR pilotās request to go 90% to the green side and reach the west bank with everything he had, a collision might have been avoided but given the GRās momentum, both vessel would have ended up aground on the west side. Their proximity to green 73 can be seen at 3:42 to 3:46 in the video.
1 Like
Voyager more than likely was already full ahead before this started, if not, didnāt have anything left in his pocket to make a difference. The pushboats I worked early on were always balls to the wall, simply because we didnāt go that fast anyway.
Iām not claiming that the pilot of the GR didnāt make any errors. Sitting at a computer scrolling through the report the errors are obvious.
Iāve made a lot of trips in the Houston Ship Canal and I know that the pilots there sometimes end up working with less safety margin than expected.
What Iām saying is given an unknown hindsight bias I donāt feel like Iād be qualified to make a judgment in this case one way or the other.
1 Like
Well, I finally got around to reading the accident report from the NTSB. My opinion here is based on the report. I have not talked to the pilots involved but have been in many similar situations and close calls.
Once again the NTSB missed an opportunity to go a little deeper into an investigation. Over the years they have conducted many investigations where a ship loses control in situations that are routine for other ships in the same circumstances. The case of the Eagle Otome in Sabine Pass was one. The Orange Sun in NY harbor was another. Probably the Evergiven in Suez as well. There have been many more.
Thousands of ships transit the Houston Ship every year and sea speed across Galveston Bay is routine. Speed could only be a contributing factor - maybe 10% of the blame. This collision was most likely because the ship was built with the bare minimum size rudder designed to keep course in open deep water. It was obviously inadequate for maneuvering of any kind and particularly not for shallow water narrow channels. The full NTSB report actually says the rudder was designed for fuel efficiency which probably means she was also designed to be directionally unstable so as to minimize the amount of rudder necessary to course correct. I have talked to naval architects and read articles confirming the priority of fuel efficiency over maneuvering ability. One architect even admonished a pilot asking about this practice by saying ships spend 99% of their time at sea so they are designed for that environment. Golf clubs spend 99% of their time in a bag in the trunk of a car, but they still have heads on them to hit the ball. The pilots in this incident were actually recorded on the Voice Data Recorder discussing how bad the ship handled in exactly this manner, but the NTSB did not follow up.
The report said the Genesis River met the IMO maneuvering standards but they should have looked deeper into its zig-zag test. I would guess it was not performed on the actual ship but on a simulation. For those who donāt know, a zig-zag test is supposed to be part of the sea trials of the ship. The ship is supposed to be loaded and at full speed. The rudder is put over 20 degrees (for a 20-20 test). When the course has changed by 20 degrees the rudder is shifted to 20 degrees in the opposite direction. When she is 20 degrees on the other side of the initial course the rudder is shifted again. This is repeated several times. The test is how far she continues to turn before checking up each time. A ship that is dynamically unstable will have difficulty in passing this test. Zig-zag results are not posted on the bridge even though they could be useful to the pilot. In Houston we tried to study poorly handling ships by comparing rudder size and looking into the zig-zag tests. The rudder size info was mostly not available although the few we got showed a significant variation and, importantly, a correlation between size of rudder and maneuverability. Zig-zag info was likewise mostly not available. Sometimes the test appeared to have been performed using simulation. I remember one where the overshoot was exactly 19.4 degrees. Every time. There was also one, the old Jo Lonn, where the captain was curious enough to dig up the test. It was a live test with the real ship. Buried in the appendix was this noteā¦ ālost control of ship during testā.
Last time I looked there was no guidance from the vetting agencies on rudder size except for something vague from Det Norske.
Why not look into why ship after ship steers so badly?
Easier to blame the pilot.
The report outlines a scenario in which the intersection of the Bayport channel with the main ship channel played a major role because of the variations in channel width.
I have transited that area many times as a pilot before retirement. There is an effect, but not really significant.
Iād propose an alternate explanation. The previous meetings were in the straight parts of the channel. These meetings only require the ship to turn 4-6 degrees. The collision occurred after the ship made a turn in the channel of 15 degrees and while that turn was accelerated by the passing stern of the inbound ship. That is enough of a turn to make a loaded ship that is directionally unstable very difficult to check up and steady on the new course. You can start the turn with 10 degrees of rudder, but once she begins swinging full rudder in the opposite direction is barely adequate.
This might not be what happened - I didnāt talk to the pilots - but it is more likely than the haunted Bayport channel scenario.
8 Likes
Well said. Having served on a ā Resolution Bayā class container ship in the late 70ās with twin screw and a single centreline rudder I know exactly where youāre coming from.
1 Like
Thanks 
I missed your commentary in the other thread.
The worst of the worst. The SL-7ās were that way and some Disney cruise ship that used to call in Houston were like that. Should be illegal.
That makes sense.
The rudder design being a latent error, revealing itself only when combined in an unexpected way with other factors.
Not really latent in the sense of being a hidden defect. According to the full NTSB report, both pilots remarked on the shipās sluggishness and their conversation about it was recorded on the Voice Data Recorder - before the accident. The NTSB chose to ignore the issue. The question is: why?
ARPA is useless in a narrow channel where the vessels are passing within a few hundred feet of each other and the seascape is littered with beacons, ranges snd buoys.
Another question might be, if they noticed the sluggishness early on, why be on sea speed?
1 Like
My 2 cents - probably not worth that much. A few observations from someone who has spent some time in the Houston Ship Channel.
On speed - I have found very few pilots over the years that understand or appreciate saturation speed. On many occasions we have had to tell pilots that adding additional turns will just make the ship shake a bit more, and you will not get any appreciable increase in speed. At least 90 % of the time - after getting such information - the pilots say thanks - now can I get those turns.
More because of, than instead of - the poor handling ship, I find the pilot(s) at fault here. Being aware of this condition it would be prudent to operate at a slower speed, and have the availability of the increase in rudder power you get when propeller RPMS are greater than ship speed. Seems it would have been prudent to keep some turns up your sleeve.
Also - one huge advantage pilots have over ships meeting at sea is they know the person on the other vessel. Both pilots, aware of the poor handling vessel should have considered and arranged for the best place for them to meet. As above, in one of the few turns in the entire transit was probably not the right place to meet.
I have rode with, and know quite a few Houston Pilots and they are almost to a person great at their job. I would also venture that the 2 pilots on these ships have forgotten more about driving ships up and down this ditch than I will ever know.
But pilots are human - and in looking at this there is a flavor of a lack of appreciation of the risk caused by a poor handling ship.
One of the greatest accident risk, and one we all are susceptible to is complacency. There might have been some of that at work here.
6 Likes
That post seems unnecessarily proscriptive and challenging for a friendly discussion among sailors.
The pilotās PPU in this case is better and it can be discussed.
It is not uncommon to find the shipās position displayed incorrectly on the ECDIS. Even an error of a few meters can be problematic in a narrow channel. Sometimes the ECDIS didnāt even show the ship in the channel.
The Houston pilotās PPUs use a custom made chart. The NOAA charts had several known errors: new docks not shown, mislabeled features, and even showed a phantom dock that didnāt exist. In the trial phase of use they found errors in the channel drawing itself. The Bolivar Roads reach was off by 90ā. It was decided to go with a new chart built from scratch using the best DGPS tech available. Over 250 docks were walked with the equipment and the channel was conscientiously surveyed and checked.The port authority was supportive. Not surprisingly, the resulting chart was orders of magnitude better than the old chart upon which ECDIS is based, some or even most of which was created using techniques like horizontal sextant angles and celestial navigation. This pilot chart is regularly updated as new construction appears. It is so accurate that it is often used it as as a principle source of information when turning ships in a basin (yes, they listen to what the mate, the tugs and the boatmen say as well).
Long before AIS was available the program was written to show where the other ships in the channel were using data from the other PPUs. Now both are displayed. The PPU data is more accurate. Many ECDIC displays do not have the ability to overlay AIS.
The pilots decided that shipās heading was an indispensable feature and the earliest models used a clever antenna arrangement to do that. Gyro input is not always available on shipās ECDIS.
Lastly, the PPU sits at the conning station where the pilot has easy access to it as opposed to being in a remote corner of the bridge as it frequently is.
Neither ECDIS nor the radar was a factor in this accident anyway.
4 Likes
Wheelhouse is a very good product. Itās also helpful that the pilots are incredibly familiar with the software, and the measurements are usually more accurate than than the ships ECDIS.
1 Like