X-Press Pearl: What went wrong? The scientific basis

Thursday, 17 June 2021 00:00 -     - {{hitsCtrl.values.hits}}

The fate of MV X-Press Pearl (Photo Credit: SLAF Media)  

 


The overview of the accident

MV X-Press Pearl was a container ship carrying approximately 1,500 containers that arrived at Colombo on the late hours of 19 May. Awaiting the port entry, it was then anchored outside the harbour at about 9.5 NM (about 18 km) away from the shoreline. A fire was reported at around noon the following day. Despite swift firefighting, the fire was not able to be extinguished and continued to escalate, partly supported by the monsoon winds.

After several days of fire, on 25 May morning, a violent explosion occurred onboard the vessel. Within hours, various debris were washed ashore along the coastline, including a large amount of polymer beads and a gummy textured dark substance which is suspected to be self-polymerised resins. By 27 May, the fire had largely consumed much of the flammable substances and the intensity of the fire was reduced to a pool fire deep down in the ship bottom. 

The next day, at around 23:00 time, a second explosion was reported. By 1 June, the fire was completely extinguished, and the salvors were able to board the ship for inspection. Attempts were made next day to tow it further away into the sea, and then a large section of the hull was found to be disintegrated and the engine room area was observed to be flooded. Shortly after, the ship started to sink. 

By the end of the day, the aft section had settled on the sea floor at a depth of about 21 m, while the forward section was still slowly sinking. The middle section had been completely submerged in water, while a pile of container rubble was still visible above the water level. The building structure located at the aft will continue to be visible above the surface due to its height. It is evident that the ship is now a wreck in the SL waters, needing to be further dealt with in future. 



The initiating events 

On 11 May, about eight days prior to the arrival at Colombo, an acid leak was observed on a container carrying 25 tonnes of nitric acid. This was when the ship was at Hamad port in Qatar. Attempts of discharging this leaking container at Hamad port was rejected by the port authorities. A similar rejection happened when the ship was at Hazira, India, four days later (15 May). The continued leaking of acid can be the precursor to the subsequent fire that occurred five days later at Colombo. 

There could be several exothermic chemical reactions responsible for the initiation of this primary fire. The nitric acid (HNO3) itself does not burn as a fuel. But it is a strong oxidising agent and it can violently react with many organic matters as well as with certain metals. Nitric acid reacts with organic carbon to generate nitrogen dioxide (NO2) gas (or sometimes nitric oxide NO) and carbon dioxide (CO2). This is a highly exothermic reaction (i.e. it generates a lot of heat). This heat further accelerates the reaction rate. If the heat is not removed (e.g. by firefighting water) the reaction can become a runaway reaction or in other words a ‘thermal explosion’. 

Generation of NO2 gas was evident during the first day of the fire initiation, as reddish-brown smoke was seen emanating from the top forward section of the ship. NO2 is a toxic gas that causes severe air pollution. 

The fire can also thermally decompose nitric acid to generate NO2 and oxygen gas (O2). The releasing oxygen can further accelerate combustion, even if the location is not well-ventilated (e.g. deep down in the cargo hold, or inside a container).



Contribution from other hazardous substances

Apparently, there was only one nitric acid container (with 25 tonnes of acid) stored in the forward section of the ship. This is very likely to be the place where the acid leak started. The adjacent container was filled with epoxy resin. Also, the nearby containers had vinyl acetate, methanol, and lithium-ion batteries – all flammable and potentially explosion-causing agents. 

The epoxy resin itself is flammable, and if mixed with nitric acid it will decompose exothermically, giving off more heat and starting a fire. 

Further, the reaction between nitric acid and methanol produces methyl nitrate (CH3NO3), which is a strong explosive and also a highly flammable substance. 

Lithium in the lithium-ion batteries may react with nitric acid to produce lithium nitrate and nitric oxide gas (NO). The reaction is exothermic (generates further heat). Further, Li-ion batteries can catch fire and once started it is very difficult to extinguish. The fire could often re-emerge once the firefighting /cooling is stopped. 

The ship’s cargo contained a number of other flammable and combustible materials at different locations. Vinyl acetate, aluminium smelting by products, sodium methylate, as well as HDPE (high density polyethylene) and LDPE (low density polyethylene) were among them. 

Aluminium smelting by products can react with water to generate highly flammable and toxic gases such as hydrogen, acetylene, methane, ammonia and even phosphine. This means even the fire water used in firefighting could have led to further escalation of the fire, if water had reached the locations where these aluminium smelting by products were stored (below deck).

Sodium methylate reacts with water generating methanol (which is flammable/potentially explosive).

The ship contained a large amount of HDPE and LDPE bags. These are combustible polymer beads that can quickly fuel an ongoing fire, leading to further escalation of the fire.

Further, the ship was reported to have approx. 700 tonnes of fuel; mostly the heavy fuel oil for engines. A significant quantity of lube oils, and also some quantities of fuel sludge (resulting from onboard fuel oil purification) were present. All these are combustible materials.



What caused the explosions?

As mentioned before, there were many agents that could have led to explosions. In addition to nitric acid’s explosive reactions with organic matter, other agents such as methanol, vinyl acetate, aluminium smelting by products, and sodium methylate (in alcohol) could also have contributed to the occurrence of the first explosion (on 25 May). Among them, methanol and aluminium smelting by products can be direct contenders. Due to the very high congestion on this container ship, a gas explosion could have easily occurred when a flammable gas cloud was accumulated within a partially confined space, such as the below deck cargo holds, or even within a single container. 

The second explosion reported on 28 May could very well be due to the partially extinguished pool fires in the ship’s fuel tanks. It is a well-known fact that extinguished or partially extinguished ventilation-controlled fuel pool fires can lead to gas explosions, as unburnt but continuously emitting flammable gases are present. In such scenario, letting the fire continue is a better option that should be considered during firefighting.



Applicable international regulations

The container ship was filled with a large amount of hazardous cargo. Transport of such large quantities with many different varieties of chemical substances poses a large risk factor. That makes it vulnerable to various chemical reactions; while also making firefighting very challenging. The International Maritime Dangerous Goods Code (IMDG code) developed by International Maritime Organization (IMO) is applicable to this cargo. 

Accordingly, the relevant packaging, categorising, labelling, segregation, stowage, etc. must have been properly done in accordance with the IMDG code. Further, the ship itself should have been correctly certified by its classification society (reported to be a DNV class ship) to carry this kind of hazardous cargo. The ship’s design and the onboard safety systems (including firefighting) are dictated by the relevant classification class of the ship. In addition, various other IMO regulations and guidelines are also applicable; including the International Code for Fire Safety Systems – FSS code.



Was the accident preventable?

It might have been preventable, should common sense have prevailed during the days prior to the initial fire event on 20 May. First, the leaking acid container did mean that either proper packaging was not done (as per IMDG code), or proper handling of the container had not happened. 

When the acid leak was first noticed, the ship had requested both Hamad and Hazira ports to allow unloading the leaking acid container. Unfortunately, both ports had rejected the request. This cannot be considered as an act of goodwill, and must be addressed as a matter of urgent attention in future formulations of international maritime law and the standards applicable to potentially hazardous situations onboard ships. Nonetheless, if the ship had sufficiently emphasised the gravity of the situation to the ports is an open question. 

The next obvious decision the ship could have made was to discharge the leaking acid container at sea. This is totally allowed as per IMDG code. If the ship’s master comes to the judgment that a certain cargo is immediately hazardous to the lives of the crew, such goods can be disposed in the sea. In this case, the onboard cranes apparently had access to the relevant container location and had capacity to lift the less than 30 MT container and drop it in the sea. 

If any of the above were not done, then the last remaining action that could have prevented a fire from occurring was to continuously cool the area subjected to the acid leak and simultaneously dilute the leaking acid with ample amount of water (until the leaking container could be removed from the ship). Whether this had been done properly or not, should be the subject of further investigation. 

Further, the ship did not apparently notify Colombo port about the acid leak issue when they were arriving (until the fire started next day), delaying any possibility of assisted control of the situation. Those last 12 hours or so could have made a big difference in the eventual outcome. 



What are the eventual environmental consequences?

As per the cargo load list released to the open media (via Right to Information requests), X-Press Pearl contained various chemical substances including a number of environmentally hazardous organic substances, such as acetyl cedrane used in perfumery products that is categorised as a substance posing acute and chronic toxicity to marine life. However, it is likely that many of such organic substances are decomposed and burnt in the fire event, thus avoiding them being directly released to the marine environment. 

The ship was reported to have about 700 MT of fuels and lube oils. It can reasonably be expected that a significant portion of this may have also been burnt in the fire. However, some fuel amounts might still be trapped in the fuel tanks of the ship (which were either not ruptured open during the fire, or the pool fires at those fuel tanks were ventilation controlled, and also possibly extinguished by firefighting). In that case, these oils (if remaining) are pending to be released to the environment. 

While the ship was still afloat during the fire, the bottom sections were evidently filled with a large amount of water (i.e. used in firefighting, and also the water leaked through the damaged hull). This water must have been contaminated with various toxic and environmentally harmful chemical ingredients; including potentially persistent and bio accumulative organic substances. As the ship has partially sunk now, this cocktail of liquid chemicals (mixed with water) must have already been released into the sea. 

The observations made on 2 June indicated a yellowish substance floating near the ship. This is suspected to be originating from a large sulfur stock on the ship. This is indicative of the release of any liquid contents or other unburnt substances collected at the bottom sections of the ship. Nevertheless, the wave action and the sea currents should lead to rapid dispersion and dilution of any chemical concentrations; including the release of any acidic or alkaline chemicals such as nitric acid and sodium hydroxide (which was present in a large quantity). 

The ship’s cargo load list also indicates a large amount of scrap metals and metallic Lead slabs. These heavy metals can cause a considerable marine pollution, and can cause long term harm to human health, as many of these metals are slowly released to seawater (due to chemical and biochemical reactions) and then they bioaccumulate and bioconcentrate along the food chains, eventually entering human body via sea food consumption. 

A large-scale marine and coastal pollution has happened due to the spreading of micro-plastics in the form of polymer beads. However, apart from some dead fish and several sea turtles, no mass scale fish death has been reported so far. The observed few deaths could have been caused by the reduced dissolved oxygen levels in the water closest to the burning ship, resulted from increased temperatures. Nonetheless, large temperature increases cannot occur much further away from the burning ship as the mixing action of waves and currents quickly equalise sea water temperature to ambient condition. In addition, surface level emissions, chemical toxicity effects, and slightly changed pH levels could also have contributed to some destruction of sea life near the location. 

The fire had released a substantial amount of air pollutants, which were directly carried towards the land due to the prevailing weather conditions. A significant part of the fire duration may have proceeded at relatively lower combustion temperatures. Such low temperature incomplete burning of plastics releases various harmful fumes including polychlorinated dibenzo-p-dioxins and polychlorinated dibenzo furans; known to be very harmful and persistent pollutants. These pollutants are basically scattered and settled within the island. The smoke plumes from the fire should have had many other toxic gases and components such as NO2, NO, PH3, CO, NH3, various organic vapours, as well as particulate matter such as hazardous ash and soot. 

(Note: The facts presented in this article are based on information available on public domains. The analyses and opinions are based on the author’s long-term experience in the field, and are not connected to any institutional interest.)


The writer holds a Ph.D., an M.Sc.Tech. (Norway), M.Sc., B.Sc.Eng. (1st Hons., UoM), AIChE, AMIE(SL).)


 

COMMENTS