The tragic implosion of the OceanGate Titan submersible two years ago, claiming the lives of all five onboard, sent shockwaves around the globe, highlighting the perilous nature of deep-sea exploration.
Now, an exclusive look into tens of thousands of internal OceanGate emails, documents, and photographs, along with interviews with former employees and third-party suppliers, reveals a disturbing inside story. These materials, provided exclusively to WIRED by anonymous sources, paint a picture of a company culture where safety concerns were repeatedly raised by engineers and experts but were often dismissed or ignored by OceanGate CEO Stockton Rush.
Stockton Rush, a former flight engineer and tech investor, cofounded OceanGate in 2009 with a vision for commercial and research trips to the ocean floor. He spoke with grand ambition, once telling a reporter, “I wanted to be the first person on Mars until I realized there was nothing there.” He saw the ocean as the true frontier, full of undiscovered life, stating, “My goal is to move the needle.”
Initially, OceanGate used older, steel-hulled submersibles. But by 2013, Rush was pursuing a “revolutionary new manned submersible” he called Cyclops 2. Its key innovation was a lightweight carbon fiber hull designed to carry more passengers than traditional spherical cabins and dive much deeper than existing subs.
Rush’s ambitious vision for the Titan was to transport paying adventurers to the Titanic wreck, resting 3,800 meters below the ocean’s surface, where the pressure is an astounding 6,500 pounds per square inch compared to the 14.7 psi experienced at sea level.
Testing the innovative design was crucial, and scale models were built for pressure testing. On the afternoon of July 7, 2016, a three-foot-long scale model of Cyclops 2 was being tested in a high-pressure facility at the University of Washington in Seattle. Water was pumped in, steadily increasing the pressure to mimic a dive.
At about the 73-minute mark, as the pressure gauge reached 6,500 psi – the equivalent of the Titanic’s depth – a sudden, violent roar erupted. The tank shuddered.
“I felt it in my body,” one OceanGate employee wrote in an email that night, describing the event. “The building rocked, and my ears rang for a long time.” He added, “Scared the shit out of everyone.”
The model had imploded at a depth equivalent to the Titanic, thousands of meters short of the safety margin OceanGate had designed for. In the high-stakes world of crewed submersibles, such a failure would typically send engineering teams back to the drawing board.
However, the exclusive documents and interviews reveal that Rush’s company did not take this step. Instead, within months of the implosion, OceanGate began building a full-scale Cyclops 2 based on the design of the imploded model. This full-sized vessel, the Titan, would go on to successfully reach the Titanic in 2021 and return for expeditions in 2022 before its fatal dive.
Carbon fiber is indeed a strong material, often stronger and lighter than titanium, making it attractive for engineering applications like aerospace. However, engineers Mark Negley and William Koch of Boeing Research & Technology, who produced a preliminary design report for OceanGate in October 2013, highlighted specific challenges for deep-sea applications.
They pointed out that carbon fiber can weaken progressively, sometimes unexpectedly, and the manufacturing process itself can introduce defects that compromise its strength. The risk of such defects increases with the number of layers used; Titan’s hull would ultimately have 660 layers.
To mitigate these risks, the Boeing engineers recommended rigorous quality assurance during manufacturing and subsequent ultrasound testing of the hull to detect defects or delaminations. This advice, detailed in their 70-page report, appears to have been a crucial early warning.
Early scale model testing of the carbon fiber hull components also showed worrying signs. In June 2015, a one-third scale model with carbon fiber end domes failed at pressures equivalent to around 3,000 meters. Although the cylinder portion performed better with solid aluminum ends, subsequent tests with new carbon fiber domes in March 2016 again resulted in implosion at 3,000 meters.
The alarming fourth scale model test, which ultimately failed at a depth of 4,500 meters due to aluminum caps, revealed a concerning safety factor of merely 1.18 for dives to the Titanic’s depth. For context, James Cameron’s Deepsea Challenger had a safety factor of 1.36, while the submersible Alvin, which originally explored the Titanic, boasted a much safer rating of 1.8 or higher.
Despite clear evidence that the model design was inadequate and dangerous below the target depth, Rush assured shareholders that they would conduct extensive durability tests on a new cylinder, a plan that former employees revealed was never executed due to Rush’s overconfidence in OceanGate’s computer simulations.
The final Titan design incorporated titanium end domes, a change from the tested carbon fiber domes. However, a former employee familiar with Rush’s decision said the CEO balked at the cost of commissioning models to test the interaction between the new materials, relying instead on analysis.
One former employee remarked on this troubling approach, saying, “The modeling says it’s OK. The analysis says it’s OK,” further commenting on the irony that the same model is used for aircraft. Yet, experts caution that carbon fiber reacts differently under high external pressure, like that found underwater, compared to internal pressure in an aircraft cabin, making it especially critical for underwater applications.
Submersible experts outside OceanGate stressed the need for extensive testing on new designs. Adam Wright, who worked on a carbon-fiber sub for explorer Steve Fossett, stated, “We did at least 10 scale-model pressure hulls that we tested to destruction.” OceanGate, by contrast, tested its model hull to destruction only once and didn’t test the final configuration with titanium components.
Chase Hogoboom, president of Composite Energy Technologies, confirmed carbon fiber is a viable material but requires significant investment in engineering and manufacturing controls over many years. He noted, “It takes millions of dollars and many years, but it’s not rocket science. It’s just connecting the dots.” Despite this, OceanGate engineers later found the full-size Titan hull was too thick for portable ultrasound scanners, and a coating applied by the manufacturer further blocked signals. A former employee said Rush decided moving the sub to a lab for scanning was too expensive, resulting in no scans being made, contrary to the advice of Boeing and OceanGate engineers.
Further warnings arose concerning the Titan’s viewport. It was a new design by OceanGate’s director of engineering and manufactured by Hydrospace Group. Will Kohnen, Hydrospace CEO, expected thorough testing according to rigorous standards like those set by the American Society of Mechanical Engineers, which involve testing multiple windows to destruction, cycling under pressure, and long-duration stress tests.
“The more innovative you get, the more testing you’ve got to do,” Kohnen said. He grew concerned in late 2017, telling WIRED, “Over a period of years, it was pretty obvious that OceanGate wasn’t going to do the testing.” Former OceanGate employees corroborated that the viewport was not tested to ASME standards.
In November 2017, as a final attempt, Kohnen emailed Rush offering a significant discount on a second viewport using a design certified to 4,000 meters, which could be swapped out quickly. Rush told him he wasn’t interested.
When Hydrospace delivered the viewport to OceanGate in December 2017, it was rated for a mere 650 meters, a fraction of the depth needed for the Titanic. An independent analysis indicated that the design could fail after only a few attempts at 4,000 meters, yet OceanGate proceeded to install the viewport and marketed its first Titanic expedition for the upcoming May.
Warnings were also escalating within the company; in January 2018, David Lochridge, the director of marine operations, expressed serious safety concerns regarding the Titan in a quality-control report, detailing 27 issues related to the carbon-fiber hull and component seals, only to be fired the very next day by Rush.
Lochridge later filed a whistleblower report, but Rush sued him, resulting in a settlement where Lochridge dropped his complaint and signed an NDA. The pattern of dismissing internal dissent was evident in OceanGate’s culture, where employees who questioned the rapid pace or decisions were sometimes dismissed as overly cautious or even fired.
:max_bytes(150000):strip_icc():focal(773x260:775x262)/titan-submersible-061724-1-82f74247e2ef4a2e80ddba10c72b6581.jpg)
The concerns extended across the industry. Will Kohnen remembers thinking, “We have a rogue element within the submersible industry,” fearing that a failure could negatively impact deep-sea exploration broadly. In March 2018, he drafted a letter, signed by over 30 crewed submersible experts, urging Rush to seek testing and certification from an accredited outside group like DNV or the American Bureau of Shipping.
While it has been widely reported that Rush dismissed such certification, the documents reveal OceanGate *did* pursue certification with DNV in 2017 until Rush learned the estimated cost of around $50,000. In an email, Rush wrote he had “grown tired of industry players who try to use a safety argument to stop innovation,” stating, “Since [starting] OceanGate we have heard the baseless cries of ‘you are going to kill someone’ way too often.”
Perhaps the most alarming communication came from Boeing’s Mark Negley on March 30, 2018, who, after reviewing the hull’s design, bluntly warned Rush of a high risk of failure before reaching 4,000 meters—emphasizing a complete lack of safety margin in their design.
Negley urged Rush to “Be cautious and careful” and included a graph charting strain on the submersible against depth, which starkly displayed a skull and crossbones in the region below 4,000 meters, a vivid representation of the potential for catastrophic failure at Titanic depths.
Rush, however, seemed largely unfazed by these repeated warnings. His confidence was reportedly bolstered by a real-time health monitoring system designed by engineer Mike Furlotti, intended to detect the sounds of carbon fibers breaking under compression.
OceanGate’s theory was that the hull would be noisy initially but quiet down over repeated dives to the same depth, with increased noise indicating a need to surface immediately. However, industry experts were highly skeptical. Adam Wright noted the system’s uncertainty: “you just don’t know when the end point is,” adding, “You don’t know how many pops is too many, and it could be different for every vessel.”
In September 2017, an engineer raised concerns about the accuracy of the system meant to track fiber breakage. Although an external consultant later approved it in 2018, he grew apprehensive after Rush claimed that it could detect “micro-buckling” well before a failure, suggesting that the CEO may have exaggerated its capabilities.
The extensive documentation and interviews reveal a troubling trend of favoring ambition and cost over safety protocols and expert warnings. From constructing a full-scale vessel based on a model that imploded before reaching its intended depth to dismissing severe concerns about manufacturing flaws and hull integrity, the tragic trajectory leading to the Titan’s disaster is marked by missed chances to avert catastrophe.
Related posts:
The Titan Submersible Disaster Shocked the World. The Exclusive Inside Story Is More Disturbing Than Anyone Imagined
OceanGate adviser rips US rescue response — but firm faces claim it took 8 hours to raise alarm
