Jan 272015

Three men lay more than a hundred yards from the thick torn metal that once covered the top forward ballast tank, they were dead.

In the gathering darkness, in the roughening seas around the ship, the bodies of four other men were being carried away on the current, three of them never to be found. Inside the gray powder-coated ballast tank, burned and injured one man lived. He would not survive his injuries.

The last sound he heard, if he heard it, before the massive explosion may have been the quiet pop of a light-bulb breaking…


bosunblue.jpgabblue.jpgabblue.jpgabblue.jpg abblue.jpgabblue.jpgabblue.jpgabblue.jpg

Their names were Xin Zeng Wang; Dee Qiang Han, Zhen Yan, Jun Ji, Dan Zhao and Yi Li, Jun Kong and Fan Jun Meng. Real names, real people. They were eight of the 23 strong compliment of the Hong Kong flagged geared bulk carrier Nego Kim in November 2001.

Like the rest of the crew they came from the People’s Republic of China.

The Ship

Nego Kim

The Nego Kim (Now the Sinfos Bay) in 2006, from www.shipspotting.com

The Nego Kim was 26,951 tonnes deadweight, 167.2 metres length overall with a beam of 26 metres.

At 0200 on 17th November 2001 she arrived off Port Dampier Western Australia on a time charter with a cargo of 17,000 tonnes of scrap metal. She was to pick up another load of scrap at the Material Offloading Facility Wharf at Dampier for discharge in Indonesia and Singapore.

Spraying A Way To Doom

It was a confusing arrival. The Master tried to contact Dampier Port Control for instructions but failed. He didn’t know that the port control no longer operated on a 24 hour basis. He was heard by a radio operator at the Hammersley Iron Company which owned its own wharf, who told him to contact the ship’s agent for instructions.

The master didn’t have the agent’s home number or cellphone number.

It was a weekend so the Nego Kim 15.4 miles off Dampier that Saturday and Sunday to wait for instructions.

Meanwhile, the crew continued the ship’s planned maintenance programme which involved cleaning and painting the interiors of the top ballast tanks. Over the previous week the two manhole covers of the port topside ballast tank had been removed and the inside of the tank cleaned and the surfaces prepared for painting.

port top ballast tank

None of the eight ratings survived the incident so we can’t be absolutely sure of every detail but the investigation and forensics evidence suggests that it went something like this:

It’s now 0800 hours on Sunday, November 18. The chief officer has given safety instructions to the bosun that safety clothing is to be worn, there is to be no smoking and respirators are to be worn inside the tank.

The mate checks the atmosphere in the tank. It’s 21 per cent oxygen so it’s regarded as safe. He doesn’t use the ship’s explosimeter in part because it’s not considered necessary, the ship’s only carrying scrap iron, but mainly because it’s in Singapore being repaired. He signs the company checklist and confined space entry permit.

The men don cotton overalls and safety shoes and work begins.

A 19mm compress air hose is fed through the after manhole with its discharge end laid on the deck of the ballast tank near where the work is to be done. A 300 mm electric fan is put on the deck beside the manhole to blow air into the tank.


A fan similar to this was set up to ventilate the tank

It’s dark in the tank so a 300 watt cargo light with a green circular shade is rigged with a lanyard for a deck hand to raise and lower it as needed. Sometimes a second man is in the tank to direct the light.


The epoxy paint is to be applied by an airless spraygun so a pneumatic pump is set up on deck and the hose led through the after manhole to the tank. A paint reservoir is set up next to the pump beside a 20 litre drum for mixing the p


Communication between the deck and elsewhere on the ship is by VHF radio.

Like the fan and the lamp, the radio was not explosion-proof or intrinsically safe, after all, nobody saw much reason for such safety measures, they were only going to paint the tank.

At 1300 the chief officer checks the oxygen levels in the tank again and they are satisfactory.

The bosun begins mixing the paint and hardener and put it in the paint reservoir. If someone had read the manufacturers materials safety data sheet he’d have known that mechanical mixing should have used, but there isn’t a MSDS on the ship and it doesn’t matter anyway because there isn’t a mechanical mixer on the ship either.

lamp before

A lamp similar to the one used in the port top forward ballast tank

Jun Ji puts a mask over his nose and mouth, a visor over his eyes, gets into the tank and starts spraying.

After five minutes or so, the spray gun develops a leak so work stops while it is repaired. At 14.30 the chief officer again checks the oxygen in the tank and finds it satisfactory.

It’s hot, the temperature on the bridge is 26 degrees Celsius, in the tank it is as much as 38 degrees celsius. The paint is difficult to mix and hardening fast in the heat so thinners is added to keep it workable. If someone had read the non-existent data sheet they’d have known that the maximum amount of thinners that should be added was 5 per cent. But they didn’t so they keep on adding thinners until the level reaches between 30 and 50 per cent of the paint.

Thinners is a mixture of hydrocarbon-based solvents. As Jun Ji sprays, the tank fills with hydrocarbon fumes. The fumes are heavier than air so they collect near the floor.

When hydrocarbons like this reach a certain percentage in air they reach what’s called the Lower Explosive Limit or LEL. At that point, if given a source of ignition, they will ignite or explode. Another element is the temperature at which they will ignite or explode if there is a source of ignition. It’s called the flashpoint. The temperature inside the tank was well above the flashpoint of the thinners.

The fan and the air hose together do not deliver enough air to keep the fumes inside the tank below their LEL.

Within as little as forty minutes of work, Jun Ji is inside a bomb waiting for its fuse to be lit.

At about 15.30 the chief officer gives the VHF radio to the deck fitter and at 1600 takes over the anchor watch on the bridge. In another hour work will be finished for the day.

As the end of the work day approaches the men start to withdraw equipment from the ballast tank. As the crewman with the lamp pulls it up he drops it, or knocks it, the bulb breaks.

At 16.40, the ballast tank erupts, tearing a five metre gash in the deck, raising it by three metres and curling the steelplate. Three men are thrown the length of the deck. They die instantly. Four more men are thrown overboard, a flaming drum of thinners spins across the deck. A mixture of burning paint and thinners roast the aft end of the tank.

fromair2.jpg fromair.jpg

frombow.jpgmai011118_a.jpg fromdeck.jpgmai011118_b.jpg

These photographs show the enormity of the explosion

The crew extinguished the fire with water and dry powder and found Jun Ji still alive, with burns over 90 per cent of his body. They helped him out through the ruptured deck plating where he fell unconscious. It took until December 4 for him to die in a Perth hospital without recovering consciousness.


Jun Ji was pulled from this tank with 90 per cent burns. He died nearly three weeks later

The emergency response was hampered by confusion, approaching darkness and weather conditions. About 24 hours later the body of one of the men blown overboard was found, the others are still missing four years later.

Searching For The Spark

What caused the explosion? And how do you stop it happening to you?

What we’re looking for is the spark that set it off.

Nobody seems to have been smoking on deck and there’s nothing to suggest that a metal object was accidentally dropped into the ballast tank causing a spark.

On deck, the VHF radio wasn’t explosion-proof or intrinsically safe and simply pushing a button could have produced the spark that ignited the flammable paint vapours. Given that there was wind, it’s unlikely that the concentration of flammable vapours on deck would have been high enough to ignite and spread the explosion into the ballast tank. We can’t be sure because the radio, and the man who held it, have not been found.

What else was on deck? The fan. That wasn’t explosion-proof or intrinsically safe either. It could have produced the spark that set off the explosion. Again, it was on deck where there probably would not have been enough vapour to ignite.

My bet is on the cargo lamp. You already know it wasn’t intrinsically safe or explosion-proof. A drop or knock could have broken the light bulb exposing the hot filament to the explosive vapours. If the bulb wasn’t firmly in it socket it could have caused a spark. If the cement holding the glass of the bulb to the bulb cap failed, the explosive vapours would have been sucked into the bulb and met the hot filament.


The remains of what was probably the deadly lampshade, its protective mesh was found inside the tank

In fact, the light shade was found outside on deck but the protective mesh that was over the bulb was found on the bottom of the ballast tank, which suggests that the explosion started just underneath the light shade.

The truth is, it really doesn’t matter whether it was the radio, the fan or the lamp that set off the explosion, any one of them could have done. None of them should have been used in or near a potentially explosive atmosphere.

Why was the atmosphere in the tank so explosive? After all, it was ventilated with a compressed airline and a fan. The compressed air line only delivered a little more than 2 cubic metres a minute. The fan was rated to deliver 61 cubic metres a minute, it actually delivered far less and much of what was delivered was lost because the fan was blowing through the manhole at an angle.

The explosive vapours were heavier than air, the fan blew from the top of the tank, so it probably wasn’t doing much to reduce the concentration of the vapours. With this kind of fan, trunking should have been used to carry the air to the lower part of the tank, that way, the vapours may not have reached an explosive level.

So, an underpowered fan delivered insufficient air to the wrong place and the tank became a bomb.

The paint manufacturer recommends using an exhaust fan to remove the vapours. That, too, would have helped keep them below the level at which they would explode.

Basically, the wrong equipment was being used the wrong way.

Why weren’t they aware that the atmosphere in the tank was explosive? They didn’t have an explosimeter to test it and, anyway, seem to have only considered it when carrying a potentially inflammable cargo. They did not understand the risk. No Materials Safety Data Sheet, MSDS, was carried on board but the company’s work order for the job insists that paint and unknown materials should be treated as chemicals for safety purposes.

Why was the atmosphere so explosive? The instructions for the paint insist that no more than 5 per cent thinners should be added but the bosun used around 30 per cent, six times more. Had he followed the instructions eight men, including the bosun, might still be alive.

Why did he use so much thinners? Probably because it was hard to mix the paint and its hardener thoroughly by hand and it became difficult to handle.

Why did that happen? The paint manufacturer recommends using a pneumatic mixer but the Nego Kim didn’t have one on board so the men on the Nego Kim were doing it manually and having a hard time of it. The shortage of that one piece of equipment may have laid the trail to tragedy.

So, how do you protect yourself? Always consider paint to be potentially explosive and use safety procedures with that in mind. Use intrinsically safe equipment, and use the right equipment for the job. Follow the manufacturers instructions. Even if you think you’re familiar with the materials you’re handling, check the MSDS if its available, if it isn’t treat the materials as hazardous. Your life and those of your crewmates may depend it.

This is Bob Couttie wishing you safe sailing.

Official ATSB Report

Sorry, the comment form is closed at this time.