The National Transportation Safety Board (NTSB) has concluded its investigation and has issued a full report about the incident on the towing vessel Spence, which caused it to list badly before finally sinking. The incident occurred on 14 December 2015 about 115 nm north of Cartagena, Colombia.
The NTSB report says that the list increased despite efforts by the crew to correct it. Consequently, the captain activated the vessel’s emergency position indicating radio beacon (EPIRB), and the crew climbed onto the barge that the Spence was towing. The towing vessel sank shortly thereafter. US Coast Guard District 7 received the EPIRB alert and directed the Coast Guard cutter Decisive to the scene. On arrival, the Decisive rescued the four crew members from the drifting barge. Three crew members sustained non-life threatening injuries.
The Coast Guard Marine Safety Center (MSC) conducted a post-sinking analysis of the Spence to determine the most likely source of flooding during the accident. The MSC was provided with copies of some of the original construction drawings and the 1974 stability letter listing lightship weight, vertical center of gravity (VCG), and longitudinal center of gravity (LCG).
The MSC analysis noted that because these parameters were not updated during the vessel’s 41 year service life to reflect material additions, removals, and repairs, the MSC had to estimate the vessel’s displacement and LCG at the time of the sinking using photographs of the vessel’s departure condition and tank levels reported by the engineer during his interview. A range of possible VCGs were evaluated to ensure the vessel’s actual condition was bracketed by the analysis. The MSC used the crew’s description of how the Spence behaved as it sank in an attempt to validate or discount possible flooding scenarios. The MSC analyzed several flooding scenarios.
NTSB says that flooding the potable water tank alone would not have caused the vessel to list, nor the aft deck to submerge as the crew had described. The second scenario indicated that complete flooding of only the rudder compartment would submerge half the aft deck, but would not cause a list. The third scenario showed that with only the aft void completely flooded the entire aft deck submerged, but the vessel again did not list.
The MSC surmised that because none of these scenarios matched the crew’s description of the sinking, gradual water ingress into the aft void was more likely than sudden flooding. The MSC then analyzed gradual flooding of the aft void in 10-percent volumetric increments. The results showed that for a range of vessel VCGs, when the aft void was flooded approximately 70–90 percent, the vessel would likely enter a lolling condition, resulting in a sudden list to starboard of 20–30 degrees, consistent with crew reports.
The NTSB investigation determined that, at a minimum, downflooding could have occurred through several gooseneck vents at the aft end of the superstructure, which likely submerged in the lolling condition. Additional analysis with the aft void flooded 70 percent and the engine room progressively flooding through downflooding points indicated the vessel would sink by the stern without capsizing.
The MSC stated that the most likely sinking scenario involved gradual flooding in the aft void, which caused the aft deck to submerge and the vessel to enter a lolling condition, resulting in a sudden list to starboard. In this state, water would then have likely entered other spaces through downflooding points, causing progressing flooding and sinking by the stern without capsizing.
The MSC analysis stated that not enough information was available to comment on the possibility of repairs having compromised the hull of the vessel.
Click to read the NTSB report in full: Sinking of Towing Vessel Spence Report