Copyright USGenNet Inc., 2015 All Rights Reserved USGenNet Data Repository Please read USGenNet Copyright Statement on this page: Transcribed and submitted by Linda Talbott for the USGenNet Data Repository http://www.us-data.org/ =========================================================================== Formatted by USGenNet Data Repository Chief Archivist, Linda Talbott All of the above information must remain when copied or downloaded. =========================================================================== U.S. Marine Board of Investigation Marine Casualty Report SS EDMUND FITZGERALD, Sinking in Lake Superior CONCLUSIONS ----------- 1. Preface The SS EDMUND FITZGERALD left Superior, WI, on the afternoon of 9 November 1975, enroute Detroit, MI, with a full cargo of taconite pellets. That evening, and the next day, FITZGERALD proceeded eastward in Lake Superior, on a course north of the charted lanes due to the weather, heading towards Whitefish Bay and the Locks at Sault Ste. Marie, MI. At the same time, a severe November storm was crossing Lake Superior and, as a result, FITZGERALD encountered worsening weather throughout the early hours of the 10th of November, and by that afternoon, was experiencing winds in excess of 50 knots and seas approaching 16 feet. At approximately 1510, 10 November, FITZGERALD reported damage, but did not, at that time or in subsequent communica- tions, indicate that it was of a serious nature or that there was any immediate concern for the safety of the vessel. No distress message was received. FITZGERALD sank sometime after 1910, 10 November 1975, at a position 46°59.9'N, 85°06.6'W, approximately 17 miles from the entrance to Whitefish Bay, MI. There were no survivors and no witnesses to the casualty. Information available to the Marine Board consists of testimony of people who were on board other vessels in the area at the time FITZGERALD was lost, of people who had served on FITZGERALD prior to its last voyage, of employees of the company which operated the vessel, of other persons familiar with the vessel or similar vessels or its cargo, of personnel of the Coast Guard and of the American Bureau of Shipping who had conducted inspec- tions and surveys on the vessel, of Coast Guard personnel who participated in the extensive search which followed the report of its loss, of personnel from the National Weather Service concerning weather at the time of the loss, of personnel at 89 --------------------------------------------------------------------------- the facility where the vessel loaded its last cargo, and of information from the several underwater surveys which were conducted on the wreckage which was found on the bottom of Lake Superior. Information available is incomplete and incon- sistent in the following particulars. a. Position. The only information available on the position and trackline of FITZGERALD is in the weather reports sent by FITZGERALD and in testimony of the Master and Watch Officers of the SS ARTHUR M. ANDERSON, which was following it visually and on radar. The weather reports from FITZGERALD scheduled at 1300 and 1900, 10 November, were not received. The position of FITZGERALD relative to that of ANDERSON cannot be reconstructed. Information available is based on the recollections of the Master and Watch Officers on ANDERSON, since the relative position of FITZGERALD was observed intermit- tently on the radar, but not recorded. Testimony on these observations is inconsistent. For example, the Officer on watch on ANDERSON recalled that FITZGERALD was "a shade to the right of dead ahead," as FITZGERALD passed northeast of Caribou Island, while the Master thought it was a point to a point and a half to the right at that time. The Master and the Watch Officers on ANDERSON testified at length as to the position and trackline of ANDERSON in the afternoon and evening of 10 November. An analysis of this testimony shows that the vessel was navigated by radar ranges and bearings, that, at times, positions were deter- mined but not logged, that course changes were made without simultaneous determination of position, that positions were determined as much as twenty minutes from the time that course changes were made, and that the courses steered varied from the 90 --------------------------------------------------------------------------- course logged because of the expected drift. The Marine Board attempted to reconstruct the trackline of ANDERSON and found that in order for the vessel to have steered the courses and have been at the positions at the times testified to, the speed of the vessel would have varied from a low of 5 mph to a high of 66 mph. But the Master testified, and the engineering log confirmed, that throughout the period, ANDERSON maintained a steady speed, turning for 14.6 mph. Accordingly, it is con- cluded that the times and positions reported by officers of ANDERSON were not sufficiently accurate to allow the trackline of either FITZGERALD or ANDERSON to be reconstructed. b. Difficulties Reported by FITZGERALD. FITZGERALD reported the loss of two vents and some fence rail, indicating that topside damage had occurred to the vessel. The flooding which could be expected to result from the loss of any two tank on tunnel vents would not be serious enough, by itself, to cause the loss of the vessel. FITZGERALD reported, at the same time, that it had devolped a list. The existence of the list which would result from flood- ing of any two ballast tanks, a tunnel, or a tunnel and a ballast tank would not, of itself, indicate damage sufficiently serious to cause the loss of the vessel. FITZGERALD reported that steps were being taken to deal with the flooding on the list, and that two pumps ("both of them") were being used. FITZGERALD had four 7000-gpm pumps and two 2000-gpm pumps available, indicating that the flooding was evaluated by personnel on board FITZGERALD as not sufficiently serious to create a danger of loss of the vessel. FITZGERALD reported difficulties with its radars, and requested ANDERSON to provide navigational information. 91 --------------------------------------------------------------------------- FITZGERALD reported slowing down to allow ANDERSON to catch up. This action might have been taken because the Master of FITZGERALD knew or sensed that his problems were of a more serious nature than reported to ANDERSON. c. Underwater Survey. The underwater survey showed that mud covered a majority of the wreckage, that the midships section of the hull was completely disrupted, and that the stern section was inverted. Movement of the survey vehicle disturbed the mud, which limited visibility and made it diffi- cult to identify individual components of the wreckage. How- ever, the survey provided the Marine Board valuable information with respect to the vessel's final condition and orientation. 2. In the absence of more definate information concerning the nature and extent of the difficulties reported and of problems other than those which were reported, and in the absence of any survivors or witnesses, the proximate cause of the loss of the SS EDMUND FITZGERALD cannot be determined. 3. The most probable cause of the sinking of the SS EDMUND FITZGERALD was the loss of buoyancy and stability which resulted from massive flooding of the cargo hold. The flooding of the cargo hold took place through ineffective hatch closures as boarding seas rolled along the Spar Deck. The flooding, which began early on the 10th of November, progressed during the worsening weather and sea conditions and increased in volume as the vessel lost effective freeboard, finally resulting in such a loss of buoyancy and stability that the vessel plunged in the heavy seas. 4. The following factors contributed to the loss of FITZGERALD: a. The winter load line assigned to FITZGERALD under the changes to the Load Line Regulations in 1969, 1971 and 92 --------------------------------------------------------------------------- 1973 allowed 3 feet, 3-1/4 inches less minimum freeboard than had been allowed when the vessel was built in 1958. This overall reduction in required freeboard also reflected a reduc- tion in Winter Penalty for Great Lakes vessels. Not only did the reduction in minimum required freeboard significantly reduce the vessel's buoyancy, but it resulted in a significantly increased frequency and force of boarding seas in the storm FITZGERALD encountered on 10 November. This, in turn, resulted in an increased quantity of water flooding through loosely dogged hatches and through openings from topside damage. b. The system of hatch coamings, gaskets, covers and clamps installed on FITZGERALD required continuing mainten- ance and repair, both from routine wear because of the frequent removal and replacement of the covers and from damage which regularly occurred during cargo transfer. That the required maintenance was not regularly performed is indicated by the fact that the crew of the vessel had no positive guidelines, in the form of Company requirements or otherwise, concerning such maintenance. That the required repairs were not regularly performed as damage occurred is indicated by the fact that significant repairs had been required during the previous winter lay-up period and by the fact that more repairs of the same nature were expected, since a general item to repair hatch covers and coamings had been included in the work list for the winter lay-up which FITZGERALD was approaching when it was lost. It is concluded that the system of cargo hatch coamings, gaskets, covers and clamps which was installed on FITZGERALD and the manner in which this system was maintained did not provide an effective means of preventing the penetration of water into the ship in any sea condition, as required by Coast Guard Regulations. c. Whether all the cargo hatch clamps were properly fastened cannot be determined. In the opinion of the Marine 93 --------------------------------------------------------------------------- Board, if the clamps had been properly fastened, any damage, disruption or dislocation of the hatch covers would have resulted in damage to or distortion of the clamps. But, the underwater survey showed that only a few of the clamps were damaged. It is concluded that these clamps were the only ones, of those seen, which were properly fastened to the covers and that there were too few of these and too many unfastened or loosely fastened clamps to provide an effective closure of the hatches. d. The cargo hold was not fitted with a system of sound- ing tubes or other devices to detect the presence of flooding water. It is not known whether any efforts were made to deter- mine if water was entering the cargo hold. If the hold had been checked at a time when the level of the water was below the cargo surface, the extent of flooding could not have been determined. It is inconceivable that flooding water in the cargo hold could have reached a height to be seen, without a seasoned Master taking more positive steps for vessel and crew safety than were reported. Therefore, it is concluded that the flooding of the cargo hold was not detected. e. The cargo hold was not fitted with transverse water- tight bulkheads. As a result, the flooding water which entered could migrate throughout the hold, extending the effect of the flooding and aggravating any trim which existed. 5. At sometime prior to 1530 on 10 November, FITZGERALD experi- enced damage of sufficient magnitude to cause the Master to report topside damage and a list. Significantly, the Master of FITZGERALD reported the damage rather than the incident which caused it. It is the opinion of the Marine Board that the incident, while possibly of a serious nature, was not of such extent as to have caused, by itself, the loss of the vessel and, 94 --------------------------------------------------------------------------- further, that the full extent of the incident was not perceived by vessel personnel. The Master noted the list and topside damage and incorrectly concluded that the topside damage was the only source of flooding. He began what he believed were ade- quate, corrective measures - pumping spaces which would receive flooding from damaged vents - and thus felt the problems were under control. The topside damage could have been caused by the vessel striking a floating object which was then brought aboard in the heavy seas. This also could have resulted in undetected damage opening the hull plating above or below the waterline and additional unreported damage to topside fittings, including hatch covers and clamps. Intake of water into the tunnel or into one or more ballast tanks through the damaged vents and opened hull would have produced the reported list and increased the rate of cargo hold flooding. The most likely area of damage would have been in the forward part of the ship. The vessel had entered a snow storm approximately one-half hour before the topside damage was reported. In addition, FITZGERALD's radars were reported inoperative shortly after the damage was reported, and may have been malfunctioning for some period before the report. Both the reduced visibility from the snow storm and the radar malfunction would, in the opinion of the Marine Board, have reduced the liklihood that the crew of the vessel could have detected the object in sufficient time to take effective action to avoid it. The topside damage could have been caused by some unidenti- fied object on board breaking away in the heavy seas. Flooding through such damage could have caused a list. While there were objects on deck which might have come adrift and knocked off a vent cap or damaged a hatch coaming, the only items on deck which had enough mass to do sufficient damage to the hull to 95 --------------------------------------------------------------------------- cause a sustained list were a hatch cover, the hatch cover crane, or the spare propeller blade. If such extensive damage had occurred, a seasoned Master would have reported it. Such a report was not received. The topside damage and list could have been caused by a light grounding or near grounding on the shoals north of Caribou Island. Although their testimony is not fully consistent, both the Master and the Watch Officer on ANDERSON indicated that FITZGERALD passed within a few miles of Caribou Island and that they had a conversation concerning the closeness of FITZGERALD to the shoals north of the island. It is considered possible that a light grounding or near grounding on these shoals could have occurred. The vessel could have been damaged from the grounding, from the effect of the violent seas which would be expected near the shoals, or from the shuddering that the vessel would have experienced as it passed near the shoals. The damage could have been on deck, below the water line, or both, leading to the reported topside damage and list. The Marine Board is unable to reconstruct the trackline of FITZGERALD south of Michipicoten Island, however, FITZGERALD was observed to pass two to three miles off Michipicoten Island West End Light from which position a single course change to 141°T would have taken the vessel directly to Whitefish Point on a track well clear of the shoal areas off the northern tip of Caribou. Had there been a delay in making the course change after passing Michipicoten, FITZGERALD would have passed closer to the shoals. But, the distance between Michipicoten and the shoals is such that it appears that a delay in making the course change of upwards of an hour would have been required to cause FITZGERALD to have actually reached the shoals. The list could have been caused by a localized hull struc- tural failure, resulting in the flooding of the ballast tank or tanks. There is no correlation between such an occurrence and the reported loss of vents and fence rail. The survey of those parts of the wreckage which could be seen showed no evidence of brittle fracture. 96 --------------------------------------------------------------------------- The Marine Board concludes that the exact cause of the damage reported cannot be determined, but that the most likely cause was the striking of a floating object. 6. In the opinion of the Marine Board, the flooding from the damage reported, and from other damage which was not detected, most likely occurred in the forward part of the vessel, resulting in trim down by the bow. By the time the damage was reported by FITZGERALD, the flooding of the cargo hold had reached such an extent that the cargo was saturated and loose water existed in the hold. Because of the trim by the bow, this water migrated forward through the non-watertight screen bulkheads which separated the cargo holds, further aggravating the trim and increasing the rate of flooding. 7. Because there were neither witnesses nor survivors and because of the complexity of the hull wreckage, the actual, final sequence of events culminating in the sinking of the FITZGERALD cannot be determined. Whatever the sequence, however, it is evident that the end was so rapid and catastrophic that there was no time to warn the crew, to attempt to launch lifeboats or life rafts, to don life jackets, or even to make a distress call. Throughout November 10th the vessel was subjected to deteriorating weather and an increasing quantity of water on deck. With each wave that came aboard, water found its way into the cargo hold through the hatches. As the vessel lost freeboard because of this flooding and as the sea conditions worsened, the frequency and force of the boarding seas in- creased, and so did the flooding. The Master of the vessel reported that he was in one of the worst seas that he had ever seen. It is probable that, at the time he reported this, FITZGERALD had lost so much freeboard from the flooding of the 97 --------------------------------------------------------------------------- cargo hold that the effect of the sea was much greater than he would ordinarily experienced. Finally, as the storm reached its peak intensity, so much freeboard was lost that the bow pitched down and dove into a wall of water and the vessel was unable to recover. Within a matter of seconds, the cargo rushed forward, the bow plowed into the bottom of the lake, and the mid- ships structure disintigrated, allowing the submerged stern section, now emptied of cargo, to roll over and override the other structure, finally coming to rest upside-down atop the disintigrated middle portion of the ship. Alternatively, it is possible that FITZGERALD sank as a result of a structural failure on the surface, resulting from the increased loading of the flooding water. However, this is considered less likely because such a failure would have severed the vessel into two sections on the surface, and one or the other, if not both sections, would have floated for a short while. With the weather conditions that existed at the time FITZGERALD was lost and, in particular, with the winds in excess of 50 knots, if either or both of the pieces had floated for any time, significant drifting would have occurred. But, the survey of the wreckage showed that the two main pieces were within a ship length, thus little or no drifting took place. 8. There is no evidence that the crew of FITZGERALD made any attempt to use any lifesaving equipment, or that lifesaving equipment or its performance contributed in any way to this casualty. The condition of the lifeboats recovered indicates that the boats were torn away from their chocks, grips and falls. The condition of the life rafts recovered indicates that they were released from their float-free racks and inflated as they were designed to. One raft was damaged, partly when it floated onto the rocky shoreline and partly by a search party 98 --------------------------------------------------------------------------- which punched holes in it to allow water to drain out during the recovery operation. Testimony of witnesses indicates that a successful launching of a lifeboat would have been extremely difficult in the weather and sea conditions which prevailed at the time FITZGERALD was lost. This testimony also indicates that Great Lakes mariners have little confidence that lifeboats could be launched success- fully in other than moderate wind and sea conditions, and given the choice, they would use the inflatable rafts as the primary means of abandoning a sinking ore carrier. Their confidence in the capability of the rafts was tempered by stated beliefs that a raft could not be boarded safely once it was launched and water- borne and that they would inflate it on deck and wait for it to float free from the sinking vessel. This illustrates that although Great Lakes mariners understand the difficulties inherent in dis- embarking from a stricken vessel their level of understanding of the use and capability of inflatable life rafts is inadequate. In the opinion of the Marine Board, the appraisal by crewmen that they have small chance of survival on abandoning a stricken vessel in a rough seaway could influence them to stay with the stricken vessel rather than attempt abandonment. The present requirement for posting a placard containing life raft launching instructions is not considered sufficient to train crewmembers in the proper use of this primary lifesaving equipment. The placard is, however, considered a valuable aid in assisting and reinforcing other crew training. Lifeboat drills were held on FITZGERALD during the 1975 season, but were not held on a weekly basis as required by regulations. The level of training of the crew in the use of lifeboats and life rafts is indeterminate. 99 --------------------------------------------------------------------------- There is no evidence to indicate that any of the crewmembers of FITZGERALD escaped from the vessel at the time of its loss. However, if they had, their chances of survival would have been significantly enhanced if they had been provided with equipment to protect them against exposure. 9. The twenty-nine crewmen on board FITZGERALD are missing and presumed dead. 10. It was fortunate that the Steamer ARTHUR M. ANDERSON was in the area of an in radiotelephone communication with FITZGERALD on the afternoon and evening of 10 November. Without the presence of this vessel, the loss of FITZGERALD would not have been known for a considerable period of time, possibly not until the follow- ing day, and, at the latest, when the vessel failed to arrive at the unloading dock. 11. The testimony of witnesses indicates a conflict as to the time that the Coast Guard was first notified of the problems with FITZGERALD. The Marine Board concludes that the first notification that the Coast Guard received of the problem with FITZGERALD was at approximately 2025 Eastern Standard Time on 10 November in a radiotelephone call from CAPT Cooper, Master of ANDERSON. At the time of this call, the actual loss of FITZGERALD was neither compre- hended by CAPT Cooper nor conveyed to the Coast Guard. The Coast Guard radio watchstander who received the call attempted to com- municate with FITZGERALD, without success, and advised the Rescue Coordination Center. The second call from CAPT Cooper to the Coast Guard, at approximately 2100, 10 November, did express a grave concern that FITZGERALD was lost, and rescue efforts were initiated. It is concluded that the time period which elapsed in evaluating and reporting the loss of FITZGERALD did not contribute to the casualty or high loss of life, because FITZGERALD sank suddenly, with all hands trapped on board. 100 --------------------------------------------------------------------------- 12. In the opinion of the Marine Board, in a tragedy of this magnitude, occurring, as this one did, in extreme weather conditions, vessels in the area and SAR aircraft must be relied upon as the first source of assistance. The response by the merchant vessels in the area to the Coast Guard's request for assistance was in keeping with the finest traditions of mariners. The response of the vessels ARTHUR M. ANDERSON and WILLIAM CLAY FORD is considered exemplary and worthy of special note. These vessels proceeded to the scene on the night of 10 November and searched under conditions of extreme weather and sea on 10 and 11 November. The response of the Canadian vessel HILDA MARJANNE, which got underway but was forced back by weather, is also worthy of note. The response by Coast Guard SAR aircraft from Air Station Traverse City was timely. The first aircraft was not launched until 51 minutes after it was ordered because it was necessary to load flares for the night search. The launching of three aircraft within one hour and thirty-five minutes is within the response requirements called for by the Ninth Coast Guard District SAR Plan. The request for and dispatch of additional SAR aircraft from Coast Guard Air Station Elizabeth City, NC, from the U.S. Navy, from the Michigan Air National Guard, and from Canadian SAR forces was also timely. The only Coast Guard surface unit in an SAR standby status which was close enough to respond within a reasonable time and was large enough to cope with the weather and sea conditions which prevailed at the time was the Buoy Tender WOODRUSH at its home port in Duluth, MN. WOODRUSH, on a six-hour standby status, was underway within two and one-half hours. The Marine Board concludes that the response by the WOODRUSH was timely. The wind and sea conditions precluded the use of the Harbor Tug 101 --------------------------------------------------------------------------- NAUGATUCK stationed at Sault Ste. Marie, which had operating limitations imposed on its use outside harbor waters. The small craft designed for coastal operations which were available in Lake Superior were unsuitable for search 15 miles offshore in the high sea state then existing. It is concluded that there is a need for additional surface forces with SAR capability to improve the overall search and rescue posture in Lake Superior. 13. Because ANDERSON was following FITZGERALD, providing navigational assistance and observing FITZGERALD to be on a trackline heading for the entrance to Whitefish Bay and because the wreckage was found on a trackline headed for the entrance to Whitefish Bay, it is concluded that the outages of Whitefish Point light and radio beacon did not contribute to the casualty. 14. The progress of the severe storm which crossed Lake Su- perior on 9 and 10 November was adequately tracked by the National Weather Service and the weather reports and weather forecasts adequately reflected its path and severity. Weather forecasts were upgraded in a timely manner and a special warning was issued. Estimates of wind velocity by persons on vessels in the storm were higher than those forecast and also higher than those reported by shoreside stations, however, the overall severity of the storm was generally as forecast and reported. It is concluded that mariners on Lake Superior on 10 November were adequately warned of the severe weather and that the Master of FITZGERALD was aware of the severity and location of the storm. 15. Testimony of licensed Great Lakes mariners indicates the cargo hold of a Great Lakes ore carrier cannot be dewatered if it is loaded with a cargo of taconite pellets. The Marine Board is unable to determine the validity of this as a general proposition or whether it affected the loss of FITZGERALD. 102 --------------------------------------------------------------------------- 16. The Loading Manual which was developed for FITZGERALD did not comply with the requirements of the Load Line Regu- lations. Since the only loading information available to the Marine Board is the total cargo carried on down-bound voyages, whether FITZGERALD was ever subjected to unacceptable stresses cannot be determined. 17. The underwater survey of the wreckage and the detailed study of the photographs taken show no apparent relationship between the casualty and the discrepancies found and reported at the Spar Deck Inspection conducted on 31 October 1975. 18. The hydrographic survey performed by CSS BAYFIELD basically confirmed the data indicated on chart L.S.9 and Canadian chart 2310. In addition, this survey showed that the northern end of the shoals north of Caribou Island extends approximately one mile further east than indicated on Canadian chart 2310. 19. The nature of Great Lakes shipping, with short voyages, much of the time in very protected waters, frequently with the same routine from trip to trip, leads to complacency and an overly optimistic attitude concerning the extreme weather hazards which can and do exist. The Marine Board feels that this attitude reflects itself at times in deferral of mainten- ance and repairs, in failure to prepare properly for heavy weather, and in the conviction that since refuges are near, safety is possible by "running for it." While it is true that sailing conditions are good during the summer season, changes can occur abruptly, with severe storms and extreme weather and sea conditions arising rapidly. This tragic accident points out the need for all persons involved in Great Lakes shipping to foster increased awareness of the hazards which exist. 103 --------------------------------------------------------------------------- 20. There is no evidence of actionable misconduct, inattention to duty, negligence, or willful violationi of law or regulation on the part of licensed or certificated persons, nor evidence that failure of inspected material or equipment, nor evidence that any personnel of the Coast Guard, or any other government agency or any other person contributed to the cause of this casualty. 104 ===========================================================================