LOUISIANA OFFICE FLORIDA OFFICE 3501 HOLIDAY DRIVE, SUITE #203-A OCEAN-OIL P. O. BOX 47188 NEW ORLEANS, LA 70114 ST. PETERSBURG, FL 33743-7188 CELL (24 HOURS): (504) 367-4072 EXPERT WITNESS, INC. 566 VILLA GRANDE AVE. S. FAX: (504) 367-3790 ST. PETERSBURG, FL 33707 NAVAL ARCHITECTS and MARINE ENGINEERS PH: (727) 347-2556 Email: Oceanoilpazos@cs.com MARINE CONSULTANTS FAX: (727) 343-9717 www.siterrific.com/pazos CELL (24 HRS): (727) 415-7192 Email: Hectorpazos75597@aol.com ARTICLE #4 - GROUP 1: MARITIME DISASTER SHIFT OF CARGO RESULTED IN THE LOSS OF ALMOST THE ENTIRE CREW By: Hector V. Pazos, P.E., Naval Architect, Marine Engineer and Registered Mechanical Engineer (www.siterrific.com/pazos) Copyright 2007 All Rights Reserved Summary of the Accident: The M/V Salvador Allende, a large dry cargo carrier, capsized and sunk on or about December 9, 1994, in the Atlantic Ocean, with the loss of almost the entire crew (29 people were lost). Only two (2) members of the crew survived, the third engineer, Mr. Alexander Tarenov, and the second mate, Mr. Ivan Skiba. During the arbitration procedures in New York, these two officers, provided a review of the events preceding the casualty, which allowed the author of this article to conclude that the shift of the cargo due to improper loading procedures was the cause of the capsizing. Main Cause of Disaster: A cargo of rice was shipped in Flexible Intermediate Bulk Containers (FIBC), some of which may have been damaged by forklift operations during loading. The improper securing of the cargo combined with the rice becoming loose due to damaged FIBC was the main cause of the capsizing of the M/V Salvador Allende. The FIBC bulk bags are commonly woven from heavy duty, UV treated polypropylene resin or other flexible fabrics available in a range of fabric weights. When first manufactured in the late 1950 s or early 1960 s, the FIBC were constructed with heavy duty PVC coated nylon or polyester where the cut sheets are welded together to form the FIBC.
Tragic Note: A tragic note of this disaster was that when the captain gave the order to abandon ship, the first lifeboat was boarded primarily by the female crewmembers, which represented one third of the crew. Regretfully, during the boarding, a large wave hit the lifeboat before reaching the surface of the ocean, throwing everybody overboard and no one aboard the lifeboat could be rescued. a). Summary of the testimony of Mr. Tarenov: Mr. Tarenov, 37 years old, Ukrainian National, had an operating engineer diploma from the ODESSA Port Authority. He joined the M/V Salvador Allende on October 15, 1994 in ESTONIA, while the vessel was taking a cargo of metals and coal to carry to Baltimore, USA. He was responsible for the diesel generator, compressed air system, fuel system and other auxiliaries. He had two watches. From 4:00 a.m. to 8:00 a.m. and from 4:00 p.m. to 8:00 p.m. His stateroom was located on deck C, room No. 78. In describing the tanks of the M/V Salvador Allende, Mr. Tarenov indicated that tanks 1, 2, 3, 4, 9, and 10 were used only as ballast tanks. Tanks 7 and 8 could be used for fuel or ballast. During the last voyage of the Salvador Allende, tanks 7 and 8 were carrying fuel. Also, fuel was being carried in tanks 5, 6, 7, 8, 11, 12, 14, 28, 29, 30, 45, 46, 47, 54 and 55. Also small tanks: 31, 32, 33, 34 and 35 were either day tanks or settling tanks. From Baltimore, the vessel went to New Orleans, Louisiana, and then to Freeport, Texas, to load rice. After loading, the vessel left Freeport with the tanks full with water (although the testimony does not indicate if these tanks were pressed-up, that is, without free surface). The vessel left Freeport with approximately 720 tons of fuel on December 1, 1994, and no problems were noted with the hull, machinery and equipment. The daily fuel consumption was 24 to 26 tons, including fuel oil, diesel oil, etc. After using the fuel from tanks 28, 29 and 30, which may have contained approximately 75 tons, Mr. Tarenov started using from tanks 7 and 8, about December 8. From December 4 to December 8, he consumed 100% of tank #8 and partially tank #7. No transfer of fuel took place. The reason for using fuel from tank #8 is because about December 5 th, it was noticed that the vessel started to develop a port list of about 1 degree. The list was reported to the chief engineer on December 6 th and to the captain on December 8 th. Between December 5 th and 8 th, the weather conditions were worsening, and up to December 8 th, Mr. Tarenov was able to compensate the 1 degree list by consuming fuel from tank #8 only, on the port side. The vessel was making about 14 knots, turning about 112 to 114 RPM. No water was detected in the cargo holds on December 3 rd, 4 th and 7 th, but nothing was done to find the reason for the list to port. 2
On December 8 th, an attempt was made to level the vessel by ballasting. During the morning watch of Mr. Tarenov, the list to port was ½ of a degree. The weather worsened in the evening and the roll amplitude was about 4 degrees to 5 degrees. When Mr. Tarenov started his afternoon watch at 1600, the list was again 1 degree to port, so Mr. Tarenov informed the captain that the fuel was being used from tank #8 (port side), but that also there were 70 to 75 tons of fuel less on the port side than on the starboard side. This is when the captain ordered Mr. Tarenov to pump out 30 tons of ballast from ballast tank #10 to overboard, correcting the 1-degree port list. Mr. Tarenov went to bed about 10:30 p.m., but due to the rolling motions, he found it difficult to remain in bed. Nevertheless, he fell asleep, but woke up at about 11:30 p.m., because the vessel developed a strong list to port of about 30 degrees, while rolling substantially getting at times close to 0 degree, but never reaching 0 degree. When he got out of bed and looked thru the porthole, the vessel was lying on its side. He heard (thru the P.A. system) the captain ordering everybody to the ship s bridge with emergency equipment. While Mr. Tarenov was putting on clothes and a thermal jacket, the vessel took another strong list to port. Nevertheless, he managed to reach the bridge (without his pants on). The vessel settled at about 40 to 45 degrees list, while rolling back up to approximately 15 degrees, that is, a total amplitude of roll of 30 degrees. When he arrived to the bridge deck, he was able to see thru portholes in the navigator s room that part of the deck was underwater and also found that the vessel s heading had changed to port. About midnight, the main engine stopped, but was later restarted, and the vessel took another strong list to port getting to approximately 50 degrees. A discussion regarding pumping ballast into starboard tank took place, but all the ballast tanks on the starboard side were full. At this time an SOS was sent and the order to abandon ship was given. During the abandon ship activities, the lifeboats were torn off and the lifer rafts were driven away by wind and waves. The first lifeboat was hit by a large wave while being lowered and lost with all the people aboard the lifeboat (mostly women crewmembers). Shortly after 4:00 a.m., Canadian rescue planes found the vessel. About this time the list of the vessel reached 90 degrees and started sinking, submerging the stern at about 5:30 a.m. Mr. Tarenov eventually ended up in the water and saw the vessel for the last time, in an almost vertical position at about 6:30 a.m. about one hour after he hit the water. The generator was operating until the stern of the vessel began to sink. During her last voyage, the bilges of all the holds were pumped out periodically. To execute this task the corresponding valves have to be opened and closed by hand. These valves are reached thru a pipe tunnel located on the double bottom at the centerline of the vessel. The original design of the vessel included remote control valves, but this system was not functioning when the vessel left the shipyard. Until December 8 th, there were crewmembers handling the valves in 3
the pipe tunnel, but, after the vessel took a 40-degree list, no one would volunteer to get in the pipe tunnel. Although Mr. Tarenov testified that he is not an expert in cargo loading, he made the following statements regarding the cargo: There was no cargo in the hatch coaming space of hold no. 4 and the space between the top of the cargo and the main deck was about 70 centimeters. The cargo in no. 3 hold was approximately as high as in hold no. 4. The cargo of plastic material bags (FIBC) was lashed with polyester ropes. b). Summary of the testimony of Mr. Skiba: Mr. Skiba graduated in 1992 from the Naval Academy in the City of Odessa. Mr. Skiba joined the M/V Salvador Allende in July 1994, as second mate, while the vessel was being repaired. He had watched duty from noon to 1600 and from midnight to 4:00 a.m. He was responsible for loading, safety transportation of the cargo and unloading. While the vessel was at the shipyard, among other items, he inspected the rubber sealing of the hatch covers, which was being replaced, and the ventilation system, and accepted the vessel jointly with the ship s chief officer. He also inspected the weather doors of the mast houses and the cargo holds and found the wooden battens in good condition. Another item he found in good condition was the manhole covers, located at the double bottom ceiling or tank tops. During a trip prior to sailing to Freeport, Texas, the M/V Salvador Allende carried ore concentrate from Brazil, in big bags made of a material similar to the rice bags transported during her last trip, but more rigid or hefty, with an approximate weight of two tons. Upon arrival in Baltimore, Maryland, the USCG inspected the vessel and no deficiencies were noted. When Mr. Skiba received information, while in Baltimore, indicating that the vessel will be loading rice, he prepared a preliminary cargo plan assuming that the rice will be in bags of about 50kg each and that the destination was Sweden. The cargo holds were washed, cleaned and inspected by Mr. Skiba, the chief officer, the electric engineer and the chief engineer in preparation for the cargo of rice. Upon arrival to Freeport, a group apparently representing cargo owners inspected the holds to ascertain their fitness to receive the cargo. At this time Mr. Skiba found out that the rice would be in one ton bags (FIBC), resulting in the need to prepare a new cargo plan and stability calculations. Mr. Skiba went to the warehouse to familiarize himself with the bags. He measured a few bags concluding that they were approximately 1.2 meters in height and about the same in diameter or lengthwise. 4
The bags were made of plastic and a forklift damaged some during loading. The top of the bags was made of transparent material, sewn along the top or main bag. In other words, the bags were made of 2 materials, one for the body or lower portion, and one for the top. The bags were found to be susceptible to easy damage when contacting or getting banged against metallic structure. The bags previously carried from Brazil were made of much more durable and thicker material. During the last day of loading, the discharge destination was changed to Helsinki, Finland. At the destination, there was a draft restriction to 7.3 M. Hence, a modified cargo plan had to be developed for the 7,404 metric tons of rice. Mr. Skiba discussed the use of dunnage with the captain and found that apparently the charter papers indicated the use of paper as separation dunnage material, which he found unacceptable. Hence, wooden pallets and planks of wood were used, primarily under the bags, but also as separators between the bags and the metallic parts of the vessel. To obtain the pallets and wood planks, the shipper (the sender of the cargo) was contacted. The actual loading of cargo in the vessel started when the issue of the dunnage was settled. Cranes and forklifts were used to stow the cargo in the holds. First, on the port side, then starboard, aft and bow side, but Mr. Skiba felt that the shore side personnel loading the cargo were inexperienced. Mr. Skiba met with the supervisor of the stevedore and went with him into the holds to explain to the stevedores how to install the dunnage. At this time, some cargo had been already loaded and Mr. Skiba was horrified the way that it was stowed. Hence, the cargo was rearranged. Because the stevedores were not following Mr. Skiba s instructions properly, after several discussions, Mr. Skiba informed the captain and he participated in the discussions. The testimony appears to indicate that the crew of the vessel installed additional battens as instructed by the master. Pallets were used on decks, close to the bulkhead and partitions, on frames, etc., to prevent damage to the bags. Also, paper separations were used. Mr. Skiba believes that two types of bags were loaded. The red stitched bags had opaque tops and the blue stitched bags had clear tops, which may indicate a difference manufacturer. All damaged bags were returned to the dock. The spaces between bags were filled as tight as possible. The shore personnel did the lashing, but as the ship officers were not satisfied with the quality of the lashing, the crew of the ship redid the lashings. The shipper did not provide lashing materials, hence, whatever materials available aboard the vessel were used, specifically: polypropylene rope, hemp rope. Some of the ropes used were tightened (pulled) using crane equipment on the vessel. If the distance to stretch the rope was large, steel wire was used with 2 turnbuckles on each side. To protect the bags secured with wire rope, pallets were used in a vertical position. There was concern regarding the vessel structure used in conjunction with the lashings, particularly the rings. Mr. Skiba indicated that 200 large size automobile tires were also loaded to secure the cargo by squeezing the tires between the pallets. There was no cargo loaded in the port and starboard of no. 1 tween deck. These spaces were used to keep pallets that belong to the ship. Mr. Skiba indicated that his stability calculations comply with the requirement of the Russian rules and the range of stability was 60 degrees. The metacentric height upon departure was 72 to 5
74 cm. Ballast was carried in the forepeak, tanks 3 and 4 and 120 tons in the after peak. Draft at departure was approximately 7.93M. After leaving port, ballast was added in tanks 9 and 10. These tanks were not full. Some 20 tons were needed to fill them to capacity. Mr. Skiba indicated that it was dangerous to fill them completely because the possibility of breaking the hatches (by hatches, probably Mr. Skiba was referring to manholes). In other words, they were not pressed-up. Tanks 9 and 10 are indicated in the drawings to be fuel tanks, but during the shipyard work, they were probably converted to ballast tanks. They were full, but not pressed-up. The statement of facts indicates 673 tons of fuel and diesel and 125 tons of freshwater. The vessel was navigating a course provided by a weather routing service, going from the Straits of Florida to the Azores, just a little South of the recommended course. The bilges of the cargo holds were checked for water twice a day. Also, all ballast tanks and fresh water tanks were checked. No changes were made to the water ballast during the voyage and/or December 8 th. Pumping of the bilges was done during each watch when necessary. On December 8 th, during Mr. Skiba s watch from noon to 4:00 p.m., at about 3:00 p.m., he was ordered to perform test pumping from all the bilges and found no water in the bilges of the holds. During the 2 nd day or 3 rd day of the voyage, the master and the chief mate inspected hold #1 tween deck. They reinforced the lashing a little bit and found the cargo to be normal. For this inspection, the hatch covers were opened and then closed and dogged down. Before departure, the captain and the chief mate inspected upper tween no. 3. On December 8 th, on Mr. Skiba s watch from 0000 hours to 0400, the wind was 10 to 15 meters per second, the seas Beaufort scale was about 4 or 5, and the vessel was rolling 5 to 6 degrees both sides. The speed was about 13 knots and the course about 70 degrees. In consideration of the weather, the captain decided to alter course to 90 degrees when Mr. Skiba took his watch at 1200 on December 8 th. At this time, the center of the cyclone was 45 knots to the north and was northbound while the vessel was heading east. The winds were coming from the West by South, or from the starboard quarter, blowing at 20 meters per second. By the end of Mr. Skiba s watch the seas were 7 to 8 beaufort. The swells were huge and the vessel was rolling 6 to 7 degrees, with a period of roll of 20 seconds. The metacentric height was determined to be 68 centimeters, on the basis of the 20-second period of roll, which indicated that the ship was operating under normal conditions, during the 1200 to 1600 watch on December 8 th. Mr. Skiba indicated that the wide-wave (probably swells) had a period of 20 seconds. During and after Mr. Skiba s watch of 1200 to 1600, the weather was worsening dramatically. After having dinner, about 9:00 to 10:00 p.m., Mr. Skiba returned to the bridge finding the storm had increased and the roll was getting worse, with many items falling down. The captain and the 3 rd mate were on the bridge and the vessel was rolling 14 degrees to port side and 8 degrees to starboard side, indicating that the vessel had developed a static list to port of 5 to 6 degrees. Mr. Skiba went to his cabin and returned to the bridge about 11:30 because the list, the roll and pitching of the vessel were worsening. He heard the captain telling the third mate to stop 6
pumping port tank no. 10 and to turn the rudder 15 degrees to the right. A few minutes later the vessel sustained a heavy roll to port and a few minutes later, the engine stopped. Mr. Skiba was in communications with the engine room. The main engine could not be restarted. The captain made several announcements thru the P. A. system requesting the crew to gather at the bridge and activated the automatic emergency system. Mr. Skiba gave the radio operator the coordinates obtained by GPS. At this time, the static list was in excess of 20 or 25 degrees. At the time Mr. Skiba reached the bridge an attempt was being made to pump out overboard ballast from tanks no. 10 and no. 4, but there was a major confusion and the chief engineer told Mr. Skiba that they could not do it because they had to open the valves, which were in the pipe tunnel and due to the list, no one was willing to get in the tunnel. When Mr. Skiba arrived to the bridge he heard the captain ordering to stop pumping ballast overboard from tank no. 10. When the announcement was made for the crew to gather at the bridge, all the exterior and interior lights were turned on, and Mr. Skiba observed that the port side of the deck was at the water level and while rolling the water reached to half of the hatch coamings. As the crew arrived to the bridge, the captain ordered the crew to go to the lifeboats and order to start transmitting SOS signals. Mr. Skiba went down to get dressed with environmental clothing and upon his return to the bridge, the lights were on, no one was on the bridge and the list was approximately 45 degrees. Thereafter, the port lifeboat was lost; life rafts were lost and about 2:00 a.m., the starboard lifeboat was also damaged and drifting away. It was about 5:00 a.m. when Mr. Skiba could last observe the vessel s lights. The M/T Torungen eventually rescued Mr. Skiba. Conclusions: The primary cause of the loss of the M/V Salvador Allende was the shift of the cargo of rice, combined with the unexpected difficult or extreme environmental conditions, and inability to operate the bilge pumping system. The shift of cargo was the result of the combined effect of the following detrimental actions and procedures by various parties: a). The lack of appropriate lashing materials and appropriate dunnage. b). The apparently unsuitable forklift trucks, which should have been adapted for cargo operations with FIBC s, damaging the FIBC. c). The apparently insufficient strength of the FIBC s provided by the cargo interest. d). The lack of experience and/or knowledge of the shore personnel (stevedore, cargo surveyors, etc.) in charge of loading the vessel. e). The apparent insufficient strength of pad eyes and rings available in the holds of the vessel for proper anchoring of the lashing systems. f). The inability to operate the valves related to the bilge system because the remote control system was not functioning, and, when the vessel developed a 40 degree list, no one from the crew would volunteer to get in the pipe tunnel to operate the valves by hand. 7
g). Failure of the officers to determine and connect the cause of the 1 degree list to port developed 3 to 4 days after departure. About the Author: Hector V. Pazos is a Naval Architect, Marine Engineer and a Registered Mechanical Engineer and has been engaged in Accident Investigation/Reconstruction for more than 35 years. He has been retained as an Expert Witness in over 1,200 Maritime cases, related to both commercial vessels and pleasure crafts, for both defense and plaintiff. He has offices in St. Petersburg, FL and New Orleans, LA. He can be reached by phone at (727) 347-2556 or (504) 367-4072 or by email at: Hectorpazos75597@aol.com or Oceanoilpazos@cs.com. For more information, please visit our website at: www.siterrific.com/pazos. 8