The Titan Submersible Implosion Was “an Accident Waiting to Happen”

By Ben Taub

The primary task of a submersible is to not implode. The second is to reach the surface, even if the pilot is unconscious, with oxygen to spare. The third is for the occupants to be able to open the hatch once they surface. The fourth is for the submersible to be easy to find, through redundant tracking and communications systems, in case rescue is required. Only the fifth task is what is ordinarily thought of as the primary one: to transport people into the dark, hostile deep.

At dawn four summers ago, the French submariner and Titanic expert Paul-Henri Nargeolet stood on the bow of an expedition vessel in the North Atlantic. The air was cool and thick with fog, the sea placid, the engine switched off, and the Titanic was some thirty-eight hundred metres below. The crew had gathered for a solemn ceremony, to pay tribute to the more than fifteen hundred people who had died in the most famous maritime disaster more than a hundred years ago. Rob McCallum, the expedition leader, gave a short speech, then handed a wreath to Nargeolet, the oldest man on the ship. As is tradition, the youngest—McCallum’s nephew—was summoned to place his hand on the wreath, and he and Nargeolet let it fall into the sea.

Inside a hangar on the ship’s stern sat a submersible known as the Limiting Factor. In the previous year, McCallum, Nargeolet, and others had taken it around the Earth, as part of the Five Deeps Expedition, a journey to the deepest point in each ocean. The team had mapped unexplored trenches and collected scientific samples, and the Limiting Factor’s chief pilot, Victor Vescovo—a Texan hedge-fund manager who had financed the entire operation—had set numerous diving records. But, to another member of the expedition team, Patrick Lahey, the C.E.O. of Triton Submarines (which had designed and built the submersible), one record meant more than the rest: the marine-classification society DNV had certified the Limiting Factor’s “maximum permissible diving depth” as “unlimited.” That process was far from theoretical; a DNV inspection engineer was involved in every stage of the submersible’s creation, from design to sea trials and diving. He even sat in the passenger seat as Lahey piloted the Limiting Factor to the deepest point on Earth.

After the wreath sank from view, Vescovo climbed down the submersible hatch, and the dive began. For some members of the crew, the site of the wreck was familiar. McCallum, who co-founded a company called EYOS Expeditions, had transported tourists to the Titanic in the two-thousands, using two Soviet submarines that had been rated to six thousand metres. Another crew member was a Titanic obsessive—his endless talk of davits and well decks still rattles in my head. But it was Paul-Henri Nargeolet whose life was most entwined with the Titanic. He had dived it more than thirty times, beginning shortly after its discovery, in 1985, and now served as the underwater-research director for the organization that owns salvaging rights to the wreck.

Nargeolet had also spent the past year as Vescovo’s safety manager. “When I set out on the Five Deeps project, I told Patrick Lahey, ‘Look, I don’t know submarine technology—I need someone who works for me to independently validate whatever design you come up with, and its construction and operation,’ ” Vescovo recalled, this week. “He recommended P. H. Nargeolet, whom he had known for decades.” Nargeolet, whose wife had recently died, was a former French naval commander—an underwater-explosives expert who had spent much of his life at sea. “He had a sterling reputation, the perfect résumé,” Vescovo said. “And he was French. And I love the French.”

Ben Taub’s 2020 account of an expedition to the deepest point in each ocean.

When Vescovo reached the silty bottom at the Titanic site, he recalled his private preparations with Nargeolet. “He had very good knowledge of the currents and the wreck,” Vescovo told me. “He briefed me on very specific tactical things: ‘Stay away from this place on the stern’; ‘Don’t go here’; ‘Try and maintain this distance at this part of the wreck.’ ” Vescovo surfaced about seven hours later, exhausted and rattled from the debris that he had encountered at the ship’s ruins, which risk entangling submersibles that approach too close. But the Limiting Factor was completely fine. According to its certification from DNV, a “deep dive,” for insurance and inspection purposes, was anything below four thousand metres. A journey to the Titanic, thirty-eight hundred metres down, didn’t even count.

Nargeolet remained obsessed with the Titanic, and, before long, he was invited to return. “To P. H., the Titanic was Ulysses’ sirens—he could not resist it,” Vescovo told me. A couple of weeks ago, Nargeolet climbed into a radically different submersible, owned by a company called OceanGate, which had spent years marketing to the general public that, for a fee of two hundred and fifty thousand dollars, it would bring people to the most famous shipwreck on Earth. “People are so enthralled with Titanic,” OceanGate’s founder, Stockton Rush, told a BBC documentary crew last year. “I read an article that said there are three words in the English language that are known throughout the planet. And that’s ‘Coca-Cola,’ ‘God,’ and ‘Titanic.’ ”

Nargeolet served as a guide to the wreck, Rush as the pilot. The other three occupants were tourists, including a father and son. But, before they reached the bottom, the submersible vanished, triggering an international search-and-rescue operation, with an accompanying media frenzy centered on counting down the hours until oxygen would run out.

McCallum, who was leading an expedition in Papua New Guinea at the time, knew the outcome almost instantly. “The report that I got immediately after the event—long before they were overdue—was that the sub was approaching thirty-five hundred metres,” he told me, while the oxygen clock was still ticking. “It dropped weights”—meaning that the team had aborted the dive—“then it lost comms, and lost tracking, and an implosion was heard.”

An investigation by the U.S. Coast Guard is ongoing; some debris from the wreckage has been salvaged, but the implosion was so violent and comprehensive that the precise cause of the disaster may never be known.

Until June 18th, a manned deep-ocean submersible had never imploded. But, to McCallum, Lahey, and other experts, the OceanGate disaster did not come as a surprise—they had been warning of the submersible’s design flaws for more than five years, filing complaints to the U.S. government and to OceanGate itself, and pleading with Rush to abandon his aspirations. As they mourned Nargeolet and the other passengers, they decided to reveal OceanGate’s history of knowingly shoddy design and construction. “You can’t cut corners in the deep,” McCallum had told Rush. “It’s not about being a disruptor. It’s about the laws of physics.”

Stockton Rush was named for two of his ancestors who signed the Declaration of Independence: Richard Stockton and Benjamin Rush. His maternal grandfather was an oil-and-shipping tycoon. As a teen-ager, Rush became an accomplished commercial jet pilot, and he studied aerospace engineering at Princeton, where he graduated in 1984.

Rush wanted to become a fighter pilot. But his eyesight wasn’t perfect, and so he went to business school instead. Years later, he expressed a desire to travel to space, and he reportedly dreamed of becoming the first human to set foot on Mars. In 2004, Rush travelled to the Mojave Desert, where he watched the launch of the first privately funded aircraft to brush against the edge of space. The only occupant was the test pilot; nevertheless, as Rush used to tell it, Richard Branson stood on the wing and announced that a new era of space tourism had arrived. At that point, Rush “abruptly lost interest,” according to a profile in Smithsonian magazine. “I didn’t want to go up into space as a tourist,” he said. “I wanted to be Captain Kirk on the Enterprise. I wanted to explore.”

Rush had grown up scuba diving in Tahiti, the Cayman Islands, and the Red Sea. In his mid-forties, he tinkered with a kit for a single-person mini-submersible, and piloted it around at shallow depths near Seattle, where he lived. A few years later, in 2009, he co-founded OceanGate, with a dream to bring tourists to the ocean world. “I had come across this business anomaly I couldn’t explain,” he recalled. “If three-quarters of the planet is water, how come you can’t access it?”

OceanGate’s first submersible wasn’t made by the company itself; it was built in 1973, and Lahey later piloted it in the North Sea, while working in the oil-and-gas industry. In the nineties, he helped refit it into a tourist submersible, and in 2009, after it had been sold a few times, and renamed Antipodes, OceanGate bought it. “I didn’t have any direct interaction with them at the time,” Lahey recalled. “Stockton was one of these people that was buying these older subs and trying to repurpose them.”

In 2015, OceanGate announced that it had built its first submersible, in collaboration with the University of Washington’s Applied Physics Laboratory. In fact, it was mostly a cosmetic and electrical refit; Lahey and his partners had built the underlying vessel, called Lula, for a Portuguese marine research nonprofit almost two decades before. It had a pressure hull that was the shape of a capsule pill and made of steel, with a large acrylic viewport on one end. It was designed to go no deeper than five hundred metres—a comfortable cruising depth for military submarines. OceanGate now called it Cyclops I.

Most submersibles have duplicate control systems, running on separate batteries—that way, if one system fails, the other still works. But, during the refit, engineers at the University of Washington rigged the Cyclops I to run from a single PlayStation 3 controller. “Stockton is very interested in being able to quickly train pilots,” Dave Dyer, a principal engineer, said, in a video published by his laboratory. Another engineer referred to it as “a combination steering wheel and gas pedal.”

Around that time, Rush set his sights on the Titanic. OceanGate would have to design a new submersible. But Rush decided to keep most of the design elements of Cyclops I. Suddenly, the University of Washington was no longer involved in the project, although OceanGate’s contract with the Applied Physics Laboratory was less than one-fifth complete; it is unclear what Dyer, who did not respond to an interview request, thought of Rush’s plan to essentially reconstruct a craft that was designed for five hundred metres of pressure to withstand eight times that much. As the company planned Cyclops II, Rush reached out to McCallum for help.

“He wanted me to run his Titanic operation for him,” McCallum recalled. “At the time, I was the only person he knew who had run commercial expedition trips to Titanic. Stockton’s plan was to go a step further and build a vehicle specifically for this multi-passenger expedition.” McCallum gave him some advice on marketing and logistics, and eventually visited the workshop, outside Seattle, where he examined the Cyclops I. He was disturbed by what he saw. “Everyone was drinking Kool-Aid and saying how cool they were with a Sony PlayStation,” he told me. “And I said at the time, ‘Does Sony know that it’s been used for this application? Because, you know, this is not what it was designed for.’ And now you have the hand controller talking to a Wi-Fi unit, which is talking to a black box, which is talking to the sub’s thrusters. There were multiple points of failure.” The system ran on Bluetooth, according to Rush. But, McCallum continued, “every sub in the world has hardwired controls for a reason—that if the signal drops out, you’re not fucked.”

One day, McCallum climbed into the Cyclops for a test dive at a marina. There, he met the chief pilot, David Lochridge, a Scotsman who had spent three decades as a submersible pilot and an engineer—first in the Royal Navy, then as a private contractor. Lochridge had worked all over the world: on offshore wind farms in the North Sea; on subsea-cables installations in the Atlantic, Indian, and Pacific oceans; on manned submarine trials with the Swedish Navy; on submarine-rescue operations for the navies of Britain and Singapore. But, during the harbor trial, the Cyclops got stuck in shallow water. “It was hilarious, because there were four very experienced operators in the sub, stuck at twenty or twenty-five feet, and we had to sit there for a few hours while they worked it out,” McCallum recalled. He liked and trusted Lochridge. But, of the sub, he said, “This thing is a mutt.”

Rush eventually decided that he would not attempt to have the Titanic-bound vehicle classed by a marine-certification agency such as DNV. He had no interest in welcoming into the project an external evaluator who would, as he saw it, “need to first be educated before being qualified to ‘validate’ any innovations.”

That marked the end of McCallum’s desire to be associated with the project. “The minute that I found out that he was not going to class the vehicle, that’s when I said, ‘I’m sorry, I just can’t be involved,’ ” he told me. “I couldn’t tell him anything about the Five Deeps project at that time. But I was able to say, ‘Look, I am involved with other projects that are building classed subs’—of course, I was talking about the Limiting Factor—‘and I can tell you that the class society has been nothing but supportive. They are actually part of our innovation process. We’re using the brainpower of their engineers to feed into our design.

“Stockton didn’t like that,” McCallum continued. “He didn’t like to be told that he was on the fringe.” As word got out that Rush planned to take tourists to the Titanic, McCallum recalled, “people would ring me, and say, ‘We’ve always wanted to go to Titanic. What do you think?’ And I would tell them, ‘Never get in an unclassed sub. I wouldn’t do it, and you shouldn’t, either.’ ”

In early 2018, McCallum heard that Lochridge had left OceanGate. “I’d be keen to pick your brain if you have a few moments,” McCallum e-mailed him. “I’m keen to get a handle on exactly how bad things are. I do get reports, but I don’t know if they are accurate.” Whatever his differences with Rush, McCallum wanted the venture to succeed; the submersible industry is small, and a single disaster could destroy it. But the only way forward without a catastrophic operational failure—which he had been told was “certain,” he wrote—was for OceanGate to redesign the submersible in coördination with a classification society. “Stockton must be gutted,” McCallum told Lochridge, of his departure. “You were the star player . . . . . and the only one that gave me a hint of confidence.”

“I think you are going to [be] even more taken aback when I tell you what’s happening,” Lochridge replied. He added that he was afraid of retaliation from Rush—“We both know he has influence and money”—but would share his assessment with McCallum, in private: “That sub is Not safe to dive.”

“Do you think the sub could be made safe to dive, or is it a complete lemon?” McCallum replied. “You will get a lot of support from people in the industry . . . . everyone is watching and waiting and quietly shitting their pants.”

“It’s a lemon.”

“Oh dear,” McCallum replied. “Oh dear, oh dear.”

Lochridge had been hired by OceanGate in May, 2015, as its director of marine operations and chief submersible pilot. The company moved him and his family to Washington, and helped him apply for a green card. But, before long, he was clashing with Rush and Tony Nissen, the company’s director of engineering, on matters of design and safety.

Every aspect of submersible design and construction is a trade-off between strength and weight. In order for the craft to remain suspended underwater, without rising or falling, the buoyancy of each component must be offset against the others. Most deep-ocean submersibles use spherical titanium hulls and are counterbalanced in water by syntactic foam, a buoyant material made up of millions of hollow glass balls, which is attached to the external frame. But this adds bulk to the submersible. And the weight of titanium limits the practical size of the pressure hull, so that it can accommodate no more than two or three people. Spheres are “the best geometry for pressure, but not for occupation,” as Rush put it.

The Cyclops II needed to fit as many passengers as possible. “You don’t do the coolest thing you’re ever going to do in your life by yourself,” Rush told an audience at the GeekWire Summit last fall. “You take your wife, your son, your daughter, your best friend. You’ve got to have four people” besides the pilot. Rush planned to have room for a Titanic guide and three passengers. The Cyclops II could fit that many occupants only if it had a cylindrical midsection. But the size dictated the choice of materials. The steel hull of Cyclops I was too thin for Titanic depths—but a thicker steel hull would add too much weight. In December, 2016, OceanGate announced that it had started construction on Cyclops II, and that its cylindrical midsection would be made of carbon fibre. The idea, Rush explained in interviews, was that carbon fibre was a strong material that was significantly lighter than traditional metals. “Carbon fibre is three times better than titanium on strength-to-buoyancy,” he said.

A month later, OceanGate hired a company called Spencer Composites to build the carbon-fibre hull. “They basically said, ‘This is the pressure we have to meet, this is the factor of safety, this is the basic envelope. Go design and build it,’ ” the founder, Brian Spencer, told CompositesWorld, in the spring of 2017. He was given a deadline of six weeks.

Toward the end of that year, Lochridge became increasingly concerned. OceanGate would soon begin manned sea trials for Cyclops II in the Bahamas, and he believed that there was a chance that they would result in catastrophe. The consequences for Lochridge could extend beyond OceanGate’s business and the trauma of losing colleagues; as director of marine operations, Lochridge had a contract specifying that he was ultimately responsible for “ensuring the safety of all crew and clients.”

On the workshop floor, he raised questions about potential flaws in the design and build processes. But his concerns were dismissed. OceanGate’s position was that such matters were outside the scope of his responsibilities; he was “not hired to provide engineering services, or to design or develop Cyclops II,” the company later said, in a court filing. Nevertheless, before the handover of the submersible to the operations team, Rush directed Lochridge to carry out an inspection, because his job description also required him to sign off on the submersible’s readiness for deployment.

On January 18, 2018, Lochridge studied each major component, and found several critical aspects to be defective or unproven. He drafted a detailed report, which has not previously been made public, and attached photographs of the elements of greatest concern. Glue was coming away from the seams of ballast bags, and mounting bolts threatened to rupture them; both sealing faces had errant plunge holes and O-ring grooves that deviated from standard design parameters. The exostructure and electrical pods used different metals, which could result in galvanic corrosion when exposed to seawater. The thruster cables posed “snagging hazards”; the iridium satellite beacon, to transmit the submersible’s position after surfacing, was attached with zip ties. The flooring was highly flammable; the interior vinyl wrapping emitted “highly toxic gasses upon ignition.”

To assess the carbon-fibre hull, Lochridge examined a small cross-section of material. He found that it had “very visible signs of delamination and porosity”—it seemed possible that, after repeated dives, it would come apart. He shone a light at the sample from behind, and photographed beams streaming through splits in the midsection in a disturbing, irregular pattern. The only safe way to dive, Lochridge concluded, was to first carry out a full scan of the hull.

The next day, Lochridge sent his report to Rush, Nissen, and other members of the OceanGate leadership. “Verbal communication of the key items I have addressed in my attached document have been dismissed on several occasions, so I feel now I must make this report so there is an official record in place,” he wrote. “Until suitable corrective actions are in place and closed out, Cyclops 2 (Titan) should not be manned during any of the upcoming trials.”

Rush was furious; he called a meeting that afternoon, and recorded it on his phone. For the next two hours, the OceanGate leadership insisted that no hull testing was necessary—an acoustic monitoring system, to detect fraying fibres, would serve in its place. According to the company, the system would alert the pilot to the possibility of catastrophic failure “with enough time to arrest the descent and safely return to surface.” But, in a court filing, Lochridge’s lawyer wrote, “this type of acoustic analysis would only show when a component is about to fail—often milliseconds before an implosion—and would not detect any existing flaws prior to putting pressure onto the hull.” A former senior employee who was present at the meeting told me, “We didn’t even have a baseline. We didn’t know what it would sound like if something went wrong.”

OceanGate’s lawyer wrote, “The parties found themselves at an impasse—Mr. Lochridge was not, and specifically stated that he could not be made comfortable with OceanGate’s testing protocol, while Mr. Rush was unwilling to change the company’s plans.” The meeting ended in Lochridge’s firing.

Soon afterward, Rush asked OceanGate’s director of finance and administration whether she’d like to take over as chief submersible pilot. “It freaked me out that he would want me to be head pilot, since my background is in accounting,” she told me. She added that several of the engineers were in their late teens and early twenties, and were at one point being paid fifteen dollars an hour. Without Lochridge around, “I could not work for Stockton,” she said. “I did not trust him.” As soon as she was able to line up a new job, she quit.

“I would consider myself pretty ballsy when it comes to doing things that are dangerous, but that sub is an accident waiting to happen,” Lochridge wrote to McCallum, two weeks later. “There’s no way on earth you could have paid me to dive the thing.” Of Rush, he added, “I don’t want to be seen as a Tattle tale but I’m so worried he kills himself and others in the quest to boost his ego.”

McCallum forwarded the exchange to Patrick Lahey, the C.E.O. of Triton Submarines, whose response was emphatic: if Lochridge “genuinely believes this submersible poses a threat to the occupants,” then he had a moral obligation to inform the authorities. “To remain quiet makes him complicit,” Lahey wrote. “I know that may sound ominous but it is true. History is full of horrific examples of accidents and tragedies that were a direct result of people’s silence.”

OceanGate claimed that Cyclops II had “the first pressure vessel of its kind in the world.” But there’s a reason that Triton and other manufacturers don’t use carbon fibre in their hulls. Under compression, “it’s a capricious fucking material, which is the last fucking thing you want to associate with a pressure boundary,” Lahey told me.

“With titanium, there’s a purpose to a pressure test that goes beyond just seeing whether it will survive,” John Ramsay, the designer of the Limiting Factor, explained. The metal gradually strengthens under repeated exposure to incredible stresses. With carbon fibre, however, pressure testing slowly breaks the hull, fibre by tiny fibre. “If you’re repeatedly nearing the threshold of the material, then there’s just no way of knowing how many cycles it will survive,” he said.

“It doesn’t get more sensational than dead people in a sub on the way to Titanic,” Lahey’s business partner, the co-founder of Triton Submarines, wrote to his team, on March 1, 2018. McCallum tried to reason with Rush directly. “You are wanting to use a prototype un-classed technology in a very hostile place,” he e-mailed. “As much as I appreciate entrepreneurship and innovation, you are potentially putting an entire industry at risk.”

Rush replied four days later, saying that he had “grown tired of industry players who try to use a safety argument to stop innovation and new entrants from entering their small existing market.” He understood that his approach “flies in the face of the submersible orthodoxy, but that is the nature of innovation,” he wrote. “We have heard the baseless cries of ‘you are going to kill someone’ way too often. I take this as a serious personal insult.”

In response, McCallum listed a number of specific concerns, from his “humble perch” as an expedition leader. “In your race to Titanic you are mirroring that famous catch cry ‘she is unsinkable,’ ” McCallum wrote. The correspondence ended soon afterward; Rush asked McCallum to work for him—then threatened him with a lawsuit, in an effort to silence him, when he declined.

By now, McCallum had introduced Lochridge to Lahey. Lahey wrote him, “If Ocean Gate is unwilling to consider or investigate your concerns with you directly perhaps some other means of getting them to pay attention is required.”

Lochridge replied that he had already contacted the United States Department of Labor, alleging to its Occupational Safety and Health Administration that he had been terminated in retaliation for raising safety concerns. He also sent the OSHA investigator Paul McDevitt a copy of his Cyclops II inspection report, hoping that the government might take actions that would “prevent the potential for harm to life.”

A few weeks later, McDevitt contacted OceanGate, noting that he was looking into Lochridge’s firing as a whistle-blower-protection matter. OceanGate’s lawyer Thomas Gilman soon issued Lochridge a court summons: he had ten days to withdraw his OSHA claim and pay OceanGate almost ten thousand dollars in legal expenses. Otherwise, Gilman wrote, OceanGate would sue him, take measures to destroy his professional reputation, and accuse him of immigration fraud. Gilman also reported to OSHA that Lochridge had orchestrated his own firing because he “wanted to leave his job and maintain his ability to collect unemployment benefits.” (McDevitt, of OSHA, notified the Coast Guard of Lochridge’s complaint. There is no evidence that the Coast Guard ever followed up.)

Lochridge received the summons while he was at his father’s funeral. He and his wife hired a lawyer, but it quickly became clear that “he didn’t have the money to fight this guy,” Lahey told me. (Lochridge declined to be interviewed.) Lahey covered the rest of the expenses, but, after more than half a year of legal wrangling, and threats of deportation, Lochridge withdrew his whistle-blower claim with OSHA so that he could go on with his life. Lahey was crestfallen. “He didn’t consult me about that decision,” Lahey recalled. “It’s not that he had to—it was his fight, not mine. But I was underwriting the cost of it, because I believed in the idea that this inspection report, which he wouldn’t share with anybody, needed to see the light of day.”

A few weeks after Lochridge was fired, OceanGate announced that it was testing its new submersible in the marina of Everett, Washington, and would soon begin shallow-water trials in Puget Sound. To preëmpt any concerns about the carbon-fibre hull, the company touted the acoustic monitoring system, which was later patented in Rush’s name. “Safety is our number one priority,” Rush said, in an OceanGate press release. “We believe real-time health monitoring should be standard safety equipment on all manned-submersibles.”

“He’s spinning the fact that his sub requires a hull warning system into something positive,” Jarl Stromer, Triton’s regulatory and class-compliance manager, reported to Lahey. “He’s making it sound like the Cyclops is more advanced because it has this system when the opposite is true: The submersible is so experimental, and the factor of safety completely unknown, that it requires a system to warn the pilot of impending collapse.”

Like Lochridge, Triton’s outside counsel, Brad Patrick, considered the risk to life to be so evident that the government should get involved. He drafted a letter to McDevitt, the OSHA investigator, urging the Department of Labor to take “immediate and decisive action to stop OceanGate” from taking passengers to the Titanic “before people die. It is that simple.” He went on, “At the bottom of all of this is the inevitable tension betwixt greed and safety.”

But Patrick’s letter was never sent. Other people at Triton worried that the Department of Labor might perceive the letter as an attack on a business rival. By now, OceanGate had renamed Cyclops II “Titan,” apparently to honor the Titanic. “I cannot tell you how much I fucking hated it when he changed the goddam name to Titan,” Lahey told me. “That was uncomfortably close to our name.”

“Stockton strategically structured everything to be out of U.S. jurisdiction” for its Titanic pursuits, the former senior OceanGate employee told me. “It was deliberate.” In a legal filing, the company reported that the submersible was “being developed and assembled in Washington, but will be owned by a Bahamian entity, will be registered in the Bahamas and will operate exclusively outside the territorial waters of the United States.” Although it is illegal to transport passengers in an unclassed, experimental submersible, “under U.S. regulations, you can kill crew,” McCallum told me. “You do get in a little bit of trouble, in the eyes of the law. But, if you kill a passenger, you’re in big trouble. And so everyone was classified as a ‘mission specialist.’ There were no passengers—the word ‘passenger’ was never used.” No one bought tickets; they contributed an amount of money set by Rush to one of OceanGate’s entities, to fund their own missions.

“It is truly hard to imagine the discernment it took for Stockton to string together each of the links in the chain,” Patrick noted. “ ‘How do I avoid liability in Washington State? How do I avoid liability with an offshore corporate structure? How do I keep the U.S. Coast Guard from breathing down my neck?’ ”

But OceanGate had a retired Coast Guard rear admiral, John Lockwood, on its board of directors. “His experiences at the highest levels of the Coast Guard and in international maritime affairs will allow OceanGate to refine our client offerings,” Rush announced with his appointment, in 2013. Lockwood said that he hoped “to help bring operational and regulatory expertise” to OceanGate’s affairs. (Lockwood did not respond to a request for comment.) Still, Rush failed to win over the submersible industry. When he asked Don Walsh, a renowned oceanographer who reached the deepest point in the ocean, in 1960, to consult on the Titanic venture, Walsh replied, “I am concerned that my affiliation with your program at this late date would appear to be nothing more than an endorsement of what you are already doing.”

That spring, more than three dozen industry experts sent a letter to OceanGate, expressing their “unanimous concern” about its upcoming Titanic expedition—for which it had already sold places. Among the signers were Lahey, McCallum, Walsh, and a Coast Guard senior inspector. “OceanGate’s anticipated dive schedule in the spring of 2018 meant that they were going to take people down, and we had a great deal of concern about them surviving that trip,” Patrick told me. But sea trials were a disaster, owing to problems with the launch-and-recovery system, and OceanGate scuttled its Titanic operations for that year. Lochridge broke the news to Lahey. “Lives have been saved for a short while anyway,” he wrote.

OceanGate kept selling tickets, but did not dive to the Titanic for the next three years. It appears that the company spent this period testing materials, and that it built several iterations of the carbon-fibre hull. But it is difficult to know what tests were done, exactly, and how many hulls were made, and by whom, because Rush’s public statements are deeply unreliable. He claimed at various points to have design and testing partnerships with Boeing and NASA, and that at least one iteration of the hull would be built at the Marshall Space Flight Center, in Huntsville, Alabama. But none of those things were true. Meanwhile, soon after Lochridge’s departure, a college newspaper quoted a recent graduate as saying that he and his classmates had started working on the Titan’s electrical systems as interns, while they were still in school. “The whole electrical system,” he said. “That was our design, we implemented it, and it works.”

By the time that OceanGate finally began diving to the Titanic, in 2021, it had refined its pitch to its “mission specialists.” The days of insinuating that Titan was safe had ended. Now Rush portrayed the submersible as existing at the very fringe of what was physically possible. Clients signed waivers and were informed that the submersible was experimental and unclassed. But the framing was that this was how pioneering exploration is done.

“We were all told—intimately informed—that this was a dangerous mission that could result in death,” an OceanGate “mission specialist” told Fox News last week. “We were versed in how the sub operated. We were versed in various protocols. But there’s a limit . . . it’s not a safe operation, inherently. And that’s part of research and development and exploration.” He went on, “If the Wright brothers had crashed on their first flight, they would have still left the bonds of Earth.” Another “mission specialist” wrote in a blog post that, a month before the implosion, Rush had confessed that he’d “gotten the carbon fiber used to make the Titan at a big discount from Boeing because it was past its shelf-life for use in airplanes.”

“Carbon fibre makes noise,” Rush told David Pogue, a CBS News correspondent, last summer, during one of the Titanic expeditions. “It crackles. The first time you pressurize it, if you think about it—of those million fibres, a couple of ’em are sorta weak. They shouldn’t have made the team.” He spoke of signs of hull breakage as if it were perfectly routine. “The first time we took it to full pressure, it made a bunch of noise. The second time, it made very little noise.”

Fibres do not regenerate between dives. Nevertheless, Rush seemed unconcerned. “It’s a huge amount of pressure from the point where we’d say, ‘Oh, the hull’s not happy,’ to when it implodes,” he noted. “You just have to stop your descent.”

It’s not clear that Rush could always stop his descent. Once, as he piloted passengers to the wreck, a malfunction prevented Rush from dropping weights. Passengers calmly discussed sleeping on the bottom of the ocean, thirty-eight hundred metres down; after twenty-four hours, a drop-weight mechanism would dissolve in the seawater, allowing the submersible to surface. Eventually, Rush managed to release the weights manually, using a hydraulic pump. “This is why you want your pilot to be an engineer,” a passenger said, smiling, as another “mission specialist” filmed her.

Last year, a BBC documentary crew joined the expedition. Rush stayed on the surface vessel while Scott Griffith, OceanGate’s director of logistics and quality assurance, piloted a scientist and three other passengers down. (Griffith did not respond to a request for comment.) During the launch, a diver in the water noticed and reported to the surface vessel that something with a thruster seemed off. Nevertheless, the mission continued.

More than two hours passed; after Titan touched down in the silt, Griffith fired the thrusters and realized something was wrong.“I don’t know what’s going on,” he said. As he fiddled with the PlayStation controller, a passenger looked out the viewport.

“Am I spinning?” Griffith asked.

“Yes.”

“I am?”

“Looks like it,” another passenger said.

“Oh, my God,” Griffith muttered. One of the thrusters had been installed in the wrong direction. “The only thing I can do is a three-sixty,” he said.

They were in the debris field, three hundred metres from the intact part of the wreck. One of the clients said that she had delayed buying a car, getting married, and having kids, all “because I wanted to go to Titanic,” but they couldn’t make their way over to its bow. Griffith relayed the situation to the ship. Rush’s solution was to “remap the PS3 controller.”

Rush couldn’t remember where the buttons were, and it seems as though there was no spare controller on the ship. Someone loaded an image of a PlayStation 3 controller from the Internet, and Rush worked out a new button routine. “Yeah—left and right might be forward and back. Huh. I don’t know,” he said. “It might work.”

“Right is forward,” Griffith read off his screen, two and a half miles below. “Uh—I’m going to have to write this down.”

“Right is forward,” Rush said. “Great! Live with it.”

Shipwrecks are notoriously difficult and dangerous to dive. Rusted cables drape the Titanic, moving with the currents; a broken crow’s nest dangles over the deck. Griffith piloted the submersible over to the wreck, and passengers within feet of it, while teaching himself in real time to operate a Bluetooth controller whose buttons suddenly had different functions than those for which he had trained.

“If you’re not breaking things, you’re not innovating,” Rush said, at the GeekWire Summit last fall. “If you’re operating within a known environment, as most submersible manufacturers do—they don’t break things. To me, the more stuff you’ve broken, the more innovative you’ve been.”

The Titan’s viewport was made of acrylic and seven inches thick. “That’s another thing where I broke the rules,” Rush said to Pogue, the CBS News journalist. He went on to refer to a “very well-known” acrylic expert, Jerry D. Stachiw, who wrote an eleven-hundred-page manual called “Handbook of Acrylics for Submersibles, Hyperbaric Chambers, and Aquaria.” “It has safety factors that—they were so high, he didn’t call ’em safety factors. He called ’em conversion factors,” Rush said. “According to the rules,” he added, his viewport was “not allowed.”

It seemed as if Rush believed that acrylic’s transparent quality would give him ample warning before failure. “You can see every surface,” he said. “And if you’ve overstressed it, or you’ve even come close, it starts to get this crazing effect.”

“And if that happened underwater . . .”

“You just stop and go to the surface.”

“You would have time to get back up?” Pogue asked.

“Oh, yeah, yeah, yeah. It’s way more warning than you need.”

John Ramsay, who has designed several acrylic-hulled submersibles, was less sure. “You’ll probably never be able to find out the source of failure” of the Titan, he told me, in a recent phone call from his cottage in southwest England. But it seems as though Rush did not understand how acrylic limits are calculated. “Where Stockton is talking about those things called conversion factors . . .”

Ramsay grabbed a copy of Stachiw’s acrylic handbook from his spare bedroom. When Stachiw’s team was doing its tests, “they would pressurize it really fast, the acrylic would implode, and then they would assign a conversion factor, to tabulate a safe diving depth,” he explained. “So let’s say the sample imploded at twelve hundred metres. You apply a conversion factor of six, and you get a rating of two hundred metres.” He paused, and spoke slowly, to make sure I understood the gravity of what followed. “It’s specifically not called a safety factor, because the acrylic is not safe to twelve hundred metres,” he said. “I’ve got a massive report on all of this, because we’ve just had to reverse engineer all of Jerry Statchiw’s work to determine when our own acrylic will fail.” The risk zone begins at about twice the depth rating.

According to David Lochridge’s court filings, from 2018, Cyclops II’s viewport had a depth rating of only thirteen hundred metres, approximately one-third of Titanic’s depth. It is possible that this had changed by the time passengers finally dived. But, Lochridge’s lawyer wrote, OceanGate “refused to pay for the manufacturer to build a viewport that would meet the required depth.”

In May, Rush invited Victor Vescovo to join his Titanic expedition. “I turned him down,” Vescovo told me. “I didn’t even want the appearance that I was sanctioning his operation.” But his friend—the British billionaire Hamish Harding, whom Vescovo had previously taken in the Limiting Factor to the bottom of the Mariana Trench—signed up to be a “mission specialist.”

On the morning of June 18th, Rush climbed inside the Titan, along with Harding, the British Pakistani businessman Shahzada Dawood, and his nineteen-year-old son, Suleman, who had reportedly told a relative that he was terrified of diving in a submersible but would do so anyway, because it was Father’s Day. He carried with him a Rubik’s Cube so that he could solve it in front of the Titanic wreck. The fifth diver was P. H. Nargeolet, the Titanic expert—Vescovo’s former safety adviser, Lahey and McCallum’s old shipmate and friend. He had been working with OceanGate for at least a year as a wreck navigator, historian, and guide.

The force of the implosion would have been so violent that everyone on board would have died before the water touched their bodies.

For the Five Deeps crew, Nargeolet’s legacy is complicated by the circumstances of his final dives. “I had a conversation with P. H. just as recently as a few months ago,” Lahey told me. “I kept giving him shit for going out there. I said, ‘P. H., by you being out there, you legitimize what this guy’s doing. It’s a tacit endorsement. And, worse than that, I think he’s using your involvement with the project, and your presence on the site, as a way to fucking lure people into it.’ ”

Nargeolet replied that he was getting old. He was a grieving widower, and, as he told people several times in recent years, “if you have to go, that would be a good way. Instant.”

“I said, ‘O.K., so you’re ready to fucking die? Is that what it is, P. H.?’ ” Lahey recalled. “And he said, ‘No, no, but I figure that, maybe if I’m out there, I can help them avoid a tragedy.’ But instead he found himself right in the fucking center of a tragedy. And he didn’t deserve to go that way.”

“I loved P. H. Nargeolet,” Lahey continued. He started choking up. “He was a brilliant human being and somebody that I had the privilege of knowing for almost twenty-five years, and I think it’s a tremendously sad way for him to have ended his life.”

Lahey dived the Titanic in the Limiting Factor during the Five Deeps expedition, back in 2019. I remember him climbing out of the submersible and being upset at the fact that we were even there. “It’s a mess down there,” he recalled, this week. “It’s a tragic fucking place. And in some ways, you know, people paying all that money to go and fly around in a fucking graveyard . . .” He trailed off. But the loss of so much life, in 1912, set in motion new regulations and improvements for safety at sea. “And so I guess, on a positive note, you can look at that as having been a difficult and tragic lesson that probably has since saved hundreds of thousands of lives,” he said.

OceanGate declined to comment. But, in 2021, Stockton Rush told an interviewer that he would “like to be remembered as an innovator. I think it was General MacArthur who said, ‘You’re remembered for the rules you break.’ And I’ve broken some rules to make this.” He was sitting in the Titan’s hull, docked in the Port of St. John’s, the nearest port to the site where he eventually died. “The carbon fibre and titanium? There’s a rule you don’t do that. Well, I did.” ♦