Viewpoint: Unseaworthy Vessels One Reason Containers are Lost at Sea

The World Shipping Council reported that for 2020-2021, the two-year average annual loss of containers overboard was 3,113 units. That is more than four times the total for the previous three-year, 2017-2019, average loss of 719 units.

Two defects contribute to container stack collapses that can cause cargo to be lost at sea: First, containers sometimes are loaded contrary to the Container Securing Manual, such as stacking heavy containers over lighter ones and stacking containers too high. Secondly, container stacks are not always secured so they can withstand the most severe weather conditions expected on the voyage.

The International Maritime Organization’s Code of Safe Practice for Cargo Stowage and Securing (the IMO Code) has a number of general principles. Two of those are relevant to the defects mentioned above:

1. Personnel planning and supervising the stowing and securing of cargo should have a sound practical knowledge of the application and content of the Cargo Securing Manual.
2. Decisions taken for measures of stowage and securing cargo should be based on the most severe weather conditions which may be expected by experience for the intended voyage.

Regarding the first point, shore planners who prepare container loading plans do not possess sound practical knowledge of the application and content of the CSM; this is noted below by the P&I Club GARD reports by safety investigators following container stack collapses, also mention containers loaded contrary to the CSM.

As for the second point, if container securing systems could withstand the most severe weather conditions expected on voyages, then, provided containers were loaded in compliance with the CSM, container stacks should not collapse.

The accepted practice is that container loading plans are prepared by shore planners using computerized loading programs and CSMs provided by shipowners. Shore planners are used because masters apparently do not have the time to prepare such plans. GARD issued a report in 2016 that described how shore planners, without knowledge of vessel stability, breached the CSM when preparing loading plans:

As previously indicated, the CSM is valid only for certain metacentric height values (GM), which is problematic if the ship operates at a higher GM value.

The following are typical examples which describe the problems and explain the need for lashing software. The same CSM shows that in a certain bay on deck the containers can be stacked six tiers high, and that the tier weight from the base to the top is: 30 t, 20 t, 20 t, 15 t, 10 t, 7 t. The maximum stack weight is then 102 tons.

Containers, however, are never loaded exactly as prescribed by the CSM. If, for example, the container in the bottom tier weighs 21 tons instead of 30 tonnes, the first instinctive reaction may be that the forces will be less than the example given in the CSM, and the stowage would therefore be safe. However, the opposite is the case as less weight in the bottom tier will create higher forces as the center of gravity of the stack moves upwards.

Shipowners have been aware of container stack collapses since 2006, when results of a three-year study into container lashing were released. The study involved two ships, CMA-CGM Rigoletto and NYK Argus. Research participants included Marin (a Dutch research institute), shipowners, class societies, technology companies and lashing manufacturers. The project involved crew from these vessels completing a questionnaire with their reasons for container stack collapses.

About 25% percent of respondents stated that vessels were regularly operated in conditions outside of those described in the CSM. Typically, the GM values were above the maximum evaluated in the CSM. When sailing in partially loaded condition, this could not be avoided on many vessels.

Disappointingly, none of the recommendations proposed either a computerized loading program to ensure loading plans complied with CSMs or a safer container securing system.

A second three-year research project into container stack collapses commenced in mid-2021. It includes shipowners, P&I Clubs, class societies and industries associated with container shipping. However, its scope of work does not include the proposals mentioned earlier.

Diligent shipowners who read the safety investigation reports following container stack collapses would be aware that a principal cause of such collapses was loading plans not complying with the CSM. The report by Maritime Authority Investigation Bureau into the loss of containers from Ever Smart on Oct. 30, 2017, states:

Bay 70 was fully loaded nine tiers high with 151 x 40ft hi-cube containers. The total stack weights of the containers in all 17 rows of the stow were below the calculated limits; the weight in the starboard outer stack was just below 63t, which was 62% of the maximum permissible weight given in the CSM’s stack weight tables for a GM of 0.7m.

Nevertheless, anomalies were identified between the bay plan produced by the shore planners and approved by the ship’s master, and the requirements set out in the CSM. Of note:

• The container weights in the upper tiers of the stow exceeded the recommended limits given in the CSM’s stack weight table.
  Hi-cube containers were loaded eight high in the stow’s outer stacks(contrary to the CSM)
• The ship’s GM was above the indicative GMs provided for a fully loaded ship and that used to calculate the lashing requirements and stack weight limits for bay 70 when loaded nine tiers high with 151 containers. Therefore, the stowage plan did not comply with the requirements set out in the ship’s CSM and almost certainly contributed to the stow collapse.

The report by the Australian Transport Safety Bureau into the loss of containers from YM Efficiency on June 1, 2018, states:

There were significant deviations, particularly with respect to container weights and stack weights, tier heights, and other inconsistencies, as summarized below:

• The stowage arrangement exceeded the seven-tier limit specified (loaded to a height of 8 tier)
• Many stacks of 40-foot ‘high cube’ containers exceeded the maximum stack weights specified.
• Many container weights exceeded the weights specified for individual slots.
• There were many instances of heavy containers above lighter ones, contrary to principles of vertical distribution.

The preliminary report by ATSB into the loss of containers from APL England on May 24, 2020 states:

However, in addition to this, examination of the stowage arrangement showed that the security of the stow above the container cell guides used in bay 62 was affected by the use of high cube (2.9m (9’6″) high) as opposed to standard height (2.6 m (8’6″) containers.

The ATSB report references the earlier loss of containers from the same vessel:

At about 1500 (local time UTC + 8 hours) on Aug. 18, 2016, while transiting the Great Australian Bight, APL England lost 37 containers overboard in rough seas. The ATSB did not investigate the occurrence. The Australian Maritime Safety Authority conducted an investigation and concluded that:

• The vessel had a high metacentric height (GM) which may have contributed to generation of excessive dynamic forces leading to failure of container base sockets and collapse of the stow.
• The topmost container tier in all rows of the collapsed bay was over the recommended weight although the stack weight in each row was not exceeded.

Shipowners are required to know that the Convention for Safety of Life at Sea (SOLAS) places ultimate responsibility on masters, for the safe loading, stowing and securing of cargo/containers. In reality, masters receive loading plans perhaps a day, sometimes hours, before loading commences. This prevents masters, already overwhelmed by officialese, from ensuring that loading plans comply with the CSMs. Moreover, commercial pressures, noted in the Ever Smart MAIB Report, would preclude the majority of masters from requiring amendments to non-complying plans:

Containers could be loaded in compliance with the CSM if the computerized loading programs – used by shore planners and stevedores – were designed with fail-safe mechanisms. Using the GARD advice above as an example, the fail-safe mechanism would operate thus: if a container weighing 21 ton was allocated by a planner to a slot that had been designed by the CSM to take a container weighing 30 tons, the computer would reject the allocation.

Fail-safe mechanisms are basic features of computer programs, achievable without sophisticated computer skills. Fail-safe mechanisms in computer loading programs used by shore planners and stevedores, would ensure that every loading plan, and every container loaded onto a vessel, by stevedores, complied with the CSM, ensuring masters’ and shipowners’ compliance with their obligations under SOLAS and the Hague/Hague-Visby rules.

The IMO Code mentioned earlier requires the securing of cargo containers, to be “…based on the most severe weather conditions which may be expected by experience for the intended voyage.” A basic rule of securing cases of cargo similar in type and in the same location is to secure them as a block unit. If they were not secured as a block unit then, when the vessel rolled to port and starboard – the motion of a vessel most likely to cause improperly secured cargo to shift, break loose and get damaged. The individual cases would acquire a motion of their own, separate to the adjacent cases, and those individual motions would eventually de-stabilze all the individual cases, resulting in their collective collapse.

Container stacks could be considered as individual cases of cargo. They are vertical stacks of containers with the highest container currently about 12 feet – comprising the 12th tier above the vessel’s deck or hatch cover. At the present time, containers that are stacked up to the fifth tier, generally, are secured with lashing bars connecting the bases of each tier of containers to securing points on the deck.

It is my opinion, such containers can be considered as having been secured as block units. However, containers stacked above the fifth tier are not so secured. This is because securing would require lashing bars of greater length and weight–making them unsafe for handling by stevedores. Consequently, containers stowed between the sixth and the 12th tiers in container stacks, are not secured as block units.

Although connectors are used to tying containers above the fifth tier to one another, the ongoing collapse of container stacks speaks to their ineffectiveness. Paragraph 1.5.0 of the Ever Smart report notes the transverse accelerations experienced by containers in the upper tiers of a stack.

The magnitude of the transverse accelerations experienced by containers stowed on deck at sea differs depending on their location within a bay and their fore and aft position within the ship. The greatest transverse accelerations are typically experienced by those stowed in the upper tiers of the outer stacks.

Because the container stacks are not secured as block units, then, when the vessel rolls, each stack in the higher tiers, develops its own synchronous movement that eventually effects the securings of the containers within the stack, causing the stacks to collapse. If the container stacks were secured as block units, they would be capable of resisting the vessel’s rolling. It is unfortunate that none of the investigation reports perused by the writer have considered whether the container securing systems accorded with the IMO Code and, if so, whether the stacks would have collapsed had the containers been loaded in compliance with the CSM.

Shipowners should, as a matter of urgency, assemble a team of experts comprising P&I Clubs, class societies, hull and machinery underwriters, portsauthorities, stevedores and manufacturers of containers/lashing equipment, to design a system that secures container stacks as block units. The experts may decide to design lashing bars with carbon fiber, (considered to be superior in terms of weight, tensile and shear properties), that would permit the use of longer, but lighter lashing bars; or employ drones to handle longer and heavier lashing bars, with or without artificial intelligence; or a system that secured intermediate tiers between the sixth and 12th tiers. The probability that a safer securing system would delay a vessel’s schedule would not be acceptable, noting that a number of container ships have been slow steaming, resulting in extended schedules, since 2008.

It is inexcusable that, despite the availability of expertise, construction materials, technologies and artificial intelligence, shipowners have not taken measures to design a safer container securing system. As noted earlier, the 2021 research project is not tasked with designing a safer system for container loading plans and container securing systems.

The material provided above demonstrates that shipowners have been aware of the defects in their container loading plans and container securing systems before, and at the commencement of the voyages for a number of years. Shipowners who claim ignorance leave themselves open to criticism of willful blindness, as articulated by Lord Denning in The Eurysthenes.

If a man suspicious of the truth, turns a blind eye to it, and refrains from inquiry —so that he should not know it for certain —then he is to be regarded as knowing the truth. The ‘turning a blind eye’ is far more blameworthy than mere negligence.

And Lord Roskill in the same case:

If the facts amounting to unseaworthiness are there staring the assured in the face so that he must, had he thought of it, have realized their implication upon the unseaworthiness of his ship, he cannot escape from being held privy to that unseaworthiness by blindly or blandly ignoring those facts or by refraining from asking relevant questions regarding them in the hope that by his lack of inquiry he will not know for certain that which any inquiry must have made plain beyond possibility of doubt.

In a situation where a vessel’s container stacks collapse in heavy weather, and the investigation reveals that the vessel commenced its voyage with either one of the two known defects in question, a court could decide that the vessel was unseaworthy at the beginning of its voyage. Such a decision would be similar to the United Kingdom Supreme Court’s unanimous decision in the CMA CGM Libra.

This vessel was found to have been unseaworthy at the beginning of its voyage on May 17, 2011, from Xiamen, China, because its passage plan was defective. The court agreed with the trial judge’s decision and held that the vessel’s owners had breached their obligations under the Hague Rules of 1924.

Given the ‘essential importance’ of passage planning for the ‘safety … of navigation,’ applying the prudent owner test, a vessel is likely to be unseaworthy if she begins her voyage without a passage plan or if she does so with a defective passage plan which endangers the safety of the vessel,” the Supreme Court said.

The defect was the failure to mark, on the passage plan and the working chart – as required by the Guidelines for Voyage Planning – the shallow water areas outside the dredged channel. Shortly after the CMA CGM Libra dropped its pilot outbound from Xiamen, the master navigated the vessel outside the dredged channel, at a speed of about 11knots for about three minutes, resulting in the vessel grounding in shallow water outside the channel. The master explained that, if the shallow water areas outside the dredged channel had been marked on the working chart, he would not have left the dredged channel. The fact that the owner had provided the vessel with the wherewithal to prepare a compliant passage plan, did not absolve the shipowner from the master’s actions or from finding the vessel to have been unseaworthy at the beginning of its voyage.

The fact that navigation is the responsibility of the master and involves the exercise by the master and deck officers of their specialist skill and judgment makes no difference. The decision was that the vessel was unseaworthy at the beginning of its voyage because its passage plan was defective at that time; and the defective plan endangered the safety of the vessel.

Shipowners are operating their vessels with the two known defects, despite the costs associated with container stack collapses running in the hundreds of millions of dollars. Their willingness to accept such costs does not sit well with their concerns about the claims associated with passage planning, because those claims are a one-off and modest in value. The lack of action by shipowners to eliminate the defects in question, does little to debunk the notion that shipowners treat the cost of container collapses as inherent in operating container vessels.