NSIA Investigation: Fire due to seawater entering the ventilation system - SAFETY4SEA

2022-07-29 19:42:42 By : Ms. Sophie HU

The Norwegian Safety Investigation Authority (NSIA) published its report on the fire on board ‘MS Brim’ in the outer Oslofjord on 11 March 2021.

O n Wednesday 10 March 2021, the hybrid vessel ‘Brim’ was berthed in Fredrikstad to charge her batteries overnight. The next morning, the vessel was scheduled to sail to Sarpsborg for a planned event. On Thursday 11 March, after the event, the vessel set course for Sandefjord. The crew of four were prepared for some bad weather in the Oslofjord, and they had readied the vessel for the voyage and checked the battery and engine rooms before departure. The navigating officer, who also had the role of engineer on board, used electric power for propulsion while sailing down the Glomma river. On leaving Glomma, the diesel engines were started in order to charge the batteries.

GET THE SAFETY4SEA IN YOUR INBOX!

On entering the Oslofjord at approximately 15:00, the skipper increased the speed to initiate charging, but he reduced the speed again at 15:25 due to the sea becoming rougher. Southeasterly winds and waves hit the vessel on the port side, and the crew were under the impression that the waves were slapping against the catamaran’s tunnel top.

At approximately 15:27, after approximately 20 minutes’ charging of the batteries, all 12 battery stacks were disconnected and an alarm was transmitted to the EMS4 panel on the bridge. The skipper and engineer assessed this as being due to a ground fault in the panel. The skipper started troubleshooting from the bridge, but understood all systems to be in working order. Furthermore, he assumed that the ground fault was not critical, as they had experienced such faults several times on previous voyages. After troubleshooting for approximately two minutes, the skipper relieved the engineer at the wheel, so that he could go down into the port engine room, where the BMS panel was located, to further investigate the fault message on the panel.

The skipper reduced the speed when the EMS system also issued the warning ‘BMS serious’. The BMS system registered overheating of a module in battery stack 6, which, on the bridge, was displayed as a serious fault. The alarm provided no further details and therefore had to be investigated by checking the BMS panel in the port engine room.

Shortly after, at approximately 15:33, the fire alarm sounded on board. The skipper observed that the fire alarm panel on the bridge indicated fire in both the starboard engine room and starboard battery room.

The skipper also checked the engine status on the control panel and monitored the engine room via a camera installed in the room. The pressure and temperature readings for the diesel engines were normal.

The able seamen on the main deck called the skipper to find out why the fire alarm had sounded. They were told that the engineer was going down into the engine room to investigate. While they were on their way to the engine room, the power failed, but it came back after approximately one minute. There was no indication of battery pack consumption on the EMS panel on the bridge. On the main deck, one of the able seamen ran into the engineer, who was on his way to the engine room. The engineer found that, while many alarms had been triggered on the BMS panel, everything looked normal on the 230V panel. He registered that the light indicating that the emergency generator had started was on. The engineer found that use of the communication system to communicate with the bridge from the engine room was challenging, because of the noise from the alarms.

Because the fuel supply was shut off, the emergency generator, located on the starboard side, eventually gave up, so that control of navigation was lost and the screens and panels went black. At 15:39, emergency power was activated and only the most important electrical components were in operation. The skipper tried to steer the vessel towards Tjøme, using the port engine and the wave direction for directional stability.

At 15:41, the skipper sent a distress message on VHF channel 6. The distress call was answered by the Joint Rescue Coordination Centre (JRCC) and the skipper explained the situation. The engineer took over navigation of the vessel. The skipper continued to observe the situation and communicate from the bridge, while the able seamen readied the vessel’s immersion suits.

Immediately before the fire broke out, the battery system was disconnected as a result of a ground fault, which was indicated on the panel on the bridge. Ground faults had been a recurring problem since the vessel was new. The crew therefore perceived the alarm as ‘one of many’, and did not consider it to be serious. They had no possibility of identifying the point of origin of the ground fault alarm or ascertaining how serious it was.

The fire alarm panel indicated fire both in the engine room and in the battery room on the starboard side. This was probably because smoke had spread quickly through the fire division between the battery room and the engine room, as observed by the skipper via the surveillance camera in the engine room. It was also verified by the engineer when he arrived in the engine room. The battery room had fire insulation and was separated from adjacent rooms by fire walls designed to prevent the passage of smoke and flames for one hour. The incident showed that the fire division did not prevent the passage of smoke, which was one reason why the crew did not understand where the fire originated.

Furthermore, there was no camera surveillance of the battery room. The presence of a camera might have helped the crew to dispel the incorrect perception that it was the engine room that was on fire. The DNV’s updated classification rules from 2021 recommend camera surveillance of battery rooms to improve the crew’s situational awareness, in addition to gas monitoring for early detection of gases before they develop into smoke. These are deemed to be important aids to rapid detection of smoke development.

The skipper decided to initiate emergency shutdown of the starboard main engine and emergency generator. This led to much reduced redundancy for manoeuvring, instruments and propulsion.

Nonetheless, the NSIA cannot see that the reduction in propulsion and manoeuvring capabilities had a negative impact on the further sequence of events.

Because of the incorrect perception of fire in the engine room, the Novec fire suppression agent was first released to the engine room after approximately five minutes. When that failed to reduce the development of smoke, the crew understood that the fire was in the battery room. Hence, it  approximately seven minutes before Novec was released to the battery room. Too much time passed from the fire alarm went off until Novec was released for the fire suppression agent to have any significant effect on the battery fire.

The crew were unable to start the fire pump because the fuel supply for the emergency generator was shut off. It would not have been possible for the crew to use the fire hose in the battery room as the adjacent room was full of smoke. Moreover, use of the fire hose would probably have exacerbated the fire, as the fire pump used seawater. The extinguishing equipment available to the crew could therefore not have helped to put out the fire.

Evacuation and external extinguishing assistance was therefore the only option available to the crew. The Norwegian Society for Sea Rescue helped to cool down the hull with seawater fire pumps. The NSIA believes it could have exacerbated the fire if the water had come into contact with the batteries, which happened in a previous incident involving the car ferry ‘M/F Ytterøyningen’.

The fire on board most likely arose as a result of seawater entering the ventilation system and coming into contact with the high-voltage components of the battery system, causing shortcircuiting, electric arcs and fire. The investigation has also shown that the low IP rating enabled seawater and sea air to enter battery modules.

Late release of the fire suppression agent meant that it had little suppressive effect and did not prevent the fire from developing, but had a cooling effect for a short period. The investigation has shown that a clear extinguishing strategy that would limit the scope of damage in the event of lithium-ion battery fires is lacking.

The investigation has also identified several areas where the risks associated with the use of lithium-ion batteries was not sufficiently identified or addressed in the design. At present, DNV’s classification rules for battery safety do not sufficiently address the risks associated with the use of lithium-ion batteries on board vessels.

EXPLORE MORE AT NSIA’S ACCIDENT REPORT

Your email address will not be published. Required fields are marked *

Did you know that age matters for 2020? 19% more on the Detention Rate for Bulk Carriers 16-20 yo vs. 2019 in Australia.