The Shipwreck Hunter

by David L. Mearns

  The oil sightings seemed like compelling clues to Derbyshire’s where abouts, but it was still far from certain that they were reliable markers of the wreck’s position on the seabed more than 4,000 metres below the surface. No one had ever used leaking oil to pinpoint a wreck’s location in water this deep before. The key question was how far the oil would be displaced downstream by deep oceanic currents and surface tides as it slowly floated to the surface. At least one famous deep-sea shipwreck hunter believed the information wasn’t sufficiently reliable. Dr Robert Ballard, who discovered the wrecks of the Titanic and the German battleship Bismarck, flatly ruled out a search for Derbyshire: ‘There just isn’t enough data available to locate the area the ship went down closely enough. I think that given the lack of data, the cost of finding the Derbyshire would just be prohibitive.’

  Although one scientific study commissioned by the ITF estimated that the downstream displacement of upwelling oil could be as much as ten nautical miles, I intuitively felt it would be a lot less. What convinced me was the fact that the oil was sighted over a period of at least five days, and the description of it bubbling up to the surface. In my mind this meant it had to be coming up from the wreck in a steady, consistent stream that suggested little displacement. However, because we only had an eight-day budget to complete the search, I wasn’t going to take any chances and decided to verify the JMSA data by visiting their offices in Okinawa and double-checking the SAR team reports.

  The theory of catastrophic structural failure espoused by the DFA was based on the belief that Derbyshire’s hull separated completely and circumferentially at, or very near to, the transverse bulkhead at frame 65. One of the reasons for this long-held belief was the relatively large distance between the sighting of oil by the helicopter and the other three sightings twenty-four nautical miles to the north-north-east. To the DFA and other supporters of the frame 65 theory, this was proof that Derbyshire’s hull had indeed split and that oil was upwelling from the two separated sections of the wreck on the seabed. As Derbyshire was generally heading north at the time, they also believed that the stern section of the ship was probably located at the helicopter oil-sighting position while the forward section, including all nine cargo holds and the bow, was located where the oil was bubbling up twenty-four nautical miles away.

  To cover this theory, I originally devised a three-phase plan in which a wide-area search for the forward section of the wreck would be followed by the high-resolution sonar imaging of any wreckage found. As long as there was time remaining in the allotted eight-day search budget, which couldn’t be guaranteed, we would then attempt to locate the stern section of the wreck. This plan was accepted and approved by the ITF, but it changed completely after my visit to the JMSA in Okinawa.

  I had specifically asked the JMSA to pull out all the original records, including the SAR team reports on the four oil sightings, and to have them ready when I arrived at their office. After I explained the importance of this information to our search, they handed me a single-page summary that cross-checked their records against the four positions I had given them, which came straight out of the court’s investigation report published in 1989. All the information seemed to tally except for the sighting made by the SAR helicopter on 15 September in position 25°30’ north, 133°18’ east. Against this entry the JMSA summary had only two words: ‘no record’. When I asked what this meant, they simply said they had nothing in their records about a helicopter being used in this SAR mission. I needed to be absolutely sure of this so I pressed them again and again in case we were having a language misunderstanding, but each time they assured me they had absolutely no information about a helicopter being involved.

  I hadn’t expected this at all, but when I got back to my hotel and pulled out a large-scale chart of the area, I could see why no helicopter had been used in the search: it was because Derbyshire was lost so far from land. The JMSA operation had been organized from their regional base at Naha, Okinawa, which is 310 nautical miles from the supposed helicopter oil sighting as the crow flies. This distance (620 nautical miles minimum round-trip) would have been at the far extreme of the longest-range SAR helicopters available in 1980, which meant that if one had been used, it would have had precious little time to search once it reached the location.

  The helicopter oil sighting was an obvious piece of misinformation that had been accepted as fact for all these years simply because it was published in what people believed was an authoritative report. This was a hugely significant discovery because it meant we’d only have to search one location for the wreckage, not two. On the other hand, it went against what the DFA had believed about Derbyshire splitting and the two sections sinking in different locations, though I couldn’t let that concern me. When, a little later, I discovered the source of the helicopter position, all the pieces of the puzzle fell into place.

  At the time of the loss, Bibby Line had received a fax from their agent in Japan, Mr I. Yamada, which read: ‘Now reported us that they heard from helicopter. Quote: found some oil spot afloating [sic] at the position of 25.5 north 133.3 east hence trying further research nearby sea water surface. Unquote.’ As no helicopter was involved in the search, this report clearly refers to the sighting made by aircraft 791 in position 25°50’ north, 133°30’ east, except that an error was made when the position was quoted in decimal degrees of latitude and longitude rather than in degrees and minutes. As this information was relayed down the line, someone within Bibby Line naturally, and quite correctly, converted the 25.5 north, 133.3 east position into 25°30’ north, 133°18’ east without knowing of the earlier error. From this little piece of detective work I left Okinawa feeling 100 per cent certain that the helicopter oil sighting was essentially a phantom position and could be ignored. It also cleared up something else that hadn’t made sense to me. I never understood how Derbyshire’s forward section, minus the engine and propulsion, was supposed to have been able to travel twenty-four nautical miles further north-north-east against the southerly winds and seas of Typhoon Orchid. There was still every chance that Derbyshire split at frame 65, but it was beginning to look as if the scenario presumed by the DFA was impossible and we’d find the two sections of the wreck in the same location, close to the three confirmed sightings of upwelling oil.

  We got under way from Yokohama on the afternoon of 26 May, expecting it to take anywhere from two and a half to three days to complete the 660-nautical-mile journey to the search area. Unfortunately we set out in near-gale Force 7 conditions, which made it impossible to get any additional work done the first day of the transit. The cause of the poor weather was a tropical depression whose predicted track looked like it could be a real problem for us until thankfully it began to veer away. When you haven’t been to sea for a while, the last thing you want on the first day out is rough weather, because it makes it very difficult to get your sea legs and become accustomed to the ship’s movements. The constant pitching and rolling of the ship made the transit a bit unpleasant and laid low a couple of guys in our crew.

  Overall there were eleven of us on board, in addition to the Japanese marine crew: Mark Dickinson of the ITF, Dr Chris Davies of the University of Wales, Rory Maclean, the ITN news reporter, and myself and seven others from Oceaneering. I was the search director and supervised one shift, while the project manager Craig Bagley supervised the other. We had been forced to keep our crew size small because of ITF’s tight budget and the limited number of bunks on the Shin Kai Maru, the Japanese platform supply vessel we had hired for the project. However, the team was packed with some of the best and most experienced guys in our commercial group, and we had also brought with us a surprise the ITF was not counting on.

  Our contract with the ITF called on us to provide only the Ocean Explorer 6000 side-scan sonar. A lack of funds meant they could not afford any ROV dives to investigate or positively identify whatever was found during the search. Instead, they were counting on the high-resolution sonar images of the wrec
k to be distinctive enough to prove that Derbyshire had definitely been found and that this information met the criteria of ‘new and important evidence’ demanded by the government in order for the investigation to be reopened. From Oceaneering’s point of view, this was a poor judgement. As the reading of side-scan sonar images is an interpretive skill, the conclusions that can be drawn from it are open to debate and thus are nowhere near as definitive as the visual images taken by an ROV. Our concern was that by relying on the side-scan sonar imagery alone, the ITF was running the risk of having the government reject out of hand any evidence they put to them because the sonar imagery was deemed too questionable to be accepted as definitive proof.

  We felt so strongly about this that we decided to bring the Magellan 725 to Japan with us, at no additional cost to the ITF, just in case it was needed. I stopped short of telling Mark about it until it was about to be loaded on board the Shin Kai Maru in Yokohama. My reasoning was that I didn’t want to raise expectations too high, and that it was far better for the Magellan to be viewed as an unexpected bonus – in case it was needed to take pictures of something we found – rather than an unfulfilled promise. When I opened the container doors to reveal the ROV to Mark, our surprise had the desired effect. All he could say was ‘Thank you’ and ‘Wow, that’s great. Now I just hope we find something to dive her on.’

  Despite the rough seas, we actually made very good time during the transit and arrived on site fourteen hours ahead of schedule. That was the good news. The bad news was that the weather conditions were still too poor to launch the Ocean Explorer, so the time we gained against our schedule during the transit was subsequently lost waiting for the weather to improve. Nothing could be done about it other than to use the downtime to train the ship’s crew in running track-lines so they were ready when we needed them for the real thing. That occurred the following day, 29 May, when the Ocean Explorer was finally deployed in the early afternoon to begin the eight-hour descent towards the start of the first of seven potential track-lines. With the hopes of at least forty-four families in the UK and countless seafarers around the world riding on the outcome, the long-anticipated search for Derbyshire was finally under way.

  When the first images of the sea floor started scrolling down my screen, I wasn’t surprised to see it was a geologically active area with plenty of relief and outcropping rocks. If the upwelling oil was to be believed, Derbyshire had sunk somewhere along a steep slope that separated the Daito Ridge from an adjacent basin bearing the same name. Smack in the middle of the 200 square nautical miles I had designated for the search, the seabed dropped precipitously from 3,500 metres to well over 5,000 metres deep. Even with her enormous size, picking Derbyshire’s wreck out from all this geology was going to be extremely difficult.

  I purposely placed this first track-line (no. 4 in my grid of seven) just north of the three upwelling oil positions and parallel to the sloping seabed to make it easier for us to tow the sonar on an even plane. Although an oceanographer contracted by the ITF advised that the upwelling oil was probably drifting north with the Kuroshio current and that this indicated the wreck was located to the south, I had found more recent data during my research in Japan that indicated exactly the opposite: that the flow was actually south at about one knot, in keeping with a countercurrent flowing against the main direction of the Kuroshio. This data, along with the JMSA SAR team descriptions of the oil bubbling up, convinced me that the wreck was close by and located probably just north of the reported sightings. My best guess was that the likely displacement of the oil was no more than three nautical miles, and this was the figure I used to establish the high-probability zone within the overall search box. If I was right, there was a good chance we’d find the wreck on one of the first three track-lines.

  Four and a half hours into the search, we started picking up suspect targets at far range on the sonar’s starboard channel. The targets were clustered and clearly hard, but it was impossible to discern any recognizable shape or structure because they were at such a distance. I liked the fact that they were located to the north and midway between the two aircraft positions, but it was much too early to start getting excited. Given the order in which I had decided to run the search lines (4, 6, 3, 5, 2, 7 and finally 1), we’d get a much better look at these targets on the next most northern track-line: no. 3 in my grid and also the third line in sequence to be searched. Before that, line 6 was searched, but when nothing of note was found, it simply heightened our expectations for what we might see on the next line. Although the search was only into its second day, it was progressing very well and there was a definite sense that we were closing in on Derbyshire’s location.

  The crew driving the Shin Kai Maru had already executed the most precarious part of the turn to line 3 and were steadying the ship on course when we suffered a disaster that jeopardized the whole project. The first sign that something was wrong was a spike of noise on the sonar’s display. Within seconds, lights indicating an open circuit in the electrical core of the sonar tow cable flickered and then lit up brightly. There had been no warnings of any problems, but these symptoms meant only one thing: the tow cable had developed a fault, probably a minor ingress of water, and the sonar towfish would have to be recovered, meaning the search was over for now. As we trudged out of our operations van, disappointed at having to stop before learning what the targets on line 3 were but still thinking the problem was relatively minor, someone on the back deck shouted, ‘Hurry, before the cable snaps!’

  Short of snapping completely, the worst thing that can happen with a tow cable under tension is for some of the steel wire strands to break and ‘birdcage’ at the point where the cable exits the ship before entering the water. Seeing a cable in this condition will make your heart sink, especially when 8,000 metres of cable and your only sonar towfish are just one sudden heave of the ship’s stern away from being lost forever. The only way to save a situation like this is to gingerly haul the damaged cable back on board until it is secure on a winch, thus releasing the huge loads on the damaged section. The real trick is getting the fat birdcage past the sheaves and narrow openings through which the normally thin cable passes without making things worse or snapping any more strands.

  As everyone on my shift sprang into action, I ran up to the bridge to explain to the captain that I wanted him to maintain the ship’s speed but to change the heading in order to reduce the rolling and heaving motion of the stern as much as possible. I knew the cable already had three to four tons of load on it, but that could easily double or treble with one bad heave of the stern, causing it to snap instantly. By the time I got back to the main deck, two of our most experienced crew, Greg Gibson and Ron Schmidt, had taken control and were easing the cable over the three large internal sheaves built within our deployment crane. Once the damaged section was inboard, I could see that the break was very bad, involving strands from both the outer and inner layers of steel wire that gave the cable its strength. I could hardly believe that so few unbroken strands – perhaps a quarter of the total – were keeping the full length of the cable and the Ocean Explorer from snapping free.

  When a serious failure like this occurs at sea, it is hard not to panic. Though the potential for huge financial losses is staring you in the face, you must push it to the back of your mind and work quickly but calmly. A bad decision or action born out of panic can jeopardize not only your equipment, but someone’s life. To avoid anyone getting hurt, we had to make sure everyone was forward of the failed section in case the cable snapped and whiplashed as it left the ship. The scariest part is not knowing when this might happen. You could be literally an inch away from saving the whole rig when just one more strand snaps, causing the cable to reach its breakpoint. Fortunately, we had two of our best crew feverishly trying to save the situation from becoming a complete disaster.

  Greg and Ron had already done an amazing job, but the worst bit was still to come. In order to wind the cable up onto our traction winch, where
it could be securely held, they needed to trim all the badly frayed strands of wire so that the cable could ride cleanly in the narrow grooves of the bull wheels. As Ron lay on his stomach to grind away each of the dozen or so broken strands, his hands and face were in constant peril should the cable suddenly snap and recoil back in his direction. This selfless and brave effort to save the rest of the cable and the Ocean Explorer sonar was typical of the man. In all my years at sea I can’t recall being more relieved than I was to see what was left of the cable wind up on the winch and Ron stand up with all his fingers intact.

  Remarkably, less than nineteen hours of search time was lost while the Ocean Explorer sonar was recovered, the short end of the tow cable jettisoned and the long end wound back up on to its storage drum and electrically reconnected. We were left with about 8,000 metres of usable cable, which would be just enough to complete the search at slightly slower speeds. While we were dealing with the tow cable disaster, I had stopped thinking about the suspect targets we had found on line 4, but now that we were up and running again, I couldn’t wait to see what they looked like from a better position on line 3. However, once the sonar had passed over the target area, I was left scratching my head about what the targets actually were.

  Given that we were searching for the largest ship ever to have sunk, I was expecting to find at least some of its structure intact and easily recognizable. What we found instead could only be described as a very large patch of targets: some small, some large, but nothing on a scale that allowed me to be sure this was a shipwreck. In fact it looked more like the debris field from a plane crash, where the entire fuselage is obliterated after hitting the water surface at speed. If this was the remains of the Derbyshire, none of us could imagine what had caused the hull to be so completely shattered.