80973876 explosion and loss of the bulk carrier ythan

Introduction

Authority

Vessel Particulars & Finding of Facts

Conclusions

Recommendations &Acknowledgements

Appendix I

1

2

3

11

12

13

Inside This Report: Explosion and Loss of the Bulk Carrier YTHAN (O.N. 1693, IMO# 8117067) Off Santa Marta, Colombia on 28 February 2004

The bulk carrier YTHAN, carrying a cargo of Direct Reduced Iron (DRI)

Fines from Venezuela to China, experienced a series of explosions in the

cargo holds on 28 February 2004, while at sea in a location North of Santa

Marta, Colombia.

The explosions resulted in the sinking of the vessel. Consequently, there

were six (6) fatalities. Witness statements confirmed the death of the

vessel’s Master, and five (5) members of the Engineering staff remain

missing after the casualty and are presumed deceased.

The investigation led to the conclusion that the explosions were the result of

a dangerous concentration of hydrogen gas built up within the atmosphere

of the cargo holds, triggered by an unknown ignition source.

This Report, and any appended Maritime Administrator’s

Decision, sets forth certain findings which have been

ascertained or determined up to the time of its issuance, and is

published to alert the shipping industry and the public of the

general circumstances of this accident. While every effort has

been made to ensure the accuracy of the information contained

in this Report, this Administration and its representatives can

accept no liability for any error or omission alleged to be

contained herein.

Extracts can be published without specific permission

providing that the source is duly acknowledged,

otherwise please obtain permission prior to reproduction

of the Report.

1

MARSHALL ISLANDSMarine Investigation Report

Bulk Carrier YTHAN alongside Palua Pier during loading

Office of the Maritime Administrator

11495 Commerce Park DriveReston, Virginia 20191

Tel: +1-703-620-4880 Fax: +1-703-476-8522

Contact: Nick MakarE-mail: [email protected]

Marine Investigation Report

2

DisclaimerThe Flag Administration has a mandate to promote safety of life and property at sea, and the prevention of

pollution. This is achieved in part by conducting investigations of marine casualties involving vessels in the

registry for remedial purpose in accordance with the Republic of the Marshall Islands Maritime Act 1990

and Maritime Regulations. Marine investigations, which are administrative in nature, look to the cause of

difficulties experienced, means of avoiding them in the future, possible violations of law, and any faults or

failures on the part of personnel, shipowners or operators which might require action in respect of any licenses,

certificates or documents.

It is not the function of the Administration or the purpose of the investigation to assign fault or determine civil

or criminal liability with respect to enhancing the litigation posture of any party. The Administration must

report, or cause to be reported, the circumstances and proximate cause or causes of a marine casualty and

contributory factors. However, where it is determined that there exists evidence of criminal conduct under

the laws of the Republic of the Marshall Islands on the part of any seafarer holding a Republic of the Marshall

Islands Certificate of Competency or other official document, the matter would be referred to the Ministry of

Justice of the Republic of the Marshall Islands for appropriate action.

AuthorityAn investigation under the authority of the Marshall Islands Maritime Regulation 6.38 was conducted to

determine the cause of the casualty and recommend remedial measures.

The Republic of the Marshall Islands

3The Republic of the Marshall Islands

FINDINGS OF FACT

1. Description of the Vessel

1.1. The YTHAN was built at Nippon Kokan K. K. Shimuzu Shipyard in Japan in August 1984. She was a bulk carrier strengthened for heavy cargoes, and was of a conventional steel construction with five cargo holds forward of the accommodation structure at the stern. The engine room was below the accommodation structure, and shared a common bulkhead with No. 5 hold.

1.2. The hatch covers were of the folding steel panel type, operated by hydraulic rams, and secured by cleats and dogs. The No. 1 hold was smaller than the other four holds, and was covered by two folding hinged panels that opened aft. No. 2 – No. 5 holds were similar and covered by four panels, arranged in pairs so that two folded forward, and two folded aft. The holds were fitted with natural ventilation mushroom type vents located in the forward port and after starboard corners, and no trunking in the holds. Four cranes were fitted on the centerline, one between each hold. There was no fixed lighting in the cargo holds, and a fixed CO fire extinguishing system and a detection system was installed with coverage for each of the holds.

2. Direct Reduced Iron (DRI):(For further supplementary information regarding the production and hazards associated with DRI, please refer to Appendix I – Direct Reduced Iron (DRI))

2.1. DRI is the result of a process of converting iron ore into metallic iron without a smelting process. The process can be simply described as de-oxidizing iron ore. The direct reduction process yields DRI primarily as lumps, pellets, and Fines, which can also be formed into briquettes for shipping through either a cold or hot moulding process.

Vessel ParticularsVessel Name

YTHAN

OwnerYthan Ltd.

FlagMarshall Islands

Vessel TypeBulk Carrier

Official Number1693

IMO Number8117067

Length167.36 meters

Breadth29.50 meters

Depth14.80 meters

Gross Tonnage21,309 metric tons

Net Tonnage13,027 metric tons

Deadweight35,315 metric tons

Year Built1984

Classification SocietyLloyd’s Register

Class Notation• 100A1 Bulk Carrier

Strengthened for Heavy Cargoes, Nos. 2 & 4 Holds may be empty, ESP, L1,

ESN-Hold, +LMC and UMS

Engine/Power8716 KQ / 11,688 HP

Persons on Board27

Cargo Carried33,760 metric tons DRI Fines

Fig. 1 - YTHAN - Portside Main Deck


²

4

2.2. Cold Moulding - When the DRI briquettes are moulded at a temperature of <650° C, with the density being relatively unchanged, the product is termed Cold Moulded Briquettes.

2.3. Hot Moulding - Due to the potential hazards associated with the DRI re-oxidizing, passivation methods can be employed to reduce the reactivity of the material. The most effective method involves taking the hot DRI directly from the reduction process, or heating the DRI to a temperature of >650° C, and forming into briquettes, which increases the density to approximately 5.0 g/cc. DRI briquettes formed in this fashion are termed Hot-Moulded Briquettes, or Hot Briquetted Iron (HBI).

2.4. The direct reduction process is reversible. DRI will, when exposed to the elements (air/moisture), slowly re-oxidize in an exothermic reaction, generating heat. Furthermore, when DRI comes in contact with water, it may also slowly emit or evolve hydrogen gas. The degree of re-oxidation depends on the total available surface area of the material, the condition of the material, and the environment. Fines have the largest total exposed surface and briquettes the smallest. The quality and porosity of the DRI also plays a role in this process of re-oxidizing – iron is a relatively reactive substance, therefore pure DRI will have a much more enhanced reaction.

2.5. When stowed in a confined space, such as the cargo hold of a ship, heating of the cargo can occur. When water or moisture is present in the hold, or the cargo is loaded wet, hydrogen gas may also be produced. Therefore, the possibility exists whereby, if the temperature in the ship’s hold is high enough, and a hydrogen concentration above the explosive limit of approximately 4.1% is present, an explosion in the ship’s hold may occur.

2.6. Inerting of the ship’s hold, or proper ventilation when appropriate, and a dry hold condition will minimize the chances of such an explosion. However, smaller pockets of heat and hydrogen may still exist, and the possibility of an explosion may still be present.

2.7. The process of re-oxidation of DRI will also reduce the oxygen content in the hold when not properly ventilated, and will make entry into the hold dangerous.

2.8. DRI Fines - DRI Fines are not only the result of the direct reduction process, but also can be generated as scrapings and chips resulting from the handling of the pellets or briquettes.


As the pellets or briquettes are moved or sent along a conveyor belt, the material can rub and bounce about, chipping and scraping fragments off the edges and surfaces of the pellets or briquettes. Such fragmentary material is also regarded as DRI Fines.

NOTE:

For the purpose of the remainder of this document, the term “DRI Fines” will be used to define DRI Fines developed during the production process, and/or Fines yielded as the result of the handling of DRI product after the Hot Moulding Process (HBI Fines) or Cold Moulding Process.

3. Bulk Cargo (BC) Code – Code of Safe Practice for Solid Bulk Cargoes, 2001:3.1. The BC Code (2001) has two entries for DRI as follows:

3.2. Direct Reduced Iron, DRI, such as lumps, pellets and cold moulded briquettes. (BC No. 015)

Lumps and pellets average particle size 6 mm to 25 mm, with up to 5% Fines, under 4 mm. Cold moulded briquette approximate dimensions 35 mm to 40 mm.

It states that DRI may react with water and air to produce hydrogen and heat. The heat produced may cause ignition. Oxygen in enclosed spaces may be depleted.

A number of special requirements are listed in the Code for the shipment of this product.

3.3. Direct Reduced Iron, Briquettes, hot-moulded. (BC No. 016)

The material dimensions are approximately 90 mm to 130 mm long, 80 mm to 100 mm wide and 20 mm to 50 mm wide, with up to 5% Fines, under 4 mm.

This product is also known as HBI. It is stated that the material may slowly evolve hydrogen after contact with water. Temporary self heating of about 30° C may be expected after material handling in bulk.

A number of special requirements are listed in the Code. These requirements are less severe than BC Code No. 015 product.


5

3.4. One precaution commonly listed in the Code under both entries states, “A competent person recognized by the National Administration of the country of shipment should certify to the ship’s Master that the DRI, at the time of loading, is suitable for shipment.”

4. Existing Procedures4.1. Some ship managers involved in the successful transport of DRI Fines have developed internal procedures and instructions, in addition to those provided in the BC Code, to accommodate this cargo. Among others, the procedures contain instructions to:

Determine the cargo quality and condition, and when found necessary, to improve the cargo conditions prior to loading.

Prepare the ship prior to loading in order to properly and safely carry this cargo, including addition of facilities to measure temperatures, humidity and hydrogen levels.

Measure the hydrogen, oxygen and moisture contents of the hold and cargo every six (6) hours, during transport.

Install additional water driven fans, to enable proper ventilation of the hold.

Instruct the Officers and crew on the material characteristics and proper handling during transport.

In some cases, inert the holds, in which case only a ship equipped for this process should be utilized.

5. Ventilation/Inert Gas5.1. Due to the nature of DRI Fines with regard to the potential for the evolution of hydrogen gas during oxidation when wet, and given the very wide range between flammability limits of hydrogen (which is readily ignitable in concentrations above 4% in air), special consideration should be given to impeding the development of an explosive atmosphere when developing procedures for the transport of such cargo. Two solutions involve inerting the cargo hold after loading or ventilation.

5.2. DRI Fines are defined in the BC Code as having a diameter < 4 mm. It is understood that cargoes recently involved in explosions or near misses were comprised of particles that were significantly less than 1 mm in diameter. However, DRI Fines with a dimension of 4-6 mm are not defined by the BC Code (noting that particles 6 mm or larger are defined as


lumps or pellets), and are somewhat more prone to hazards such as self heating to extremely high temperatures (thermal runaway). Forced ventilation may contribute to or exacerbate such conditions. Therefore, DRI Fines with a dimension of 4-6 mm should be transported in an inerted condition. However, application of inert gas in a hold loaded with DRI Fines has a series of problems, including:

Hatch covers and vents installed on conventional bulkers are designed to keep weather/water out of the cargo hold, not to keep inert gas in. Therefore, it is difficult to achieve a hermetic seal with conventional bulk carrier hatch covers.

Hydrogen will mix with the atmosphere in the cargo hold; this causes a hazardous condition when the mixture has a concentration of hydrogen over 4%.

The inert gas cannot be applied in a conventional bulker as the cargo is loaded (open hatch), and can only be applied after the cargo has been loaded.

5.3. A better solution would be to transport DRI Fines greater than 4 mm in size in specialized ships, such as in the cargo holds of Ore/Bulk/Oil (OBO) carriers, which can be properly closed off and which have their own inert gas installation. Otherwise, these particular DRI Fines should not be transported in ships, but reprocessed and transformed into briquettes.

5.4. Forced ventilation of the cargo holds is another solution, provided the fan drives are intrinsically safe, e.g. water driven. However, ventilation is only suitable for DRI Fines considerably smaller in diameter than 4 mm.

6. Vessel Movements Before Loading6.1. The YTHAN loaded a cargo of dolomite in Lower Cove, Newfoundland, Canada in January 2004. This cargo was discharged in Matanzas, Orinoco River, Venezuela. On completion of discharging the dolomite, all holds were swept and washed with fresh river water. The bilges were dry with burlap covers placed over the bilge well covers.

6.2. The vessel tendered a notice of readiness on 17 February 2004, while at anchor at Palua Roads, mile 178, Orinoco River, Venezuela. The vessel berthed on 20 February 2004 at Palua pier, Copal Berth for loading.

•

•

•

•

•

•

•

•

•


6

7. Vessel Cargo7.1. The cargo, as per bill of lading, was Metallic HBI Fines, a total of 33,760 metric tons (mt) loaded in all five holds. (See Fig. 2)

7.2. The same cargo was also described with alternate names in other documents as follows:

Mate’s Receipts Orinoco Iron Remet FinesVessel’s draft survey report Remet Fines (HBI)Notice of readiness issued by the vessel

Orinoco Remet Fines in Bulk

Barge draft survey report HBI FinesThese alternate names do not adequately identify the cargo.

7.3. The correspondence between the owners and charterers is quoted as follows:

“Chrts have declared the option to load HBI and following has been agreed: All previous voyages to and including the voyage with HBI, Orinoco to the Far East to be devoid of any kind of claims what so ever. (Understood that if cargo starts to heat up in transit Orinoco to the Far East, Master will slow the vessel down.) (explanation of slowing down is that, in heavy weather the cargo begins to heat up)”

“Charterers to pay owners USD 75,000 lump sum net for the option to carry HBI/HBI fines. (HBI or HBI fines are the same product.)”

7.4. The vessel loaded 27,610 mt of cargo in all holds along side the Palua pier on 20 February 2004 through 22 February 2004. On 23 and 24 February 2004, the vessel loaded 6,150 mt of cargo in No.1, No. 3 and No. 5 holds, out of barges, at anchorage at mile 044, Orinoco River.

7.5. None of the crew observed the cargo storage area ashore. Additionally, members of the crew observed temperature readings being taken of the cargo in the ship holds by shore side personnel twice a day. The shore side personnel utilized infrared thermometers pointed at the surface of the cargo. The temperatures recorded from 20 to 22 February 2004 ranged from 31° C to 53° C. The results of the temperature readings


Fig. 2 - Photograph of cargo loaded onto YTHAN (DRI Fines)

Fig. 3 - Loading cargo onto YTHAN while alongside Palua Pier

Fig. 4 - Loading cargo onto YTHAN while alongside Palua Pier


7The Republic of the Marshall Islands

taken by the shore side personnel were not made known to the crew. Furthermore, there is no evidence of temperature readings taken from inside the loaded cargo piles.

7.6. There were no periods of rain during the loading of the vessel. During loading operations, when the cargo dropped off the conveyor belt into the hold, it was reported that a white condensate and a black dust could be seen coming from the cargo. It was also observed from photographs taken by the ship’s agent during loading operations that the cargo looked like a black powder. (See Fig. 3 and Fig. 4) Other photographs taken by the ship’s agent showed water in the barges used for lightering cargo at the anchorage prior to loading. (See Fig. 5 and Fig. 6) The density of the cargo, as reported to the vessel’s Chief Officer, was .42 cubic m/mt. Each hold was about 20% full.

7.7. After loading, the hatch coaming tables were swept and cleaned with compressed air. The hatches were then closed and swept, the main deck was swept and washed down with fresh river water, and the superstructure was also washed down with fresh river water. Finally, the vessel was fully de-ballasted.

7.8. A letter faxed to the vessel’s Master from Orinoco Iron c.a. dated 16 February 2004 states:

“Orinoco Iron Remet Fines, to be loaded on your vessel are safe to transport without the use of Inert Gas or other special precautions.”

7.9. A procedure for the handling and the transport of the Orinoco Iron Remet Fines was provided. The “Guide for Maritime Transportation of Orinoco Iron Remet” states, among other instructions:

“Orinoco Iron Remet is a sub-product from the production of HBI briquettes, which are produced by removing oxygen from iron ore lumps and pellets by reaction with hydrogen and carbon monoxide at high temperatures.”

“Remet loading should only be done in dry weather conditions. If it begins to rain during loading, the operations must be halted until the rain stops. Loading of wet cargo is permitted.”

“After washing is complete, reopen hatches for 1 or 2 days during dry weather to allow dissipation of

Fig. 5 - Loading cargo into barges alongside Palua Pier before loading onto YTHAN at anchorage

Fig. 6 - Barges alongside Palua Pier before loading onto YTHAN at anchorage. *Note water at bottom of empty barge


8

water from the Remet. If damp metallic HBI Fines were loaded, they will warm up to about 60° C (140° F) and will produce steam and a small amount of hydrogen as they dry. The hatches should be open to allow the hydrogen and water vapor to escape, since the gases are lighter than air and will rise. This is a normal condition will last 1-2 days until the metallic Remet is dry.”

7.10. As stated during interviews, the Chief Officer was aware of potential problems associated with the heating of the cargo, but did not have an understanding of the potential dangers of hydrogen production by the cargo during oxidation when wet. He found no reference of HBI Fines in Thomas’ Stowage, and subsequently looked up the IMO BC Code. References were found for BC Code No. 16, Direct Reduced Iron, DRI Briquettes, Hot-Moulded, but not for HBI Fines. No further attempts were made to research the properties of HBI Fines, and full reliance was placed on the information supplied at the terminal.

7.11. No certification of this cargo was provided to the vessel as required by the IMO BC Code which states:

“A competent person recognized by the National Administration of the country of shipment should certify to the ship’s Master that the DRI, at the time of loading, is suitable for shipment.”

8. Voyage Conditions Prior to 28 February 20048.1. On 25 February 2004 at 06:12 hours, the YTHAN weighed anchor bound for passage to China via the Panama Canal where bunkers and stores were to be taken. The hatch covers were partially opened to ventilate the holds during the down river transit. However, at 17:00 hours the hatch covers were closed for the night. Each hold had two vents with no ventilation trunking in the holds; one vent at the forward port corner of the hold, and the other one at the aft starboard corner. The vessel did not have forced ventilation in the holds.

8.2. 26 February 2004: The weather that day was good, with winds at Force 3-4, moderate seas, and no spray. In the statement from the Chief Officer, it was noticed that he observed from each of the five holds a wet path of water on the otherwise dry deck. The water was draining from the hatch coaming table drains, located at the aft end of each hatch. The drains allow condensed water accumulated on the inside of the hatch compression bar


and hatch cover gasket to drain to the open deck. The hatch covers were opened and the undersides observed to be covered in heavy sweat/condensation. Bilges were sounded and found to be dry. The hold temperatures were recorded in the morning at 55-60° C. The temperatures were monitored and recorded using an electronic probe lowered down open pipes installed in the holds along the transverse bulkheads and terminating about 1.5 meters above the tanks. No gas concentration readings were taken, and the multi-gas meter onboard did not measure hydrogen concentrations. The conditions were reported to the Master, but advice was not sought regarding the condition of the cargo. In the evening before closing the hatch covers, temperatures were recorded at 30-34° C. The hatch covers were closed at 17:30 hours with alternating cleats applied.

8.3. 27 February 2004: Similar weather and cargo conditions were observed as per the previous day. Water once again was observed coming from the hatch coaming table drains, and the morning temperatures of the cargo holds ranged from 50-60° C. The hatch covers were again partially opened, and later closed at 17:00 hours, this time with all cleats applied. Temperatures recorded prior to closing the hatches ranged from 30-34° C. The Chief Officer stated he was not worried about the observations regarding the cargo because the temperatures did drop during the day when the hatch covers were partially opened, thinking the temperatures were under control. It was also expressed that the ventilators may not have been effective due to the fact that temperatures were rising overnight when the hatch covers were closed.

Fig. 7 - View of portside forward and No. 2 and No. 3 holds.


9

9. Conditions and Events on 28 February 20049.1. Similar conditions as per the previous two days were observed. Drainage water from the hatch coaming table drains were again observed, and temperature readings from the cargo holds ranged from 56-60° C. The planned routine for that day would be to first open up the hatch covers of No. 2 hold to ventilate the hold and grease the hatch cover wheels and hinges (all other hold hatch cover wheels and hinges were greased the previous day), then to remove the cleats off the remainder of the hatch covers and open them partially to ventilate the cargo. The deck crew were split into three groups to accomplish the tasks.

9.2. At 09:50 hours, the crew were working on removing the various hatch cover cleats. No. 2 hold was open, and all of the cleats had been removed from the hatches of No. 1 and No. 4 holds. At this time, the first explosion took place in No. 1 hold. Witnesses described hearing a loud “OOMPH” noise of approximately 1-2 seconds in duration. The force of the explosion lifted the hatch covers off No. 1 hold about 2.5 m in the air, and dropped them back down onto the hatch coaming slightly twisted/misaligned.

9.3. No smoke was observed at No. 1 hold. However, the mooring line, spooled onto one of the windlass drums forward of No. 1 hold, was melted and on fire. At this time, the majority of the crew working on deck were moving forward to investigate and respond to the fire on the mooring line, along with the Master and the Chief Engineer.

9.4. At 09:53 hours, No. 5 hold exploded violently. A similar noise heard during the explosion at No. 1 hold was reported during the explosion of No. 5 hold, except significantly louder by a magnitude of 3-4, and the hatch covers were seen being completely blown off the hold into the air, with the distorted hatch panels and wreckage landing on the port side of the main deck. Afterwards, a large amount of light grey smoke was observed in front of the deckhouse, obscuring most of the accommodation and bridge. Extensive structural damage was later observed in the vicinity of No. 5 hold.

9.5. Several seconds later, No. 3 hold exploded. As a result of this explosion, the upper housing of the back of Crane #3 cab (located between No. 3 and No. 4 holds) was hit by one of the No. 3 hold hatch covers, causing hydraulic oil to spill from


the damaged cab. In addition, the boom was torn off Crane #2 (located between No. 2 and No. 3 holds), also the result of the explosion in No. 3 hold. Smoke was observed at No. 3 hold after the explosion.

9.6. Two crew members were adjacent to No. 3 hold when it exploded. Both men were thrown to the deck by the force of the explosion. One man remained conscious, and recalls seeing flame from the explosion of No. 3 hold and feeling intense heat. He sustained burns to exposed portions of his face and arms. The other crewman briefly lost consciousness, but recovered a short time later, and managed to reach safety. He sustained extensive 3rd degree burns to his body.

9.7. No. 4 hold exploded shortly after No. 3 hold. The exact time between explosions could not be confirmed but approximated to be several seconds. As with No. 5 and No. 3 holds, the explosion of No. 4 hold violently blew the hatch covers into the air, subsequently tearing off the boom of Crane #3.

9.8. The Master was reportedly struck by one of the hatch cover sections as it fell to the deck on the starboard side of No. 3 hold. It could not be confirmed if the Master was struck by a hatch cover section from the No. 3 hold or the No. 4 hold explosion. Regardless, upon receiving word, the Chief Officer and Chief Engineer checked the Master’s condition and verified no signs of life.

9.9. Thereafter, the remainder of the crew on deck during the explosions congregated on the port side of No. 1 and No. 2 holds, and proceeded aft to the boat deck via the port side. The starboard side main deck was impassable due to the fallen hatch covers, crane booms, and wreckage. They met with the remainder of the crew assembling on the boat deck. Dense black smoke was now noted emanating from No. 5 hold and the accommodations.

9.10. At this point, there were several observations made regarding the damages resulting from the explosion of No. 5 hold. The cab and boom of Crane #4 (located between No. 4 and No. 5 holds) was observed to be completely destroyed – only the pedestal remained. In addition, the Christmas Tree had fallen down onto the main deck at the port side of No. 5 hold. The deck plating on the port side of No. 5 hold was seen curved/bulged up to the height of the hatch coaming tables. Furthermore, ingress of seawater into No. 5 hold from both


10

port and starboard sides was observed from witnesses on the boat deck. There was an additional account of the deck plating at the port side of No. 3 hold being buckled as well.

9.11. The Third Officer, on the bridge at the time of the explosions, was found unconscious, lying over the chart table, bleeding from the head. He had been seriously injured and had to be carried down to the boat deck by two crew members. Injuries were mainly to the back and the back of the head. At the time of the interviews, the Third Officer had no recollection of events.

9.12. Bridge windows were observed blown inwards, and the ceiling had fallen down, with debris everywhere. The GMDSS was not operational. No distress message was sent as the telecommunications equipment and GMDSS on the port outboard bulkhead had been destroyed. The EPIRB was unable to be located at the Monkey Island.

9.13. The crew assembled at the port boat deck, due to the excessive and dense smoke on the starboard side from the accommodations. A head count was taken, and it revealed that five crew members were missing, in addition to the Master. It was quickly ascertained that the five missing crew members were from the engine department, who were occupied with duties in the engine room at the time of the explosions. A search was impossible due to worsening conditions in the accommodation. The engine room was not accessible due to missing ladders, and smoke, sparks and fire were reported. Several witnesses inside the accommodations at the time when No. 5 hold exploded recalled the vessel immediately blacked out, and felt propulsion stop. In one statement, a 1 m hole, with the edges bent upwards, was seen in the deck inside the accommodations, just above the engine room.

10. Abandon Ship10.1. The crew abandoned ship using the port lifeboat, which was not motor propelled. Excessive smoke prevented the use of the starboard lifeboat, which was motor driven. Four crew members remained on the ship to launch the lifeboat, and then jumped into the sea, one of which was able to swim to the lifeboat. The other three crew members were not able to reach the lifeboat. The lifeboat was difficult to manage by the crew with oars, and was unable to counter the rate of drift. The


three crew members remained in the water for 12 hours, arms interlocked, until picked up by a passing vessel.

10.2. At approximately 10:30 hours, on 28 February 2004, the YTHAN sank in approximately 1200 meters of water, 30 miles North of Santa Marta, Colombia. Witnesses report that the ship sank on an even keel, and the accommodation went down vertically as well.

10.3. At about 16:00 hours, the crew members in the lifeboat spotted a vessel, fired a rocket parachute signal, and activated the radar transponder (SART). The name of the vessel was the CALAPALMA, and by approximately 19:30 hours, all 18 crewmembers in the lifeboat were safely transferred onboard without incident. Also, by 22:30 hours, the three crew members adrift in the water were rescued by the SEABOARD EXPLORER II. All 21 survivors were landed at Santa Marta, Colombia that night.

11. Fatalities11.1. The Master was struck and killed by one of the hatch cover panel sections (presumably from No. 3 hold) as it fell to the deck.

11.2. Five (5) members of the Engineering staff were reported missing, and are presumed to have been killed when the explosion in No. 5 hold extended into the engine room.



11

CONCLUSIONS

1. The cargo of the YTHAN consisted of DRI Fines. Due to the ambiguity of the trade names on the various associated documents, the physical nature of the DRI Fines could not be determined. While it is apparent that the cargo was made up of Fines, it is not clear if the cargo was comprised of Fines resulting from the DRI production process, Fines generated from the handling of DRI (such as chips or scrapings from HBI or cold moulded briquettes), or a combination.

2. Multiple terms were used for the cargo which created confusion and possibly covered a range of products mixed into the cargo.

3. The condition of the atmosphere in the cargo holds of the YTHAN, on the morning of 28 February 2004, contributed to the fatal explosions of No. 1, No. 3, No. 4, and No. 5 holds. No. 2 hold did not explode because it was the only hold where the hatch covers had been opened. The exact ignition source for the explosions was not determined.

4. As reported in the crew statements, the vessel sank on an even keel. Together with the reported ingress of water in No. 5 hold, it is presumed that additionally the shell plating at No. 1 hold, and possibly in No. 3 hold, were also breached leaving one or both holds open to the sea.

5. The BC Code does not have an entry for Fines materials. The entries in the BC Code No. 015 and No. 016 limit Fines to a maximum of 5% of the cargo.

6. The BC Code requires that “a competent person recognized by the National Administration of the country of shipment should certify to the ship’s Master that the DRI, at the time of loading, is suitable for shipment.” Shippers should certify that the material conforms with the requirement of this Code. The Master did not receive this required certification prior to sailing.

7. The instructions to the vessel’s Master from the shippers stated: “Orinoco Iron Remet Fines, to be loaded on your vessel are safe to transport without the use of Inert Gas or other precautions.” These instructions do not adequately address precautions for the carriage of DRI Fines.

8. The procedure available for the handling and transport of Orinoco Iron Remet Fines is labeled: “Guide for Maritime Transportation of Orinoco Iron Remet” (Guide). Fines are not included in the title of the Guide.

9. The Guide details that the hatches have to be opened for 1 or 2 days to allow dissipation of water from the Remet. The crew only opened up the hatch covers during the day and kept them closed during the night, preventing the holds from being


ventilated which allowed the temperature in the cargo to rise to 50-60° C. This situation was reported to the Master, but further advice regarding the condition of the cargo was not sought.

10. As per correspondence quoted between the vessel owners and the charterers, the potential hazards associated with the carriage of the subject cargo were not satisfactorily conveyed:

“Chrts have declared the option to load HBI and following has been agreed: All previous voyages to and including the voyage with HBI, Orinoco to the Far East to be devoid of any kind of claims what so ever. (Understood that if cargo starts to heat up in transit Orinoco to the Far East, Master will slow the vessel down.) (explanation of slowing down is that, in heavy weather the cargo begins to heat up)”

“Charterers to pay owners USD 75,000 lump sum net for the option to carry HBI/HBI fines. (HBI or HBI fines are the same product)”

11. The vessel operators did not seem to have suitable representation at the loading terminal to ensure: 1) that the cargo was stable and safe prior to loading, 2) that the cargo was loaded onto the vessel and stowed in proper fashion, and 3) that the ship’s Officers possessed adequate competence and knowledge of the requirements to carry the cargo safely. The operators did not have their own procedures in place for the safe carriage of DRI cargoes, and were dependent on the instructions provided by the shippers for information regarding carriage and safety requirements for the cargo.

12. A wet path was observed on the upper deck caused by the water draining during the night from the hatch coaming table drains, as a result of condensation in the hold, indicating that the cargo was wet. There is no indication that the moisture content of the cargo was measured prior to loading.

13. It is unclear precisely how the hold/cargo stow temperatures were measured, but witness statements suggest that the measured temperatures represented cargo surface temperatures. The temperatures within the cargo/stow are unknown.

14. The vessel’s Chief Officer had no previous experience with the carriage of DRI cargoes. While the Chief Officer was aware of the heating problems of the cargo, there was not an understanding of the potential dangers associated with hydrogen production by the cargo during oxidation when wet. The Chief Officer consulted the BC Code for reference to DRI Fines and HBI Fines. However, there were no clear references in the Code for the handling of Fines exclusively. No further efforts were made to ascertain the properties of the cargo, and full reliance was placed on the information supplied to the ship by the loading terminal.

12

Recommendations1. DRI materials and their hazards should be properly defined. The terminology of the particular DRI product should be clear and consistent throughout all documentation, and be representative of same product. Likewise, a physical description of the material should be obtained, as specific hazards are determined based on the physical properties of the material.

2. Additional entries should be made in the BC/IMDG Codes for DRI Fines which are subject to marine transport. These entries should list the dangers and precautions to be taken for proper marine handling and transport, including the suitability of the vessel to carry such cargoes. This is being addressed at the 11th session of the Sub-Committee on Dangerous Goods, Solid Cargoes and Containers (DSC), as paper No. 4 (DSC 11/4), subject to extensive comment as per DSC 11/4/[X].

3. Vessel’s Master and crew should be properly informed and instructed on the handling of these products during marine transport, and be made aware of all associated hazards. In addition to the competent person recognized by the National Administration of the country of shipment to certify the suitability to load the DRI, the vessel’s owners and managers should be involved in the loading and transportation process. The certification from the shippers should be double checked, and records should be verified ascertaining the condition of the cargo prior to loading. The cargo should be stabilized as far as possible prior to loading, and the distinct Fines should be loaded in separate holds from other cargo. Also, consideration should be given to placing a representative onboard the vessel during the loading and voyage of the vessel/cargo.

4. In reference to the handling of DRI products during transport, special consideration should be given with regard to the potential evolution of hydrogen gas during oxidation. Due to the very wide range between flammability limits of hydrogen, a strict, no smoking policy should be enforced, and all other ignition sources should be prohibited on the deck and adjacent to cargo hold areas during the carriage of DRI products.

4.1. Furthermore, for DRI Fines with a diameter greater than 4 mm, transportation should be undertaken in an inerted condition, in specialized ships in which proper inert gas installations have been provided (such as OBOs).

4.2. For DRI Fines with a diameter less than 4 mm, proper forced ventilation is more appropriate, provided the fan drives


are intrinsically safe (i.e. water driven). Care should be taken that the inlet of the ventilation system is placed in the correct positions in order to adequately remove the hydrogen mixture.

4.3. The best overall solution would be not to transport DRI Fines in ships, but to reprocess them and transform them into briquettes.

5. It is recommended that a copy of this Marine Investigation Report be forwarded to the Secretary-General of the International Maritime Organization and other interested parties.

Acknowledgements1. The Republic of the Marshall Islands commends the valiant efforts by the Masters, Officers and Crew of the CALAPALMA and the SEABOARD EXPLORER II who, in the highest maritime tradition, diverted their vessels and heroically rescued the surviving crew of the YTHAN. 2. Dr. Alan Mitcheson BSc, PhD, CEng, FIMechE, MAIChE, FIDiagE


13

APPENDIX 1DIRECT REDUCED IRON (DRI)

Direct Reduction ProcessDirect Reduced Iron (DRI) is iron produced from iron ore (iron oxide) using the direct reduction process. The definition of reduction, as applied in this process, is to bring a substance to a metallic state by removing the non-metallic elements. In this case, iron oxide is reduced to iron, with oxygen as the primary non-metallic element removed.

Naturally occurring iron is found mainly as a compound (iron ore) from which the iron must be extracted. Typical iron ores are magnetite, hematite, siderite, etc. The compounds consist mainly of oxides (iron oxide), carbonates or sulphides, and are often combined with other elements. Iron oxide is iron chemically bonded with oxygen, which is the end product of pure iron when exposed to air – more commonly know as rust.

The extraction of iron from iron ore is accomplished by breaking the chemical bonds between the iron and oxygen atoms in the iron oxide molecules, and chemically removing the oxygen, leaving behind the iron in a free metallic form. The chemical process of removing the oxygen is termed “reduction,” as per the definition stated earlier. There are two basic variations to this process – Indirect Reduction and Direct Reduction.

Indirect Reduction is the traditional process for producing iron, accomplished in a blast furnace, whereby coke (carbon) is added to a quantity of iron ore, and extremely hot air is injected. As the coke burns in the hot air, the carbon combines with the oxygen in the air forming carbon monoxide. The carbon monoxide then reacts with the iron oxide, subsequently forming carbon dioxide, leaving behind pure iron (reduction). In this process, the iron is produced in a molten state due to the very high temperatures required to burn the coke and produce the reaction. The reason this is termed “Indirect” reduction, is because the gas used for the reduction process is indirectly generated – the coke is first converted into the reducing gas, which is then utilized in the reduction process.

Direct Reduction accomplishes this process without a requisite phase change (solid to liquid). The iron is yielded from the iron ore without melting, thereby significantly reducing the amount of energy required for the process. In the direct


reduction process, the iron ore is typically passed as pellets or granules (which are referred to as ‘fines’) through a reducing gas (mainly hydrogen with some carbon dioxide) and energy is inputted in the form of heat, but at a temperature lower than the melting point of iron. The output is iron (DRI) and carbon dioxide water vapor.

Result (Output) of the Direct Reduction ProcessThe direct reduction process yields DRI at temperatures below the melting point of the iron. As a result, the material is not produced as a liquid or a fused mass. Rather, the DRI has the same physical geometry as the iron ore introduced into the process, with a porous structure created by the individual iron particles bonded together. Therefore, the direct reduction process produces DRI as lumps, pellets or fines. Due to the porosity of the DRI, the material has a large relative surface area. The extent of porosity also has an influence on the density of the DRI, which on average is approximately 2-3 g/cc. By comparison, solid iron has a density of approximately 7.7 g/cc.

The quality of DRI outputted from the direct reduction process is expressed by the term “degree of metallisation.” The direct reduction process is not perfectly efficient. As a result, all of the non-metallic elements (such as oxygen) cannot be completely removed from the iron ore, and the final product will still contain some remnants of oxygen and other elements. The “degree of metallization” is a term used to describe the ratio of pure (metallic) iron vs. the total mass of DRI (Total Iron). As the ratio approaches 100, the DRI reaches purity.

Lastly, to facilitate storage and handling for shipping, DRI produced as lumps, pellets or fines can also be formed into


Fig. 1

surface area of the material exposed to the air/water. This is critical with regards to DRI due to the physical characteristics of the material as described earlier - DRI is very porous, of which greater than 50% of the material by volume is air. The air fills the pores and spaces between the individual grains of DRI, thereby allowing the material inside the pellet or lump to oxidize, not just the material on the exterior. For a given volume of DRI, the relative surface area is inversely proportional to the linear dimensions of the individual particles. This means that the smaller the grain particle size within the DRI pellet, the greater the relative surface area of the DRI material, thus, a greater surface area that can be exposed to air/water, giving a greater tendency for the material to re-oxidize.

Heat TransferAlso, due to the geometry of the grains within a pellet or lump of DRI, the grains have a very light contact with each other. The structure is very similar to the iron ore / iron oxide from which the DRI is derived. The rate at which heat can be transferred by conduction (from one object in direct contact with another) is dependent on the differential in the temperature between the two objects, and the total contact area between the two objects. As the contact area between grains within the structure of DRI is minimal, it has a very low specific thermal conductivity when compared to other metals. Therefore, DRI is a very poor conductor of heat.


14

briquettes, using a binder material to hold the brick together. DRI formed by this subsequent process is termed DRI, Cold Molded Briquettes. The phrase ‘Cold Molded’ is utilized because the briquettes are formed after the DRI has cooled at a temperature below 650° C. However, the physical properties of the DRI when formed into cold molded briquettes remain unaltered, and will react in the same manner as DRI lumps, pellets or fines.

MATERIAL HAZARDS

Re-OxidationThe direct production process is reversible. Iron, being highly reactive, will naturally oxidize and form rust when exposed to air and moisture. The oxygen in the air will react with DRI when exposed, and cause the material to re-oxidize. The oxidation process also occurs with a release of energy in the form of heat.

The amount of energy released during oxidation is directly proportional to the rate of oxidation. Normally, when DRI is exposed to dry air, without moisture, the oxidation occurs slowly. However, the rate of re-oxidation can be affected by a number of conditions. An increase in the rate of oxidation will be observed when the DRI is also exposed to water and becomes wet. Furthermore, when the DRI becomes wet with salt water, the oxidation rate is profoundly accelerated by the introduction of salt. Even minor concentrations of salt will have an effect. This is because the oxidation process is electrochemical in nature, and the addition of salts in the water act as a catalyst, increasing the reaction rate of the DRI.

Thermal RunawayAdditionally, an increase in the temperature of the oxidizing material will also increase the rate of oxidation. The rate at which a chemical reaction progresses is exponentially dependent on the temperature of the reacting material. For every incremental increase in the temperature of the reacting material, the reaction will experience an exponential increase in the rate of the reaction. It is possible that at a certain temperature, the rate of the reaction is such that the heat is being generated by the reaction faster than it can be dissipated, which subsequently causes the reaction to further accelerate and produce more heat. The uncontrolled progression of this phenomenon is known as “thermal runaway.”

Geometry (Surface Area)Another critical issue with DRI in the context of re-oxidation, is that the degree of re-oxidation will be influenced by the


Fig. 2

Large Contact Area

Small Contact Area


15

Furthermore, the DRI lumps and pellets themselves, being generally spherical in shape, will also have a generally small contact area from pellet to pellet. This has a compounding effect on the overall thermal conductivity within the entire pile of material. The low contact area between grains within the DRI pellet, combined with the similarly low contact area between pellets themselves greatly retards the heat transfer characteristics of the pile of DRI. (See Fig. 3)

CombustionShould a pocket of DRI within a pile begin to oxidize at a rate faster than the heat can be dissipated, it is the combination of these two characteristics of DRI (the large surface area available for re-oxidation, and the low thermal conductivity) that allow the heat generated by the oxidation of material within a pile to gather and build up, forming a ‘hot spot’. As increased temperature accelerates the re-oxidation rate, with a suitable covering of overlaying material insulating the heat and not allowing it to dissipate, conditions may be present whereby the temperature of the DRI at the hot spot can build up to the point of reaching thermal runaway. The temperature of the oxidizing material will then progress until it reaches ignition temperature – at sufficient temperature, the DRI will actually burn with oxygen present. (See Fig. 4)

The hot spot can propagate slowly within the pile, but location is the key factor. DRI oxidizing on the surface of the pile will usually not overheat, as the heat can naturally dissipate adequately enough to prevent the temperature from building up. For this reason, rain on the surface of the material will not cause overheating. However, once the surface of the pile becomes wet, no additional DRI should be added on top, nor should the pile be moved or stirred, so that the wet DRI becomes covered by a sufficient layer of material to insulate the heat generated during oxidation.

Hydrogen LiberationIn the presence of moisture, there is another undesirable effect of the re-oxidation of DRI. As the water molecules loose the oxygen atoms to the iron during oxidation, the hydrogen atoms are left behind and combine to evolve hydrogen gas. The principle hazard regarding hydrogen is that when mixed with air, it has a very wide range of flammability limits from about 4% to 75% by volume concentrations of hydrogen in air. Furthermore, it is understood that hydrogen is very susceptible to ignition. These two facts, combined with the potential for DRI to self heat, can readily lead to the development of a hazardous situation in the right conditions.


Minimal Contact Area

Fig. 3

Fig. 4


16

It should also be understood that relatively small quantities of DRI are capable of producing considerable quantities of hydrogen gas if the conditions are right, and if the entire quantity of DRI is permitted to oxidize to completion. Also, the rate at which the hydrogen gas is evolved is dependent upon the surface area of the oxidizing material available, the amount of moisture present, the amount of oxygen in that water and the temperature of the reacting material.

PASSIVATION METHODSInitially, production of DRI was limited, and the material was utilized by mills adjacent to the DRI production facilities. However, due to changes in market demands, and evolution of the steel industry (such as emergence of mini-mills), global shipment of DRI has been growing. For this reason, the safe storage and handling of DRI has become very important.

For shipping, there are several methods that can be employed to attempt to ‘stabilize’ the DRI or inhibit the oxidation process, making it less reactive and safer to transport. The three most common passivation methods are:

Ageing / Controlled OxidationBy exposing fresh DRI to air for a prescribed period of time, such as 72 hours (ageing), or by exposing the DRI to controlled amounts of oxygen at regulated temperatures and duration after the production process (controlled oxidation), a thin layer of oxidation will be allowed to form on the surface of the DRI. This thin layer of oxide will retard further oxidation, but not prevent it entirely. Also, a drawback is that this method can reduce the degree of metallisation of the DRI, and have a negative impact on its value. Also, as the oxide layer is not impervious to water, the possibility of self-heating or hydrogen formation due to further oxidation cannot be prevented.

Inhibitor ApplicationBy coating the DRI with a chemical coating, such as silicon derivatives, the applied coating will act in a similar manner as the layer of oxide, inhibiting re-oxidation. However, this process is also not completely effective due to limitations with ensuring effective coverage of the material and limited durability of the coating. Additionally, this method cannot guarantee total protection against self-heating or hydrogen evolution. (See Fig. 5)

Hot BriquettingBy modification of the morphology of the DRI material itself, the surface area of the material exposed to oxidation can be


effectively reduced (by reducing the porosity of the material), and the thermal conductivity will also be increased. This process is known as hot briquetting. The DRI is taken from the production process while still hot, or heated to a temperature above 650° C, and pressed in a roll forging process into briquettes. This forging closes the pores within the material, and increases the density to greater than 5 g/cc. The reduced surface area will have an effect of minimizing the amount of oxidation, and the increased thermal conductivity characteristics will more easily transfer and dissipate any heat generated. DRI formed by this process is termed DRI, Hot Moulded Briquettes, or Hot Briquetted Iron (HBI), and is a very successful method for reducing the reactivity of the DRI.

Before After

Fig. 5

80973876 explosion and loss of the bulk carrier ythan

Documents

marshall islands maritime

maritime regulations

marine casualty

marine investigations

criminal liability

series of explosions

proximate cause

cause of difficulties