Critical review of lessons from US warship collisions

The US Navy’s report on recent collisions of USS Fitzgerald and USS John S. McCain fails to see the true shortcomings of seamanship and procedures.

We now have the official report regarding the collisions that led to the deaths of 17 sailors on the USS Fitzgerald and USS John S. McCain earlier this year. I have read the report, and some of the conclusions were so puzzling to me that I decided to make a critical review to come up with my own conclusions. I have done this on the basis of the report itself and this article in Defense News.

I have earlier been critical about the US Navy’s procedures and level of competence in the field of navigation and seamanship. The report about USS Fitzgerald and USS John S. McCain only further deepens my distrust in the US Navy’s abilities. The performance of the crews on the two ships was appalling but the conclusions derived by the Department of the Navy are equally poor and raise serious questions about the ability of the Navy to learn from the incidents.

A few words about my own background for making these comments are appropriate before I get started. Navigation and seamanship is my field of expertise, and I have extensive experience in navigating large and small vessels in waters that are much more confined and challenging than the ones that the two collisions happened in. I have also been a professional teacher of navigation and the captain of a training ship. I do not, however, have personal experience from the US Navy, and I have never been to Japan or the Strait of Malacca, so it would be wrong to consider me a local expert on either subject. My comments are made from the perspective of a European naval officer with expert knowledge of navigation and an incapacity to comprehend the procedures of the US Navy.1

Collision between USS Fitzgerald and containership ACX Crystal

I’ll go through some of the aspects of the collision, starting with an overall assessment of the situation and then moving on to the specific conclusions that the report finds.

Reading the traffic situation

The accident happened at night in an area with a fair amount of shipping. The area was at least 10 nautical miles wide, so there was plenty of room to maneuver, and there really wasn’t any point in getting close to anybody. The accident happened as illustrated below:

USS Fitzgerald collision

It must be presumed that the only interesting ships in the picture are the Fitzgerald, Wan Hai, Crystal and Maersk Evora. My reading of the plot is that there are in fact three separate situations that are worth noticing.

Crystal follows a rather strange track, first passing the traffic separation at a small angle and then turning before entering the appropriate lane. At 0115 there was a close situation between Crystal and Maersk Evora, where Crystal should give way, but didn’t. Crystal may have followed the poorly planned route rigidly without noticing the rapidly approaching containership from the South. Is seems that Maersk Evora reacted appropriately to avoid collision by making a hard starboard turn. By observing this incident, the Officer of the Deck on Fitzgerald should have been warned to keep an extra safe distance to Crystal.

At about 0120 Fitzgerald could have made a starboard turn to pass astern of Wan Hai and Crystal. Instead the warship continued its course which creatted a critical situation when Fitzgerald crossed the bow of Wan Hai. It is difficult to gauge the distances on the illustration, but Wan Hai has probably estimated that the warship planned to squeeze in between Wan Hai and Crystal. This meant that Wan Hai was caught in an uncomfortable position where she couldn’t turn because the warship had to be expected to make a starboard turn at some point and pass closely on either the port quarter or the starboard bow.

And then, finally, there was the critical situation between Crystal and Fitzgerald which led to the collision. As the above mentioned situations had shown, neither ship seemed to have a reliable navigation team on watch, so at that point the collision was all but inevitable.

Below I will go through some of the conclusions from the report about the causes of the collision. Some of the conclusions are fine, but many of them are totally wrong, so this will as much be a critique of the report as of the actions by the crew.

High speed in a high-traffic area

The report finds that Fitzgerald was operating at a speed that was too high considering the number of other ships in the area. It is mentioned that safe speed is required by the International Rules of the Nautical Road.

As mentioned above, I will seriously question the assumption that the collision happened in a particularly difficult area to navigate. There was plenty of room for maneuvers, and the traffic density was average. Fitzgerald transited at a speed of 20 knots which seems reasonable. Judging by the illustration above, the Maersk Evora transited at a higher speed than Fitzgerald. Also, it is not necessarily more difficult to have a higher speed, it only means that you have to look a little further ahead and make your maneuvers earlier.

Fitzgerald should have turned

The report finds that Fitzgerald should have turned when the risk was present. Doh!

Poor use of radar

Apparently the radar operators on Fitzgerald did not systematically use the radar, and they were unable to make it work properly. Here’s the thing: Why do American warships even have a dedicated radar operator?

It is not very hard to tune a radar and plot the ships that are interesting. With a bit of echo trails it is easy to see which ships are relevant, and a skilled operator can get a tremendous amount of information from the radar screen by looking at it for just a few seconds. The last thing you want is an – apparently unskilled – assistant as a filter between the Officer of the Deck and his most important tool for situational awareness. It would be much better for the US Navy to teach the Officer of the Deck to operate the radar himself, just like everybody else does it.

No use of AIS

This is a ridiculous negligence. I think that many warships are too wary to transmit their own data on AIS, and that it would greatly increase the traffic safety if they used it more, but at least I can appreciate that there are situations where you don’t want to use the active mode on AIS. However, there is simply no excuse for not using the passive function of AIS to get a fast and reliable picture of the shipping around you.

Accounts from the timeline of the other collision with USS John S. McCain indicate that the AIS display on American warships lacks a user friendly design and can get “cluttered and useless”. The US Navy needs to get a better AIS integration which will show AIS targets directly on the radar screen or on an electronic nautical chart.

Lookouts only looked out on the port side

Ok, you need to look both ways. But although Crystal hit on the starboard side, the aspect from Fitzgerald would have been only a few (perhaps 20) degrees on the starboard side, so she would have been visible from the port side of the bridge through the front window. So fundamentally the lookout just has to look out.

Did not follow traffic separation schemes

The report finds that Fitzgerald did not follow the approved traffic separation schemes (TSS) in the area and was unaware of their existence. I spent more than an hour trying to figure out what TSS the report had in mind, because the official nautical charts of the area do not display such a scheme. It turns out that the Japanese Captains’ Association has established voluntary traffic separation schemes in various parts of Japan that ships are encouraged to follow. Since these schemes are not official, they are not displayed in the nautical charts, but information about them can be found in different pilot books.

Mikomoto Shima TSS

Should a ship of the Pacific Fleet be aware of such a scheme? Sure. You read the pilot books as part of your route planning. But seriously, what the heck is going on for Japan? Either you have a TSS, or you don’t.

The report also notes that:

By 0100, FITZGERALD approached three merchant vessels from its starboard, or right side, forward. These vessels were eastbound through the Mikomoto Shima Vessel Traffic Separation Scheme.

This is not what I see on the plot of the situation. Wan Hai was North of the (informal and voluntary) TSS, Maersk Evora was South of it, and Chrystal was generally going East in the Westbound lane. Fitzgerald was crossing the end point of the TSS in what appears to be a safe distance. It is reasonable to point out that the crew on Fitzgerald should have known about the existence of the voluntary Mikomoto Shima TSS, but it is hard to see that this point had anything to do with the collision, and the other traffic in the area did not seem to pay attention to the informally established TSS either.

No attempt to contact other ships by radio

The US Navy has a terrible tradition of calling everybody on Bridge to Bridge radio to make arrangements regarding the passage. While this procedure may give the impression of good seamanship, it is fundamentally a bad habit because it diverts attention from the job at hand. Contrary to popular belief, radio communications are not a quick solution, and often the consequence is merely that important decisions are delayed by 2-3 minutes.

A skilled navigator is able to combine the information from visual, radar, and AIS sources with a knowledge of the International Rules of the Nautical Road and an understanding of good seamanship. Based on that information alone, good decisions can be made, and communication between the involved ships is conducted nonverbally by clear maneuvers that are much less prone to ambiguous interpretations than radio communications.

It is obvious that radio communications could have been a good choice in the final desperate minutes before the collision between Fitzgerald and Crystal. But it is unlikely that a general compliance with the tradition of the US Navy to call everybody tirelessly would have done anything good for the shipping in the area.

Did not sound the horn

The Officer of the Deck did not sound the horn before the collision, whereby the crews on both ships would have been warned about the dangerous situation. It is possible that sounding the horn could have prepared crew members before the collision, but it is unlikely that it would have prevented the collision itself. Crystal could hardly have made any maneuvers at this point in time to avoid the accident, and on board Fitzgerald the disastrous (lack of) decisions had already been made.

Combat Information Center personnel did not comprehend the navigational situation and communicate information

It is pointed out that “[k]ey supervisors in the Combat Information Center failed to comprehend the complexity of the operating environment and the number of commercial vessels in the area.”

Apparently American procedures dictate that personnel in the CIC should interfere with the safe navigation of the ship. This sounds like a terrible division of labor and a recipe for enormous amounts of internal communications between the CIC and the bridge. All the necessary instruments are available on the bridge, and there are probably way too many people involved in the task already. Getting verbal messages from the CIC is not going to be helpful in the interpretation of the situation on the bridge. The Combat Information Center should deal with information regarding combat and leave it to the bridge to manage the navigation.

The Commanding Officer was not summoned

Obviously the Officer of the Deck on Fitzgerald was not competent to the job, so it would have been good to have a senior officer on the bridge. But generally the American procedure of having to call the Commanding Officer for minuscule reasons is perhaps the epitome of what is wrong with the Navy Way.

What you need is competent navigators that you can trust to run the show. If they have to call the CO every time there is a decision to make, they cannot grow the self-reliance that you are so dependent on when things become critical.

The report mentions several incidents before the collision where other ships were within a range of three nautical miles without a report being made to the Commanding Officer. Come on! One ship within three nautical miles does not under any circumstances qualify as a complicated situation. If you can’t trust your navigator to manage that, then he shouldn’t be trusted for watch duty at all.

You should teach your officers to be good navigators who trust their own judgment; you should not teach your officers to be insecure navigators who are afraid to act. The Officers of the Deck are the Commanding Officers of the future, and they have to learn to bear the responsibility. Otherwise they will not be prepared the day that they have nobody to ask, and everybody is looking to them for answers.

Collision between USS John S. McCain and oil tanker Alnic MC

The collision between John S. McCain and the oil tanker Alnic MC was more simple than the incident with Fitzgerald. While passing through the Eastern approaches to the Strait of Malacca, an operator’s error caused the steering control to be unintentionally transferred to another steering station, causing the bridge crew to believe that they had a technical problem while the ship was doing a slow port turn. In reality they could have just transferred the steering control back to the helmsman, but nobody had a sufficient understanding of the buttons to figure it out. To save the situation they tried to reduce the speed but inadvertently did so only on the port shaft, causing the port turn to speed up.

Lack of training

It is beyond comprehension that anybody can do duty on the bridge of a ship without knowing what the buttons do, or how the steering and maneuvering systems work. It is even stranger that there can be an entire team and nobody knows it. This is a systemically rooted incompetence problem that the US Navy should address unhesitatingly.

Speed commands were pedantic

The accidental move of steering controls happened because the crew tried to move the speed controls from the helmsman to another station called the Lee Helm. This was done because the helmsman was struggling to maintain the course while simultaneously adjusting the throttles. The CO therefore decided to separate the two things so one person would operate the rudder and another the throttle handles.

The problem is not that the CO wanted to help the helmsman or that he ordered a configuration that had not been planned ahead, as the report indicates. This is well within the boundaries of reasonable adjustments that a sensible plan must allow for and that a competent crew would have managed. The problem was that anybody was juggling the throttles in the first place.

The propulsion system on USS John S. McCain is constructed such that two things determine the effect of the propeller. There is the propeller speed given as a number of revolutions per minute (RPM) and there is the angle of the propeller blades given as a percentage of maximum pitch. This gives very good maneuverability and makes it possible to change speeds quickly. If you set the propeller RPM and pitch at a given value, the ship’s speed will fluctuate somewhat because external factors like swell, wind, and water depth influence the transformation of propeller effect into speed though the water. In addition to that, the current means that speed through the water can be higher or lower that the speed over the ground.

From my reading of the report I get the impression that speed commands were given as a set number of knots on the speed log which shows the ship’s speed through the water. In order to maintain that speed, constant adjustments were made on the handles, so if the ordered speed was 18 knots and the speed log said 18.5 knots they would reduce the throttle, and if it said 17.5 knots they would increase the throttle.

In addition to that, it seems that the navigator would perform constant calculations of the ship’s speed over the ground and suggest changes to the ordered speed through the water to make sure that a precise time schedule was followed. Speed commands were therefore changed often and by small increments, which increased the number of adjustments that the helmsman had to perform.

This part of the report illustrates the process:

Approx. 0457 JOHN S MCCAIN increased speed to 17 knots.

0459 JOHN S MCCAIN reduced speed to 16 knots.

0500 Reveille was announced to wake the crew for entering port. The Navigator informed the OOD that previous course changes to the North to avoid surface traffic had put JOHN S MCCAIN behind on its intended track and timeline and recommended an increase in speed to make 18 knots.

That’s a ridiculous way to operate a ship’s propulsion system and a pedantic approach time-space calculations. The competent way to manage a timeline is to add a buffer and to reduce speed at an appropriate time when you are sure that you can make your destination according to schedule. If you get behind schedule because you have to avoid surface traffic, it means that you are managing your timeline wrong. Speed commands should be given as a particular RPM and pitch so the helmsman doesn’t have to think about it, not as a speed in knots that requires fiddling.

No autopilot

Why on earth does an American destroyer not have an autopilot?
If they did, perhaps the helmsman would have time to think about how the buttons on the panel work.

Current does not change a ship’s course

This is a personal hobbyhorse of mine that doesn’t explain the collision but shows that the Department of the Navy does not understand the forces at play. Current does not turn a ship!

The report several times explains that the helmsman had to hold 1-4 degrees to starboard to maintain the course because the current would turn the ship. Many forces could explain the ship’s left turn, but current is not one of them. The current effects the entire mass of water around the ship whereby the stern and the bow will be moved in the same direction at the same speed.2

Wind is a much more plausible explanation to the experienced tendency of the ship to turn left.

Summary

Both collisions can be explained as a matter of incompetence on the part of the involved crews. In both cases the people on duty did not know what they were doing, so they managed to turn a routine situation into a disaster.

The crews did not follow standard US Navy procedures in their operations, but these procedures themselves show that incompetence is an institutionalized problem. The number of superfluous workflows required by US Navy procedures and the division of labor between an excessive number of assistants are counterproductive to the safe operations of the ship. What you need is one competent person to run the show on the bridge with the help of a few assistants. The US Navy approach is to spread the responsibility across many people with poor qualifications.

The Department of the Navy does not seem able to draw the relevant conclusions in their report about the accidents, because they focus solely on the mistakes made by the individuals involved. No attention is given to the faulty education system or the cumbersome procedures that allowed a culture of incompetence to develop.


  1. For a similar critique of the Navy Way from an American perspective, see Capt. Paul Lobo’s comments in The Maritime Executive
  2. Current can turn a ship if there are different currents on the stern and the bow, which can sometimes occur when navigating very close to shore (like within a hundred yards from a bank). It only happens for a short moment until the ship has passed though the boundary between the two currents. This is not what USS John S. McCain was experiencing. 

Leave a Reply