Site Formation Processes and Pollution Risk Mitigation of World War II Oil Tanker Shipwrecks: Coimbra and Munger T. Ball

Locating and documenting potentially polluting wrecks is essential in determining the risks they pose to causing oil spills as weakening hulls in the marine environment continue to corrode. Expeditions in 2019 and 2021 to two World War II oil tankers, Coimbra and Munger T. Ball, assessed the site formation processes and integrity of the hulls for pollution mitigation. This was followed by remediation that removed large amounts of oil from the tanks accessible on these wrecks. Lessons learned indicate that such approaches to shipwrecks in deeper waters may prove useful to locating wrecks and mitigating future potential spills.


Introduction
In response to concerns over the threat posed by shipwrecks leaking oil and other contaminants, Congress allocated funding in 2010 to the National Oceanic and Atmospheric Administration (NOAA) to develop a focused list and assess the risks posed by potentially polluting wrecks (PPWs). That NOAA study, completed and published in 2012, which focused on 87 potentially polluting wrecks in US waters, laid the groundwork for mitigating pollution threats from the wrecks of ships sunk carrying large amounts of fuel oil or other potentially polluting cargos. An internal database at NOAA, Resources and Undersea Threats (RUST), was reviewed as part of a larger effort to evaluate pollution risks in the US Exclusive Economic Zone (EEZ) for the Remediation of Underwater Legacy Environmental Threats (RULET) effort (Symons and Hodges 2004). This review narrowed down more than 20,000 shipwrecks to 573 that could pose a pollution risk, of which 87 were selected and prioritized as PPWs (NOAA 2013a). However, since this review, only four of these wrecks have been investigated. 1 The few cases where pollution remediation has taken place have been on known, shallow-water wrecks that were actively leaking and causing oil slicks and tarballs to appear on coastlines. However, the majority of the shipwrecks on the PPW list are reportedly in deeper water and have never been located. Furthermore, more than half of these vessels were sunk by U-boats during World War II as targets in Germany's war on American shipping to the Allied front; their fuel cargos are now around their 80th year underwater in continually corroding steel hulls. Further ocean mapping and finer-scaled efforts to locate these shipwrecks is needed in order to document and evaluate the sites, and potentially address whatever pollution risk they may pose. A key aspect of the assessment, which differed from earlier efforts and some international studies, was the application of archaeological assessment and with that an understanding of the site formation processes of the shipwrecks.
The events of World War II reached nearly every continent, and the expansive conflict left a broad impact on the planet's surface in the aftermath. The largest number of shipwrecks attributable to a single event in human history are the vessels lost in World War II; the Battle of the Atlantic alone claimed over 3000 merchant ships, 175 Allied warships, and 765 German and Axis U-boats. While a number were lost on convoy routes on the high seas, many were lost at the vulnerable spots of voyages at the beginning and end of a crossing close to port or on a coastal voyage. One such region was the Atlantic seaboard of the United States, extending into the Gulf of Mexico, where German U-boats carried out Operation Paukenschlag (Drumbeat) in early 1942 to attack and cripple American shipping of supplies-particularly fuel-to the Allied warfront in England and Europe. The geographic span of this assault, a large part of the broader Battle of the Atlantic that lasted the duration of the war, comprises a maritime cultural landscape of sunken oil tankers and freighters ranging from the coasts of New England to Florida and up to the Mississippi Delta in the Gulf of Mexico (Marx and Delgado 2013a;Church 2016;Hoyt et al. 2014). While separated by hundreds of miles, many of these shipwrecks are connected by vessel type, sinking incident, or the shared mission to deliver fuel oil across the Atlantic. Multiple oil tankers of the same type were sunk during the conflict, such as the T2-SE-A1 type tankers, including Bloody Marsh and Esso Gettysburg, or vessels owned and operated by the same company, such as Gulfoil, Gulfpenn, and Gulfstate. Similarly, many suffered their fate at the hands of the same U-boat captain, such as Reinhard Hardegan of the U-123, who sank Coimbra and Norness in January 1942 and then Gulfamerica off Jacksonville, Florida in April; and Captain Harro Schacht of U-507 who sank the freighter Norlindo, and tankers Munger T. Ball and Joseph M. Cudahy within hours of one another on May 4, 1942, followed by Alcoa Puritan and Virginia among others in the northern Gulf (Gannon 1990). The expansive maritime cultural landscape created by these actions early in America's tenure of the war are, in most cases, war graves of the sailors who went down with their ships, and many of the vessels still contain their fuel cargos or bunkers to the present day, corroding on the seabed.
The large number of merchant vessel shipwrecks that comprise the expansive maritime cultural landscape of the Battle of the Atlantic are represented in a multiple property nomination to the National Register of Historic Places entitled World War II Shipwrecks Along the East Coast and Gulf of Mexico (Marx and Delgado 2013a). This was done to ensure that historic preservation law and guidelines that regulate the activities of U.S. federal agencies (such as NOAA and the U.S. Coast Guard) properly assess the significance of the wrecks as historic and archaeological sites and take appropriate steps to mitigate the loss of historical and/or archaeological integrity in remediating pollution threats. However, the application of preservation law does not bar or limit remediation. The use of the National Register also provided a systematic means with a uniform set of criteria to determine which wrecks were significant and those that were not. The application of a multiple property approach streamlined the process of determining eligibility or listing in the National Register and proved effective.
These wartime casualties, as noted, were the result of Germany's war on Atlantic shipping to sink Allied merchant ships at a faster rate than they were being built, which with Operation Drumbeat brought the war to American shores within a month of the Pearl Harbor attack (Gannon 1990). As fuel was critical to the Allied war effort, tankers were given the highest priority, and in all, 61 were sunk by U-boats in American waters before the end of the war (Marx and Delgado 2013a, 39). This assault was in a large part counteracted by the implementation of the convoy system that had been effective in WWI.
The convoy system that developed as a result of the U-boat threat changed the way goods were moved in order to minimize the chances of being sunk. Ocean going merchant ships, typically freighters and tankers, supplied and fueled the Allied war effort and were vital to the success of strategic campaigns that led to the war's end. The Allied merchant vessel shipwrecks are significant due to their role as a link between war time goods being produced back home and troops fighting on the front line (Marx and Delgado 2013a, 49).
The significance of the 163 sunk merchant vessels included in the potential scope of the Battle of the Atlantic nomination lies in the variety of shipbuilding techniques and wartime innovation they, as a group, represent, which includes vessels constructed over the previous five decades (Marx and Delgado 2013a, 16). As older vessels were sunk, they were replaced by newer types that instead reflected a "new American style of shipbuilding" that included changes in steel manufacture, electric power, welded hull joinery, and the use of petroleum fuel over coal; the United States' shipbuilding program that accelerated due to U-boat casualties initiated the design and construction of prefabricated, welded ships, including Victory and Liberty freighters and the T2 and T3-class tankers (Marx and Delgado 2013a, 17).
The question of archaeology and archaeological significance, also addressed in the nomination, has only partly been addressed in practice but not in theory. The archaeology of oil tanker shipwrecks is unusual in the discipline of maritime archaeology because the construction of the vessels and the manner in which they sank is generally known. The role of archaeology in the investigation of these sites usually lies not in using elements of the ships design to identify the wreck, or larger questions about the battlefield, but to assess attack, sinking, and environmental damage to the structure of an individual wreck and evaluate what cargo tanks may still be intact and containing product. Rather than using shipbuilding techniques to identify the wreck, as one would do with an unknown nineteenth century wreck, for example, the ships plans are utilized to pinpoint internal bulkheads and estimate where oil may remain contained in the hull depending on its condition and orientation on 1 3 the seabed. Site assessments of potentially polluting wrecks require additional steps beyond typical archaeological characterization, as the amount of oil remaining on board must be determined through tapping the hull in numerous locations, and is often followed by pollution mitigation by way of petroleum product removal. At this point in the assessment, the history and archaeology of the shipwreck takes a back seat to what becomes an industriallevel remediation of the physical site for the safety of the environment in which the wreck lies, and adjacent environments that may also be threatened.
Oil tanker shipwrecks from World War II reflect an 'out of sight, out of mind' mentality from the period that saw thousands of ships sunk by enemy action or scuttled in the aftermath of the war as a simpler option to scrapping. This is best exemplified by the surviving ships from Operation Crossroads, irradiated by two nuclear blasts at Bikini Atoll in 1946 and sunk in locations across the Pacific in the following years to discard the irradiated and unusable hulks (Lenihan 2013;Delgado et al. 2018). The exploration for and characterization of these shipwrecks develops into an 'archaeology of discard' that is particularly relevant for this era of naval discard during and following the war . While many of the ships sunk as targets or scuttled following Operation Crossroads were emptied of hazardous materials ahead of their sinking, some were not, as evidenced by the barrels of low-level radioactive waste observed on USS Independence off San Francisco , and in the multiple actively leaking wrecks remaining at Bikini Atoll (Lenihan 2013, Brennan et al. in prep).
Oil tanker wrecks fit into the maritime cultural landscape of the war, and a landscape of environmental hazards we now face due to the attitude toward maritime assets at the time, in a more precarious way than those wrecks mentioned above. As wartime losses carrying fuel cargoes, oil tanker shipwrecks may hold a more urgent need to be located and assessed than other shipwrecks of the time. There are histories to tell-of survivors, of those who perished in the attacks, and of the lost hulls sitting on the seabed for 80 years-but the more vital story to be told is how much oil remains on the wreck, and what does the metal hull tell us about how much longer it may retain its polluting cargo. Oil slicks on the surface over many known oil tanker wrecks tell us possibly not much longer.
Mitigation of oil pollution risks has proven to be complicated and expensive, and much of this is compounded by the fact that more than half of the wrecks on the PPW list have yet to be located and are in deep water. The handful of wrecks that have been addressed have yielded varying results. Table 1 summarizes some of the mitigation efforts that have been conducted on a handful of potentially polluting wrecks with varying levels of success. Some remediations appear urgent when oil in overhead spaces leaks out during strong currents or storms but the oil remaining is actually low, for example Brigadier General M.G. Zalinski which recovered fuel but primarily oily water from within the hull. On the other hand, remediations of Prinz Eugen, the German cruiser that capsized under tow at Kwajalein following the two atomic tests at Bikini Atoll (NAVSEA 2019), and the recent effort at the Coimbra wreck in 2019 showed the levels of petroleum product that could still remain in hulls. In the case of Coimbra, which will be discussed further, significant amounts of the cargo of lubricating oil remained in the hull even despite the vessel being in three sections. The largest removal, however, was the nearly 2 million gallons of Navy Special Fuel Oil (NSFO) and diesel removed from the T3-S2-A1 auxiliary oiler, USS Mississinewa, sunk at Ulithi Atoll by a Japanese kaiten (a pilot-guided submersible with an explosive warhead) in 1944 soon after loading its cargo (NAVSEA 2004;Delgado et al. 2016:104-107). Note, however, that each of these mitigation projects was conducted on wrecks in relatively shallow water, and tackling the PPW problem at wrecks in deep water remains an untested endeavor outside of the salvage industry.
In May 2019, a multi-agency and multidisciplinary team, including Resolve Marine Group, the U.S. Coast Guard, SEARCH Inc. and teams of commercial divers and ROV operators, conducted a mission to the wreck of M/V Coimbra sunk south of Long Island in January 1942 aboard the work vessel M/V Shelia Bordelon. Following a detailed visual and sonar site assessment, divers tapped each cargo tank that was determined to remain intact and an estimated 450,000 gallons of oil was removed from the wreck, including the main cargo of lubricating oil as well as unreported Bunker C heavy fuel oil, which was found in one of the forward cargo tanks. Then in June 2021, a similar mission was conducted to the wreck of SS Munger T. Ball, sunk off Key West in May 1942. An estimated 35,000 gallons of heavy fuel oil was pumped from the wreck following the site assessments. These missions illustrate the efficacy of modern technology to mitigate the pollution risk from oil tanker shipwrecks. Furthermore, the site assessments and characterizations of the sites prior to the oil removal operations provide insights into site formation processes and impacts of more than 75 years underwater that lead to the movement of polluting materials from the hulls into the marine environment.

M/V Coimbra
Coimbra was built by Howaldtswerke A.G. in Kiel, Germany in 1937 with official number 165498. The single screw steam tanker was 422.8 feet long with a 60.4 feet beam and a gross tonnage of 6768 tons (Fig. 1). The riveted, double bottom steel-hulled tanker had two decks designed to carry bulk petroleum products, machinery positioned at the stern and two masts. The cargo area consisted of 16 water tight bulkheads extending to the upper deck and included peak tanks, deep tanks, fuel oil tanks and cargo tanks (Lloyd's Surveyor Report 1937). Coimbra was registered to a British company, Standard Transportation Company, and ran regular transport routes between Texas and Aruba to Portugal, Morocco, and Nigeria, commonly carrying refined petroleum products such as gasoline and kerosene. The vessel's operations shifted with the outbreak of World War II in 1939 on behalf of the British Admiralty to carry lubricating oil from the US to the UK, often sailing with convoys (Gordon 1991;Marx and Delgado 2013b). Coimbra became the second ship sunk during Operation Drumbeat as it sailed from New York on January 14, 1942 with a cargo of 81,000 barrels of lubricating oil headed to Liverpool, England. Following the sinking of Panamanian tanker, Norness, Reinhard Hardegen, Captain of German U-boat, U-123, spotted the unescorted Coimbra sailing along the south coast of Long Island (Gannon 1990).
The first torpedo struck the tanker amidships, lighting the bridge on fire and causing a list to starboard, and the second torpedo struck the engine room under the funnel, and the ship sank quickly by the stern.
The wreck of Coimbra was first assessed for pollution risk in 1967 by the U.S. Coast Guard and determined that residual lubricating oil cargo remained within the sunken hull. This report included statements from divers who said that the wreck was broken into three sections. Since then, technical divers have returned to the wreck and have removed objects from the site, such as portholes and china. Reports from these divers in the 1980s described the wreck and sketched the bow section resting on its starboard side with a slight turn past 90°. A study of surface currents was conducted offshore Long Island in 1976 with helicopter-dropped drifters, which was employed to model the pollution potential of surface slicks coming from the Coimbra wreck (Frey 1978). A multibeam survey by the NOAA Office of Coast Survey in 2009 from R/V Thomas Jefferson mapped the wreck and confirmed its condition in three sections, but showed that the bow section appeared to have shifted to further overturned.
The Coimbra wreck continued to leak oil. The 2019 mission to the wreck was conducted by Resolve Marine Group, who has experience dealing with mitigation and recovery efforts on modern sunken vessels, such as Manolis L off Newfoundland in 2016. The team partnered with the U.S. Coast Guard and archaeologists from SEARCH Inc. to conduct the assessment and mitigation of the pollution risk from the vessel M/V Shelia Bordelon. Operations consisted of ROV inspection of the wreck that allowed for matching the plans of the tanker to the wreck and lining up the locations of cargo tanks. Patches of the overturned hull were cleaned with wire brush and cavitation tool with the ROV so that a team of commercial divers could use a specialized drill to tap into the tanks and fit it with a cap that would allow for pumping of the tank. Fifteen of the 32 cargo and bunker tanks on board were accessible and found to be intact and still containing oil. Following the initial assessment, these tanks were pumped and an estimated 450,000 gallons of oil was recovered and disposed of at a facility in New York.
During the ROV assessment dives, the bow section was found to be almost entirely capsized with the port bilge keel facing upwards, which is farther overturned than divers stated previously but matches the more recent sonar data (Fig. 2). The midsection is more on its starboard side and less capsized than the bow, and the stern section is squarely on its starboard side. The B.L. 4″ wire mark VIII E. O.C. 1911 No. 934 naval gun remains in place and canted upward from its tripod mount on the poop deck at the stern. The ROV inspection located the smoke stack in the debris field between the stern and midships sections, which was blown off by the second torpedo strike at the engine room. Much of the reported debris and artifacts that had been on the site-including china, portholes, hatch covers, and possibly small arms and ammunition-appear to have all been removed by recreational divers, as none of this material was observed during the inspection.
The steel of the hull appears to be in fairly good condition, especially in places where the interior bulkheads were intact and containing oil; hull that was exposed to seawater on both sides, such as cargo tanks that were breached on either side of the torpedo strikes, showed more deterioration by corrosion. Oil was observed to be primarily leaking from corroding rivets. Coimbra was constructed with double plating in some locations due to internal bracing of the bulkhead construction, and there are a large number of rivets in certain places, particularly along the bilge keel at the turn of the hull. Where oil was found to be leaking out of rivets, the steel plates also appeared pitted. A small section of riveted plating was removed by divers to test the steel's friability with the drill for tapping tanks. The pitted surface and corroding rivets are apparent on this sample of hull plate in Fig. 3.
Tapping of the 15 tanks that were accessible to divers and appeared intact determined that ten contained substantial amounts of fuel product inside. Samples of the liquid in these tanks confirmed the cargo of lubricating oil, and in addition found that Coimbra had also loaded this product into its summer tanks. Unexpectedly, heavy fuel oil was found stored in the forward deep and peak tanks, which was unreported and possibly carried as additional cargo; cargo manifests during the war often did not report everything that was loaded to avoid being targeted (Marx and Delgado 2013a, 50). The interior bulkheads between these forward tanks and the forward main cargo tanks appeared intact because product did not flow between them during removal.
A crack in the hull near the keel was noted starting at Cargo Tank 4 and runs the entire length of the bow forward, indicating oil was no longer present in those starboard tanks. This crack appeared to be a separated lap joint, and internal tank structure was visible inside. We concluded that the separation of this joint was likely caused by the shifting of the bow section from its initial settling position to its more overturned present position. The starboard sideshell and deck collapsed from the weight of the hull pressing down, ripping open this area and exposing the tank structure and piping, some of which may have been from the pump room located near the bow. This event also likely caused a large plume of oil to escape from the wreck and occurred sometime between the 1980s and 2009. Scallop dredge frames were noted in four places on the wreck, two of which are wedged against the north side of the bow section (Fig. 4).
The sea scallop fishery along the Mid-Atlantic coast of the United States from Virginia to Maine has existed since the 1870s and is one of the largest commercial fisheries of the east coast. The sea scallop, Placopecten magellanicus, is a bivalve mollusk that is found along the continental shelf of the northwest Atlantic Ocean from the Gulf to St. Lawrence to North Carolina at depths ranging from 18 to 110 m (Hart and Chute 2004). Scallop dredges consist of a triangular metal structure with a flat toothed bar along the bottom that is towed over the substrate that disturbs and lifts objects on the seabed and catching them in an attached netted bag (Boulcott et al. 2014). The top of the triangular metal frame is the towing bale, which often has a pair of rubber roller wheels that roll along the seabed and keep the front of the dredge off the seabed and the teeth angled downward (Rudders et al. 2020). The area of the Coimbra wreck, south of the eastern end of Long Island, is shown to have been a prolific fishery in bycatch statistics from 1979 to 2003, particularly in the area of the 50 m isobath, which is the depth of the wreck. This area is within the Mid-Atlantic Exemption Area where scallop dredge exemptions allow for dredge gear less than 3.2 m wide (NOAA 2020).
Evidence of fishing activity in the area is prominent on the site, including nets with roller chains from trawl nets hung up on the wreck. The four scallop dredges, consisting of the triangular metal towing bale and netting, were found impacted against the wreck, including two at the junction between the bow and midsection on the north side. These have been known to slice bows and sterns off wooden shipwrecks (Brennan 2016). There is a dredge rig at the junction of the midsection and bow, wedged in between them with a small boat anchor entrained with it. The presence of these four scallop dredges suggest that this is a frequent occurrence at the wreck, as not all dredges that strike here necessarily get hung up and are cut loose. Multiple strikes from these devices may have had an impact on the stability of the bow section and contributed to it being dislodged from its sinking position, further tipping it toward near-capsized. The shifting of this part of the wreck from an estimated angle of 70° to the seabed to 30°, further pushing the bow toward a capsized position, appears to have been caused by repeated strikes by scallop dredges. The dredges found impacted against the wreck are all on the north side, suggesting a pattern of traffic in this area in the direction of the hull's rotation. In addition, this rotation of the forward section of Coimbra appears to be what caused the long crack in the hull, compromising those forward tanks as the wreck shifted over and likely resulted in a large release of oil at the time.

SS Munger T. Ball
Built in 1920 in Savannah, GA, SS Munger T. Ball was constructed by Terry Shipbuilding as Lilmae, and later sold to Sabine Transportation Company in Port Arthur, TX (Fig. 5). The vessel was a 5200 gross ton class tanker and measured 391.9 feet in length with a beam of 51.2 feet. The tanker commonly sailed a route from Texas ports to the east coast. Ball departed Smith's Bluff, Texas headed for Wilmington, North Carolina with a cargo of 65,000 barrels of gasoline at the beginning of May 1942 (Barnette 2003, 117-118).
On May 5, the day after the German submarine U-507 sank the freighter Norlindo off the west coast of Florida, the U-boat sank two tankers in quick succession, Munger T. Ball and Joseph M. Cudahy. Ball was traveling unescorted and unarmed northwest of the Dry Tortugas when it was struck by two torpedoes, the first amidships and the second in the engine room. The gasoline cargo ignited, trapping many of the crew on board. Only four survived and the tanker sank quickly, about 15 min after the second torpedo struck (Wiggins 1995, 50). Cudahy, a 430-foot, 6950 ton tanker built by the Sun Shipbuilding Co. in Chester, PA in 1921, spotted the burning Munger T. Ball and turned north to head to Tampa in a zigzag pattern when a single torpedo from U-507 struck amidships. The tanker burned and drifted until May 7 when it was sunk by gunfire from the patrol yacht, USS Coral (PY 15) as a hazard to navigation (Wiggins 1995, 24, 52).
The wrecks of these two oil tankers sunk on 5 May 1942 have been located by divers and the deep sea fishing community, but have only been dived a handful of times due to the depths and distance from shore. Divers referred to one as the "Oil Wreck" and one as the "Phosphate Wreck" and assigned them the identities Joseph M. Cudahy and Munger T. Ball, respectively. This designation was based on visual damage apparent on the wrecks. A side-scan sonar survey in 2013 by NOAA also suggested the Oil Wreck was Cudahy based on the stern section having the appearance of breaking upon impact with the seabed. This report did not note the discrepancy in the ship's length with the length of this wreck in their sonar image, nor the torpedo damage on the hull amidships. NOAA's PPW assessment states: "While the reported scuttling position of the Joseph M. Cudahy places the wreck in deep water west of the Dry Tortugas, we believe the reported position is a publication error." Their conclusion of the wrecks' identities was considered tentative (NOAA 2013b).
During our site assessment, we noted that the dimensions of the shipwreck at the location of the Oil Wreck, which was still actively leaking, were too short to be Joseph M. Cudahy but were closely in line with those from the plans for Munger T. Ball. Aside from the overall length, this included other measurements, such as the distance from the bow to the superstructure, and the beam. Cudahy was reportedly struck on the starboard side amidships at tank #4; Ball was reportedly struck twice, once amidships and once at the engine room, on the port side. However, there are conflicting reports. One account suggests the torpedoes struck different sides. With only four survivors from the Ball, it is plausible the exact nature of the attack was misreported. Both tanker wrecks lie on their starboard side, so it could be suggested both were struck on their starboard sides.
Some points in the few descriptions of the wreck from divers also help support the switched identities of the "Oil" and "Phosphate" wrecks: divers on the deeper "Phosphate Wreck" stated the bottom of one of the forward tanks was blown outward, which appeared to be the only visible damage to the ship. This would be more consistent with the single torpedo strike on Cudahy. The diver mentions looking over the engine room skylights, which would not remain intact on a ship struck by a torpedo in the engine room (NOAA 2013b). During the ROV survey of the shallower "Oil Wreck", we observed damage in two places: amidships on the starboard hull and at the engine room, which sheared off and heavily damaged the stern (Fig. 6). This break was previously attributed to the stern-first impact with the seabed, but the damage to the stern section is considerable and likely caused by the second torpedo, especially since the stern of the "Phosphate Wreck" was described as remaining intact. The wreck here is also more consistent with a ship torpedoed twice that sank within 15 min as opposed to one that drifted on fire for two days; it's rapid sinking is why the stern section, while detached, did not drift far from the main section of the wreck.
The ROV investigation of the wreck in 2021 confirmed that the tanker lies on its starboard side and is mostly turtled. The methodology during this operation followed that conducted at Coimbra; the wreck was actively leaking during the operation (Fig. 7). The ROV inspection indicated the port deck edge is visible close to the seabed while the starboard side is buried. The deck here is badly corroded with holes visible throughout, exposing the interior of some of the summer tanks and upper frames where hull plating has fallen off (Fig. 8). The wreck is broken at the engine room and one of the boilers lies on the seabed. The break is clean, and the stern section lies only a few meters away, and is slightly canted, suggesting it separated upon impact rather than at the surface. This break likely occurred when the tanker sank stern first and impacted the seabed while the bow was still at the surface. The wreck is in water shallower than the length of the vessel. One of the tank seams at the keel is split near the bow. A large area of damage is prominent about halfway down the larger bow section, which is consistent with a torpedo strike around Tank #4 as reported in the historical account.
One propeller remains visible at the stern, which only has one blade remaining, suggesting possible salvage for the brass by divers. A flat section of wreck is located past the propeller on the seabed, which appears to be the broken rudder. Beginning at the stern, the ROV conducted multiple transects longitudinally across the wreck. Damage was observed forward of the amidships torpedo hole where hull plating is missing and ribs are exposed. This is apparent down to the mudline and extends toward the bow, likely encompassing the forward tanks. One transect followed the keel from bow to stern. A small section of keel plating is missing from the forward edge of the break at the engine room, and a large line of keel plating is missing opposite this on the aft side of the break. The transect along the hull followed the highest point of the wreck along the starboard bilge keel; fishing lines were noted draped over the turn of the bilge. Most significant was the observation that the hull was entirely intact from the break at the engine room forward to the torpedo damage amidships, while the bow and stern at either end are where most damage is visible. Visual inspection along the deck edge and seabed found the two masts; one is buried in the seabed while still attached to the deck, and was shoved into the seabed during the sinking. The second mast is lying flat on the seabed with coiled cables and what appear to be trawl doors tangled around it. A hawse pipe with chain running through it was visible near the forward mast. A cargo hatch and ladderway up to the accommodation deck was visible under the deck edge as well as a second hatch on the other side of that deck.
The fantail deck is on its starboard side with the deck rising vertically, and mooring line bitts are visible still attached. Among the debris on the seabed are sections of hull plating and exposed frame ends, as well as frames that once held the stern superstructure, which in historical photos appears to have a series of railings along it that could be part of these frame ends that we see. The damage to the stern section that is apparent here is indicative of a torpedo explosion and not just splitting from impact with the seabed, which is consistent with the Ball's sinking. A wheel or pulley is visible at the end of the fantail with chain looped through it. Inspection amidships looked for the remains of the pilothouse superstructure but there was none other than exposed frames where hull plating had been ripped off. The superstructure therefore appears to be crushed under the wreck, which is nearly upside down at this area.
In addition to potential diver removal of propeller blades, and possibly unknown artifacts as occurred at Coimbra, anthropogenic impacts to the wreck largely consist of fishing net hangs. The wreck is in an area of active shrimping grounds. Pink shrimp (Farfantepenaeus duorarum) comprises a valuable commercial fishery in southern Florida waters, which developed in the early 1950s in the area around the Dry Tortugas (Schmidt et al. 2004). The shrimp spawn in coastal shelf waters and migrate to nursery grounds in Florida Bay, and the shelf between Key West and offshore the Dry Tortugas support this fishery year-round (Criales et al. 2015). Commercial vessels primarily use double-winged trawls to catch the shrimp that typically inhabit sandy bottoms in depths from 9 to 44 m (Schmidt et al. 2004). Multiple trawl nets were observed hung up on the wreck of Munger T. Ball, but are tightly wrapped and most did not appear to be floating with lines in the water column, which would have been hazardous for ROV operations. Fishing nets and lines were visible lying flat on the seabed at the stern in the area of the rudder, and one net was caught below the propeller. A separate line was seen rising up into the water column from somewhere in the break at the stern, which was carefully avoided by the ROV. Netting and lines were draped over the port deck edge in the area near the masts. Weakening hull structure due to corrosion makes the wreck potentially more susceptible to the impacts of trawling, where trawl doors and weights could punch holes in the hull. Similar impacts, including roller chains wedged against the hull, were observed at the wreck of SS Coast Trader off Washington State during pollution risk assessment (Delgado et al. 2017).

Conclusions
The recently-remediated shipwreck sites of Coimbra and Munger T. Ball are important case studies in oil tanker shipwreck site assessment and characterization, and indicative of the likely condition of many merchant ships sunk during World War II. Active oil leaks from these wrecks were observed to be coming primarily from weakened seams with corroded rivets. If these wrecks are leaking from such areas, it is past time that an effort be made to locate and assess other shipwrecks on the PPW list, particularly those in deep water and farther offshore where oil slicks may not be as easily observed. While these two example oil mitigations were both in shallow water and accessible to commercial divers, the possibility exists for hull-tapping technology to be implemented by an ROV on wrecks in deep water. Such remediations have been successful, for example the wreck of Prestige that had 14,000 tons of heavy fuel oil removed from a depth of 3500 m after it sank in 2002 off northwestern Spain (Michel et al. 2005). Deep-water wrecks have a higher cost for finding and accessing them, but with now 80 years of corrosion acting on the hulls of tankers from World War II, the potential for catastrophic release of oil cargoes continues to grow. One thing remains clear: it is easier to recover oil from a rusted shipwreck than once it has spilled into the marine environment.