Skip to content
crush depth

Titan sub implosion caused by absolutely bonkers “toxic workplace environment”

CEO Stockton Rush does not come off well.

Nate Anderson – | 295
Photo of Titan debris on ocean floor.
All that remains of the Titan now—a tail cone resting on the ocean floor.
All that remains of the Titan now—a tail cone resting on the ocean floor.
Story text

In a 300-plus page final report released today, the US Coast Guard analyzed the 2023 Titan sub implosion from every conceivable angle and came to a clear conclusion: OceanGate CEO Stockton Rush was a dangerous and deeply unpleasant boss.

His company used “intimidation tactics” to sidestep regulatory scrutiny, it was a “toxic” workplace, and its safety culture was “critically flawed.” The Titan itself was “undocumented, unregistered, non-certificated, [and] unclassed.” As for Rush, he managed to “completely ignore vital inspections, data analyses, and preventative maintenance procedures.” The result was a “catastrophic event” that occurred when 4,930 pounds per square inch of water pressure cracked the sub open and crushed its five occupants during a dive to the Titanic wreckage site.

Had Rush somehow survived, the report says, he would have been referred for prosecution.

Stockton Rush shows David Pogue the game controller that pilots the OceanGate Titan sub during a CBS Sunday Morning segment broadcast in November 2022.
OceanGate CEO Stockton Rush shows David Pogue the 2010-era game controller used to pilot the Titan sub during a CBS Sunday Morning segment broadcast in November 2022.
Credit: CBS Sunday Morning
OceanGate CEO Stockton Rush shows David Pogue the 2010-era game controller used to pilot the Titan sub during a CBS Sunday Morning segment broadcast in November 2022. Credit: CBS Sunday Morning

Throwing the controller

One small story about a video game controller shows what Rush was like to work for. You may remember Rush from an infamous 2022 CBS Sunday Morning segment, where Rush showed journalist David Pogue around the Titan sub. “We run the whole thing with this game controller,” Rush said, holding up a Logitech F710 controller with 3D-printed thumbstick extensions. Pogue chuckled, saying, “Come on!” as he covered his face with his hand.

The game controller had been used in OceanGate subs for years by that point; a 2014 video showed one being used to control the company’s earlier Cyclops I submersible. In 2016, OceanGate took the Cyclops I to dive the wreck of the Andrea Doria outside of Nantucket, Massachusetts. (Seinfeld fans will remember that an entire episode is taken up with George’s quest to get an apartment that was about to go to an Andrea Doria survivor.)

The OceanGate team spent two days at the site, running 2D and 3D scans of the sunken ship, until Rush got the Cyclops I “stuck under the bow of the Andrea Doria wreckage”—and he couldn’t get the sub free. According to the report, Rush then “experienced a ‘meltdown’ and refused to let [the assistant pilot] assist in resolving the situation. When a mission specialist suggested that Mr. Rush hand over the controller to the assistant pilot, the assistant pilot reported that the controller was thrown at him. Upon obtaining the controller, the assistant pilot was able to free the Cyclops I from the wreckage.”

Rush’s subs got tangled with the wrecks they were visiting on more than one occasion. In 2022, one year before the fatal accident, the Titan completed a successful dive to the Titanic, but it was not without incident. The Titan was piloted “by the Titanic content expert who was not an employee of OceanGate nor a qualified pilot.” When the sub reached the ocean floor, it moved closer to the wreckage and got one of its skids physically entangled in the Titanic. It was able to get free, but the report notes: “OceanGate’s lack of a risk mitigation plan for an entanglement at depth, which was highlighted by the absence of a standby ROV or secondary submersible to assist in freeing the Titan from entanglement had the situation been more serious.”

Rush also had a habit of riding roughshod over criticism or concerns. For instance, during a 2021 dive to the Titanic, “several critical equipment failures occurred, including a malfunction with the drop weight motors, which required the jettisoning of the drop weight tray to begin ascent.” But Rush didn’t want to drop the entire tray because it might disrupt future missions, since there “were no spare drop weight trays.” Instead, Rush’s plan was for the sub to “descend back to the ocean floor and remain there for up to 24 hours until the Titan’s sacrificial anodes deteriorated and released the emergency weights.”

Although this call was supposed to be made by the mission director, Rush made the decision himself. His “refusal to adhere to the Mission Director’s orders placed the Titan crew in a hazardous situation at an extreme ocean depth of approximately 3,800 m,” the report says. (In the end, the sub was able to drop some of the weight without losing the entire tray mechanism, and it ascended safely, but the incident showed “a dangerous disregard for the Mission Director’s authority and a willingness to operate Titan at depth with multiple equipment malfunctions.”)

Rush simply had little patience for safety; he fired those who tried to delay his project based on safety concerns, and if an idea was fast and easy, he was often willing to try it. Rush at one point told crew to use only four bolts to attach the Titan’s 3,500-pound dome to the sub when it was above water, despite the fact that it was made to use 18. He did this because “it took less time.” His director of engineering told him that this was a concern but was ignored. In 2021, when the Titan was being hoisted onto the deck of a ship, these bolts all sheared off, and the huge dome crashed down onto the sub’s ocean launch platform.

Death in darkness

There’s much more of this kind of thing in the report—including the time that the Titan’s thruster controls were “inadvertently reversed” and the whole mission had to be steered “backwards”—but even the long list of safety problems in the report “represent only a fraction of the incidents that occurred during Titan’s expeditions.”

Rush was under financial pressure, which appears to explain some of these decisions and may have ultimately led to his death. OceanGate decided to save money, for example, by storing the Titan sub outdoors over the Canadian winter, “exposing the hull to extreme temperature fluctuations [that] compromised the Titan’s hull integrity.”

In the end, the hull failed. The resulting implosion was “catastrophic,” and death was “immediate.” It took two seconds for the sound of the explosion to ascend through the water column. At that point, the “Titan Communications and Tracking Team on the Polar Prince [ship] heard a ‘bang’ emanating from the ocean’s surface, which the investigation later correlated to the Titan’s implosion. After that, all communications and tracking with the Titan were lost.”

Photo of Nate Anderson
Nate Anderson Deputy Editor
Nate is the deputy editor at Ars Technica. His most recent book is In Emergency, Break Glass: What Nietzsche Can Teach Us About Joyful Living in a Tech-Saturated World, which is much funnier than it sounds.
295 Comments
Staff Picks
Defenestrar
Defenestrar
I watched most of the Coast Guard hearings as several elements of this case will make into future lessons I give on safety, engineering ethics, and especially the difference between initiating event and root cause - making it really clear to management that I sometimes get to train. Root cause is almost always management.
p
p-chapman
There are so many issues with OceanGate & Stockton Rush, that it is difficult to come up with which explanation for the accident to lead with. So I see general news sources going off in one direction or another, without really handling the whole combination.

Some of Stockton's terrible behaviour we had heard about before. The Coast Guard report adds to that, showing the many times there was some possible warning sign of trouble, or some engineering issue not really investigated, that Stockton brushed aside, because he had limited money. Especially in the later years when passenger money had already come in and he needed to get trips going to the Titanic to satisfy them. Nothing could stand in the way of getting moving and making money, to keep the company from falling apart. "Push-on-itis" in aviation slang. Incidents were covered up or explained away, so that the passengers signing waivers certainly weren't giving "informed consent".

What was interesting in the report was new information about the direct technical reasons for the implosion:
The MBI determined that the probable failure point of the hull was
either the adhesive joint between the TITAN’s forward dome and the titanium segment
or the carbon fiber hull near the forward end of the TITAN.

There's strong evidence that the hull, that was cured in multiple 1" layers, had a delamination between the inner and next layer -- explaining a loud bang from about 8 dives back. So it had been damaged for some time. (Curing a 5" thick composite all at once is tough; but so is getting it right with multiple layer cures. I don't know the solution to that dilemma.)

And there may have been separation at the glue joint between the forward dome and the cylindrical hull. Yes the cap would have been pressed onto the hull by water pressure, so no problem, right? No, for if each would have deflected / squeezed a different amount by the stresses (different strain amounts in engineering terms), that sets up stresses on the joint. Which I guess they didn't properly engineer for.

Add that to other issues with the hull, for even after scrapping one composite hull with build quality issues, when they made the final one, it still had some waviness in spots which they would sand smooth, sanding through some of the carbon fibre layers. That's in addition to the voids and imperfections we knew about long ago, that were seen in the excess pieces left over on shore, where the ends of the cylinder were cut down to exact length.

Just so many issues, making it hard to keep this post of reasonable length; it gets hard to figure out some priority to write about when there are so many causes & factors involved.

The Coast Guard note that more careful examination of their built-in acoustic data and also strain gauge data would have detected that hull issue on dive 80 (while the deadly dive was #88) , which actually happened pretty much as they surfaced, and pressure was coming off the whole vessel. There was a bang that Stockton attributed to some shift of the main pressure vessel in its outside clamped on frame, or whatever. had they done graphs with the correct variables, they would have found a permanent change in the strains on the strain gauges, versus depth of water. So "something had let go".

So Stockton's gauges built into the hull, for all their imperfections and failures, could have helped avoid the accident! But yes they ended up being useless for real time monitoring of impending failure, the way he used them and boasted about the acoustic monitoring.

It didn't help that in the winter of '22 - '23 the Titan basically sat outside, unprotected, through the Newfie winter, as Stockton didn't have the money to take it somewhere better, didn't even pay for a plastic boat wrap for it, while he had discussions with a local university to take it in as a learning display, and thus avoid Canadian import taxes.

And it also didn't help that OceanGate cheaped out for the '23 season with a lesser support vessel, that couldn't hoist the Titan onboard. So the Titan was dragged literally thousands of miles through the seas behind the ship, on its little platform, as they made four trips into the Atlantic in 2023 before finally better weather allowed them to make the first dive of the season onto the Titanic. Towing the Titan around instead of having it on deck also made any inspection & maintenance much harder and less likely.

Thus after some prior delamination, and with the imperfections in the hull, it is just possible that things got worse because of being exposed to so much water, as well as freezing / thawing cycles in winter.

In 2023, they only had 4 little test dives to no more than 10m, before that final 5th expedition where they got good weather. The first deep dive of 2023, with passengers, dive # 88, was the fatal one.

Oh and Stockton deliberately obscured how well tested the Titan was, as he started counting the dive numbers with the OLD Titan, not the NEW Titan... which was called the same name and not something like Titan II. The old Titan was taken out of service after a crack was found in the hull, although the titanium end domes were reused. Taking away the 49 dives of the previous Titan, that means the fatal dive was dive # 39.... which includes all those little submergences to like 3 or 10 metres a little ways away from some port!

[Edit - I fixed the numbers in the next 2 paragraphs as one chart I took data from was a partial one. Found the 2nd part showing more dives. So the numbers are a bit higher on their deep dives.]

Hmm, so let me add up the "really deep dives": In the first year of the new Titan, 2021, they got 8 dives that were deep - at least 1700 m instead of 170 m or less. Then in 2022, they did 9 dives of 1700 m or more. Both seasons they got a bunch of Titanic dives, although not all ended up actually viewing the Titanic.

So before the fatal accident with passengers, the Titan had completed a total of 17 deep dives before the fatal one. [not 9 as I stated originally]

There are plenty more stupid events & decisions to read about in the Coast Guard report.

Excuse any errors as I'm still trying to digest the 330+ pages of the report.


(As an anecdote about nearly spotting a Titan dive: When looking through the Coast Guard list of Titan dives, I found a single one, in July 2022 in the bay of the little town of Bay Bulls, Newfoundland. Just a long shallow dive to test out fixes between two Titanic trips. I recognized that town name... I had been there once in my life... and it turns out on the very same day. So I looked through my photos from my hike along the sea coast cliffs, and out of the fog a couple hundred meters out, there appeared the support vessel I now know Stockton had used that year. He must have been there under water there while I was hiking past along the coastline....)

[minor edits for clarification]
Wheels Of Confusion
Wheels Of Confusion
James Cameron's Deep Sea Challenger (which he took to the deepest point in the Mariana Trench) wasn't a sphere - but he also wasn't making it from jury-rigged parts in order to keep the costs low, and employed very experienced engineers to design it.
The actual pressure vessel basically was a sphere. The superstructure was built around that. (Sorry for the black text on transparency for fellow Dark Mode users, but this is the best I got)

Prev story
Next story