DMAIB published its report on Maersk Essen, which on 16 January 2021 lost approximately 250 containers while the ship was en route from Xiamen, China to Los Angeles, USA. The number of lost containers were later adjusted to 750.
The incident
At 0400 on 16 January, MAERSK ESSEN was underway approximately 500 nm north of Hawaii.
At 0400, the chief officer came to the bridge and relieved the 2nd officer. During the handover, the 2nd officer informed the chief officer that increased wave heights were forecast later in the day. The chief officer could not see the sea surface due to darkness, but felt the ship moving comfortably in the sea. MAERSK ESSEN was following a heading of 087° on autopilot in port quartering seas. By now, the wind speed had increased to a strong breeze.
At approximately 0600, some of the crew were woken by the sound of cups and laptops sliding and falling off tables in their cabins as the ship took a few heavy, slow rolls. On the bridge, the chief officer saw from the inclinometer that the ship had rolled 15° to each side. The rolls stopped as quickly, as they had started. The chief officer observed nothing unusual out of the windows, and the sea state seemed not to have changed (Figure 5).
The chief officer thought nothing more of this, as it was normal for the ship to take an occasional heavier roll. Shortly after, an email with a weather warning was received from WNI advising that MAERSK ESSEN was expected to encounter deteriorating weather later that day. It also advised making heavy weather preparations, and that speed and course adjustments might be necessary to reduce the ship’s motions. The chief officer planned to inform the master of the email after his meeting with the deck ratings at 0800.
At 0745, the 3rd officer came to the bridge to relieve the chief officer. They talked about the heavy rolling earlier in the morning and the swell always being high in that sea area. The chief officer also informed the 3rd officer that the sea conditions were expected to deteriorate later in the day. The 3rd officer took over the bridge watch at 0800, and the chief officer went to the deck office to inform the deck ratings about the planned work of the day. A week had passed since the lashings had been checked and this was therefore on this day’s work schedule.
At 0804, the ship suddenly rolled approximately 15° to each side. On the bridge, the 3rd officer immediately adjusted the course 2-3° to starboard, although the vessel was already stabilising. He called the master, who was taking breakfast in the mess room, and asked him to come to the bridge. Meanwhile the deck ratings and the chief officer left the deck office and went to their cabins to secure them for heavy weather.
At 0807, MAERSK ESSEN again started to roll heavily. This time the angles of roll were greater than those previously experienced and increased with each roll cycle, resulting in the 3rd officer struggling to keep his balance and having to hold on to the bar on the conning station to stay upright. Loose items slid across the bridge, and alarms started to sound. Again, the 3rd officer altered course 2-3° to starboard using the autopilot to stop the rolling. The heavy rolling stopped two minutes after it had started.
The master, who had rushed the flight of stairs from deck B to deck G, reached the bridge 15 seconds after the ship stabilised. He immediately took the hand steering and made a large course alteration to starboard from approximately 090° to 130° and increased speed. Meanwhile, the chief officer, 2nd officer and cadet came to the bridge. They started to clean the bridge, and the chief officer gave orders for the ratings to clean in the galley and mess rooms.
At 0815, the master detected an echo on the radar close astern of the ship.
He went to the bridge wing and looked aft and saw containers hanging over the ship’s side and floating in the sea. He rushed back into the bridge and instructed the officers to fetch a camera to document the lost containers and to log the time and position.
The heavy weather emergency response flowchart was put into use, and the 2nd and 3rd officer assisted the master in communicating with relevant authorities and the company.
Concerned if any crewmember might had been on deck during the rolls, the master sounded the general alarm to initiate a head count. The head count was conducted by the chief officer at the upper deck fire station, and none of the crew were identified as missing or injured. The master instructed the crew to stay away from the deck areas, but authorised the chief officer and the dayman to carry out an initial assessment of the damages and loss of cargo. They estimated that approximately 215 containers were missing.
Analysis
#1 Loss of cargo: DMAIB has found that the cargo stowage and securing operations on MAERSK ESSEN were open to uncertainties and variabilities which could influence both the forces acting on the container stacks and the holding capacity of the cargo securing equipment. By themselves, these uncertainties and variabilities did not have the potential to cause the container stack collapses seen on MAERSK ESSEN.
#2 Heavy rolling: The investigation of the heavy rolling on the day of the accident concluded that MAERSK ESSEN most likely experienced parametric resonance, possibly in combination with pure loss of stability on a wave crest. This resulted in large roll angles building up during a six-minute period.
#3 Weather routing: DMAIB has examined the tools for predicting risk of parametric resonance made available to the company’s fleet. Common to them was that they were dependent on forecast data. Forecasts are encumbered by uncertainty and will vary depending on the weather suppliers’ data sources and calculation models. The parametric risk calculators were found to be prone to this type of uncertainty, which can result in misleading indications of risk.
Conclusion
#1 Probable cause: The investigation determined that the heavy rolling was most likely a result of parametric resonance. The acceleration forces acting on
the container stacks during the heavy rolling exposed the cargo securing equipment to stress loads which they were neither designed nor able to withstand. MAERSK ESSEN’s loading condition required the ship to avoid roll angles exceeding 19.18° in order to stay within the stress load limits defined in the ship’s loading and stability computer. This limit was exceeded at the time of the container loss.
To avoid heavy rolling, the ship used weather routing on board and received weather routing advice from an external weather service supplier. The weather routing focused on avoiding heavy weather and rough sea states. Parametric resonance was not mentioned in the correspondence between the crew on MAERSK ESSEN and the weather service supplier, and the crew had no onboard tools to analyse or monitor the risk of parametric resonance. They had to rely on reactive strategy, if the phenomena occurred. During the heavy rolling episodes, the bridge crew took action to stop the rolling motions by altering course and speed. By then, the container stacks had already collapsed.
The conditions triggering the parametric rolling of MAERSK ESSEN were within the spectrum of normal operational conditions for the ship and in normal sea conditions for the area.
These circumstances add a critical dimension to the risk of parametric rolling, because the development of the conditions required to trigger this phenomenon occurred without being evident to the crew. Consequently, normal work continued, and on the day of the accident the crew was minutes away from entering the cargo deck to check lashings, as the rolling occurred. If the crew had been on deck, the collapsing container stacks could have resulted in fatalities.
The investigation of MAERSK ESSEN’s heavy rolling accident concluded that parametric rolling was considered a rare phenomenon by the crew. However, the triggering conditions for parametric rolling were not rare for this type of ship in this sea area, and hence it is likely that the ship had experienced resonance effects earlier and on previous voyages, and that accidents are likely to occur again unless effective efforts to avoid the triggering conditions are implemented.
#2 Lessons learned: Detecting risk of parametric resonance rolling based on forecasted sea conditions can be problematic as forecasts are encumbered by uncertainty. No matter how automatised and detailed the onboard tools for monitoring parametric resonance are, they are prone to the uncertainty of the forecasts which make them unreliable as tools, unless a broad risk margin is applied.
DMAIB encourages companies and authorities to explore and test options for predicting resonance effects that are based on real-time conditions rather than forecasts.
EXPLORE MORE AT DMAIB’S REPORT ON MAERSK ESSEN
