Michigan Engineering News

A map of vessel custom-designed for Project Azorian undercover as a manganese mining ship


In the late 1960s, the U.S. caught wind that the Soviets had lost a nuclear submarine. When a covert operation to recover the sub and its secrets was launched, several Michigan Engineering alumni were central to the operation. Here is their story.

A map of vessel custom-designed for Project Azorian undercover as a manganese mining ship
This vessel was custom-designed for Project Azorian. Undercover as a manganese-mining ship, it would sail above the submarine and lift it up through a moon pool in the bottom.

This feature was authored by guest writer Randy Milgrom.

With an offer “only an idiot would refuse,” Chuck Cannon (BSE NAME ’67) was working with Global Marine, Inc.’s chief engineer John Graham on the hull design for “Glomar Explorer” – an enormous deep ocean mining ship Cannon was told would be used for an audacious Howard Hughes-backed effort to mine manganese from the ocean floor.

Around the same time, Charles Canby (BSE NAME ’69), who began his close and continuing friendship with Cannon while both were Michigan Engineering students, was hired by GMI’s Klemme Jones (BSE NAME ’49), another Michigan man, who instructed Canby to add an outsized “moon pool” at the bottom of the vessel.

Canby thought the specs seemed odd, but he never questioned them. Later, when he learned the truth, the young engineer was shocked. “It was like hearing the facts of life for the first time,” he said.

Soviets Lost, U.S. Found

The precise reason why Soviet nuclear submarine “K-129” sank remains unknown. The sub missed a scheduled midnight radio check on March 8, 1968, as it crossed the 180th meridian – and that was that. The Soviets became concerned about consecutive failed check-ins, and by the third week of March, when urgent communication requests also went unanswered, the Soviets declared K-129 “missing.”

U.S. intelligence services observed numerous Soviet subs, periscopes up, within their known patrol routes, which suggested a search might be underway. But the Soviets, who had not yet developed sea floor acoustic sensors, were using sonar – which made locating anything likely already resting on the ocean floor nearly hopeless.

The Soviets – unwilling to admit to these limited capabilities – quietly abandoned their search for the sub. But they also believed that no other foreign intelligence service would be able to find it, either.

But the U.S. had in fact established a vast network of underwater microphones in the northern Pacific, which enabled them to identify the “sonic signature” of an implosion occurring on March 8. This led Naval Intelligence to determine the site of the wreck within five nautical miles of 40.1 degrees N latitude and 179.9 degrees E longitude (in international waters very near the International Date Line) – at a depth of more than three miles.

In July 1968, the United States Navy ordered USS “Halibut” – the only specially equipped search submarine in its inventory – to find and photograph K-129. Using its collection of remote-controlled cameras, strobe lights and sonar built to withstand extreme depths, within three weeks Halibut had located and generated enough photos of the sub to enable the CIA to determine that its missiles might still be intact – a potential intelligence cornucopia. But it also knew, as revealed in a redacted declassified memo written at that time but not released to the public until 2010, that the retrieval effort (code name: Project Azorian) would be no ordinary mission. The memo presented this analysis:

“Azorian … combined … advanced technological development, complex systems engineering and testing, unusually severe cover and security requirements, a demanding mission scenario in an unforgiving marine environment, the potential for a serious confrontation with the Soviet Union, a difficult and technically unusual exploitation phase, and a high cost …” the memo reads.

But it also was inevitable that the U.S. would pursue it. Since the late 1950s, both countries had been engaged in dangerous and escalating gambles to gain military clues. Planes and pilots from both sides were occasionally lost without official explanation – and with land-based intercontinental ballistic missiles now more easily identified and targeted, nuclear submarines were recognized as difference-makers, and submarine-based forays – and losses – were on the rise.

As then-CIA director William Colby said in a 1993 interview: This sub was not on some leisure cruise. “It was pointing missiles at us!”

Group image of the lost crew of K-129
The crew of the K-129, shown here, was lost with the ship. Soon the Soviet Union gave up on its search for the submarine and its secrets.

Rough Waters, Under Cover

By November 1969, the CIA’s rigorous due diligence had led them directly to GMI – the rising offshore drilling industry leader with five heavy ocean-going drilling ships built in the last several years, including the largest ever. But many aspects of this ship would constitute first-ever achievements. The core of the mission – hauling anything this heavy from this depth – had never been attempted. And it had to work the first time.

Most of those working on the ship’s design and construction were unaware – and would remain unaware – of its true purpose. Only those who needed to know were told, while the others believed the mining manganese cover story – as did most of the rest of the world.

Alumnus John Hollett (BSE NAME ’67, MBA ’72) was brought in to work on the ocean-mining side of the house. He job was to help cultivate clients for the mining technology, which led him quite naturally to try the U.S. Navy. Unbeknownst to him, this was nosing too close to “black” projects by the CIA. When word got back to his boss, Hollett was severely chastised – and brought in on the secret.

“It became necessary for me to be ‘cleared’ to the actual salvage project since I could have compromised the whole thing by having come in a ‘back door,’” he said.

To vet him, the CIA interviewed people from his past – friends, neighbors, people he knew at U-M. Posing as representatives for the Howard Hughes venture, they asked detailed questions about drinking habits, drug use, marital fidelity and temperament.

“I remember receiving many phone calls from prior associates, asking why Howard Hughes was so interested in my total being,” he said. “I feigned total ignorance by saying that was just what this guy was like.”

Top view image of the submarine K-129
The reasons why the submarine K-129 went down were unknown. But it happened at a time when both countries were gambling to gain military clues.

But not everyone was cleared – including Steve Kemp (BSE NAME ’70), another Michigan-trained GMI naval architect – which most seemed to think was due to his well-known anti-Nixon stance. Kemp was talented and valuable enough to remain on the project, but he would need to be shielded from certain information – which he sometimes found peculiar.

When Cannon finally was briefed on the “cloak and dagger stuff” – the manganese mining story as cover for the CIA operation – it “opened up a whole new world of secrecy” that made him “personally uncomfortable.”

Hughes had agreed to lend his name and otherwise stayed away. But his representatives served in a variety of roles to extend and deepen the cover story, with each milestone – arrivals and departures from various ports; a celebratory signing to transfer the ship’s title from GMI to Hughes – glamorized by a dedicated public relations shop feeding very real-seeming tidbits to an all-too-willing press. Some crewmen even mined for actual manganese nodules and posed dockside, with buckets of their bounty, for photos that were splashed across newspaper pages worldwide.

Howard Hughes had agreed to lend his name and otherwise stayed away. But his representatives served in a variety of roles to extend and deepen the cover story.

Adding to the mission’s degree of difficulty: the Soviet sub went down in one of the roughest sections of open water on the globe. The retrieval operation would therefore need to take place within a narrow six-week, good-weather window – July 1 through mid-September – while the effect of roiling waters would still be carefully accounted for in a variety of ways.

Explorer featured three major elements: the vessel itself (built by Sun Shipbuilding); the heavy-lift pipe string (built by Lockheed); and the sub-sea grapple “claw” (built by Hughes and called the “Capture Vehicle” but nicknamed “Clementine”). Clementine, once lowered on the pipe string to the ocean floor, would grasp the wreckage and lift it back up through the moon pool.

That moon pool – an enormous center well opening in the hull, with means to be sealed off so that the sub, once captured, could be examined underwater, but in dry conditions and away from view – worried chief architect Graham no end: It was so large, when opened, that only very narrow sidewalls remained to hold the bow to the stern.

Cannon was in charge of maintaining the vessel weight and center of gravity estimate, with a directive from Graham to recalculate it monthly as Explorer’s specifications were fine-tuned. Most importantly, this meant monitoring the vessel’s stability characteristics during the ultimate heavy lift – with the moon pool open and flooded and the entire hydraulic heavy lift system, with the full weight of its heavy lift pipe string, claw and recovered submarine acting on the gimbaled platform, whose pivoted rings were designed to isolate that enormous load from its inevitable resultant pitch, keeping the platform horizontal.

Construction already had begun when Cannon realized that updated computations showed the vessel with practically zero initial stability during the heavy lift off the bottom. When Cannon told Graham, he “exploded.”

Not only had Sun already cut the steel and started wing wall module assembly, but the hull lines and structure, moon pool gates, accommodations, deckhouses, machinery spaces, mechanical systems and A-frame all were based on the established width of the beam. Also, the “company” (i.e., the CIA) would need to be told – so Graham grabbed Cannon and flew to Langley, where Graham did all the talking: This was a serious stability issue. Pipe breakage was at stake. The only solution was to widen the ship.

The quick decision that followed – to increase by 10 feet both the overall beam and the width of the moon pool – necessitated a nearly complete redesign. But its greatest impact was that Explorer – which was to be no wider than 105 feet to enable safe passage through the Panama Canal as it journeyed from Sun’s home in Chester, Pennsylvania, to Long Beach, California, where a new crew would prepare for the Pacific Ocean mission – would now have to go all the way around the southern tip of South America to get there.

While Building, Still Plotting

The tapered steel pipe string was not mechanically complex. But its stresses and loads were making mere mortals cringe. A 1970 CIA task force called it “impossible” to manufacture a pipe string strong enough to support the mission’s estimated maximum fail-safe load of 17,126,00 pounds. Curtis Crooke, GMI’s engineering vice president, managed to convince them otherwise.

But to convince himself, Crooke consulted the Army division that made World War II battleship gun barrels. Three steel companies made the rough-machined forgings. The pipe string tested well at scale, but Crooke still worried about its tensile strength, so he prevailed upon the CIA to build an expensive, customized, full-scale system capable of subjecting it to a load of 21,460,000 pounds.

Clementine would not be symmetrical (five small crane-like devices on one side and only three on the other, to accommodate how the sub was situated – and how it would need to be grabbed off the ocean floor). It contained a deployable steel mesh that could catch any missiles that might slip out during the precarious lift. And its complex operational systems required two redundant three-inch-diameter electromechanical umbilical cables to send and receive data, with controls for electrical, hydraulic, camera and other systems. No less an expert than Kelly Johnson (BSE AeroE ’32, MSE ’33) – the Michigan Engineering and Lockheed Skunk Works icon whose previous CIA work included the U-2 and the Blackbird spy planes – was consulted for his review and blessing.

Imagine standing on top of the Empire State Building … to lower the rope and grapple to the street below, snag a compact-sized car full of gold and pull the car back up.

Capture of the sub itself would be accomplished through the precise remote manipulation of the claw by crewmen needing a clear picture of the wreck and its surroundings – from three miles above. One CIA engineer offered this analogy:

“Imagine standing on top of the Empire State Building with a 4-by-8 grappling device attached to the end of a one-inch-diameter steel rope. The task is to lower the rope and grapple to the street below, snag a compact-sized car full of gold and pull the car back up to the top of the building … without anyone taking note ….”

Sun’s shipbuilders had never constructed anything this large or intricate, with 250 engineers, 20 purchasers and more than 3,000 production workers working all day, every day, to meet the rigorous timetable. It was a “pure muscle project” whose costs – projected to exceed $50 million and perhaps approach $100 million at a time when the average ship cost was only $12 million – were considered extreme. That Howard Hughes was something else!

Yet even as this costly, breakneck pace was beginning, skepticism lingered within Navy ranks that this crazy plan would work. And the Chairman of the Joint Chiefs of Staff continued to worry about the mission’s then estimated 20-30 percent chance of success – and about the nature of a Soviet response should they find out.

3D rendering of the wreck of the K-129
Artist’s rendering of the wreck of the K-129. The section the project aimed to recover was the forward one-third of the sub, which contained what the CIA hoped was a treasure trove of intelligence.

Ship Moving, Clock Ticking

On July 24, 1973, Explorer began its journey from Chester’s Atlantic coast to Long Beach’s Pacific coast – the long way around. The ship would travel 12,745 nautical miles in 50 days, seven hours, and 30 minutes, including a short layover in Bermuda and a brief stop in Chile, where its arrival on September 11 coincided with Pinochet’s launch of a successful coup to overthrow the Allende government – though Explorer managed to come and go without incident.

When Explorer arrived in Long Beach on September 30, it pulled into its pre-arranged berth alongside Howard Hughes’ famous “Spruce Goose” – another cover story masterwork.

And now it was time to alert the crew of the real objective of the project.

After working for a short time with Graham and Cannon, Canby had left GMI for a job he thought he’d love – but hated. When he returned, he “just wanted to be on a boat,” so he had joined the crew in Bermuda as a “regular seaman” – still believing he was on a great adventure to mine manganese.

Now Canby, who originally was told a cover story “so thick and thorough, it was believable to anybody,” arrived in Long Beach, where he learned “the facts of life” – which was eye-opening enough. But Cannon, for family reasons, would not be staying on to complete the mission, and Kemp still couldn’t get clearance. So Canby suddenly was named naval architect – on a ship whose secret mission was to find and recover a Soviet nuclear submarine.

Nothing could simulate actual operational conditions except the real thing. They would just have to go out and do it.

Time was running short – and adding to the stress and tension was Graham’s poor health. As a lifelong smoker the tough but beloved leader suffered from chronic emphysema, but it was getting worse. In the fall of 1973, he received a diagnosis of terminal cancer and the news hit everyone hard. Graham remained at work but was so weak that Kemp had become a kind of special assistant, making inspections and performing tasks that Graham no longer could. Though it was especially awkward, in those circumstances, to keep the truth from Kemp – they all did.

Sea trials finally began in January 1974 – and they did not begin well.

The pipe handling and heavy-lift system worked, but not especially smoothly. Tests of the moon pool “gates” – actually, nine-foot-thick doors – were conducted in especially rocky seas. Once the gates were opened they would not automatically close, and two divers had to close them manually, using ratchets. Canby and Kemp quickly revised the use manual to require that the gates be opened only in calm waters, as they had quickly determined that once they were open and secure to the ship’s bottom, the state of the sea wouldn’t matter.

The trials made it abundantly clear that though there still were bugs to work out, and that there would be breakdowns, they would just have to be fixed on the move. Nothing could simulate actual operational conditions except the real thing. They would just have to go out and do it.

On June 7, 1974, President Nixon officially authorized the mission – with the caveat that actual recovery operations not begin until after the president returned from his long-planned visit to the Soviet Union from June 27 until July 3.

Explorer arrived at the recovery site on July 4, one day after Nixon’s return to the U.S.

While Working, Soviets Watching

During much of the time that Explorer floated above its target, awaiting calmer seas and readying for the sub’s capture, a Soviet missile range instrumentation ship, an SB-10 tugboat and a companion helicopter lurked close enough to take photos.

Explorer crewmen at times exchanged messages with the instrumentation ship and calmly conveyed the manganese program “realities.” The Soviets at one point transmitted, “We wish you all the best,” and appeared to leave the scene.

The next day, however, the tug appeared, and circled close enough that Explorer had to warn it off more than once. With the tug remaining nearby, the heavy lift crew – its efforts focused on leaking hydraulics and a corresponding loss of beam pressure – very slowly lowered Clementine toward the bottom, where its cameras finally found and started to take surprisingly clear photos of the target.

With extreme caution, operators situated the claws around the sub. And on August 1, just as they were starting to actively grab it, the tugboat gently moved off.

The claw was designed to dig beneath the sub to ensure its complete capture. But the soil was harder than anticipated. As the crew, with some nervousness, tried its best to close around the entirety of the sub – the ship began to list.

The crew surrounded their catch with the containment net and paused before proceeding. Their next step would be their most dangerous. If the pipe broke, the ship could split in two.

The upward hauling went well at first. The coveted sub had risen more than 9,000 feet. Suddenly, things changed.

The heave compensators had been driven to their limit. It was clear that the load had lost some weight.

“You kind of know how it’s going on the outside by how you feel in your seat,” DMI engineer Sherman Wetmore explained. And it felt like a “small earthquake” – like “something had gone wrong somewhere on the ship.”

The heave compensators had been driven to their limit. It was clear that the load had lost some weight.

Clementine, indeed, had dropped most of its quarry – about two-thirds. “But we still had a piece of it, so keep comin’,” Wetmore thought.

Parts of the sub had shaken loose and were floating into the moon pool, and the crew worried the Soviets might notice. After all, the tug was nearby again. But as if on cue, it floated away once more. Heavy lift operators grabbed the opportunity to quickly reel in the remainder.

With Clementine back in the ship’s hull and the moon pool gates securely closed, GMI heavy lift engineer Charlie Johnson was moved by the sight of the recovered sub, resting serenely in the moon pool.

“We did it!” he exclaimed.

Aboard Explorer, the crew was inspecting and cataloguing salvaged property, and specially trained hazardous materials personnel handled potentially radioactive remains. Once the CIA had no choice but to brief Hollett, they put him through training on the remote chance they would need more crewmen aboard the Hughes Glomar Explorer – but he was never called to serve.

Graham died on August 2 – just as the sub was being captured. Cannon, who visited Graham in the hospital the day before, broke down when Graham commanded him to check on “those damned Enterprise diesels.” (Explorer had Fairbanks-Morse engines.) To fulfill Graham’s dying wish – and to honor a man who, like Moses, never reached the Promised Land – his ashes were shuttled out to Explorer and scattered in the Pacific Ocean before the ship finally set sail for home.

When it arrived in Lahaina, Hawaii, the crew – in need of a well-deserved rest – would give way to replacements. The Honolulu Advertiser ran a page one story, lauding Howard Hughes’ manganese-mining exploits.

Giving Way, Taking Stock

Explorer crewmen confirmed that several of Clementine’s claws fractured as the recovered wreckage was pulled into the hull. Blame was placed on the claw’s steel, which, while strong, was too brittle. A more ductile steel might have finished the job.

The recovered section purportedly included two nuclear torpedoes, instruments, sonar equipment, codebooks and other material of interest to the CIA – which, while remaining coy about the extent of its bounty, has publicly hailed the mission “as one of the greatest intelligence coups of the Cold War.”

It first was thought that the Soviets had been tipped off, but its intermittent lurking later was believed merely to be part of its ordinary Pacific fleet operations. Even had there been advance warning, however, the Soviets didn’t know K-129’s location. And its military command – long believing recovery impossible – was just as prone to disbelieving intelligence to the contrary.

Project Azorian’s crew also recovered the radioactive remains of six Soviet submariners – one still in his bunk, reviewing a log – and they accorded them a memorial service, with military honors, and buried them at sea. The CIA’s film of the entire operation remains classified, but the portion of the burial at sea was declassified and provided to the Russian government, along with K-129’s recovered bell, as part of a 1992 diplomatic effort.

Image of the Glomar response that was drafted after knowledge of the vessel recovery effort went public
The infamous “Glomar response” was drafted after knowledge of the recovery effort went public. This was the first use of what eventually became boilerplate language in response to inquiries on U.S. covert operations.

Coming Home, Going Back

When Explorer returned to Long Beach in September 1974, rumors floated that it was going back out. Canby, who called it the “the best job I could have,” was excited about the prospect – though apparently some of the heavy lift operators were not as interested. “The intenseness was too much,” says DMI heavy-lift operator Johnson.

What few knew, though, was that rumors were becoming reality. The CIA had prepared a cleanup mission, dubbed Matador, to retrieve what had been left behind.

But after rejecting repeated pleas from CIA Director William Colby to suppress his reporting, journalist Jack Anderson broke the real story, saying he did so because “Navy experts have told us that the sunken sub contains no real secrets and that the project, therefore, is a waste of the taxpayers’ money.”

Anderson’s February 7, 1975, page one story in the Los Angeles Times derailed Matador. With the cover now blown, many – including Kemp – felt duped. But now all those strange details he had puzzled over suddenly made sense.

And among those most disappointed was Charlie Canby.

“We took Clementine off the ship, brought it to Lockheed to put new grabbing arms on it, and brought it back onboard,” Canby says. “We would go out and test the heavy lift system and it would work so smoothly. Everything was just humming. We were all prepared to go out the next summer.”

But that was not to be. Explorer instead was mothballed for more than two decades in the north end of San Francisco Bay. In 1996, its specialized mission components were stripped and the ship was leased for deep ocean mining until just a few years ago, when Explorer finally was retired and scrapped for parts.

In 2006 the American Society of Mechanical Engineers honored Hughes’ Glomar Explorer with the prestigious Landmark Achievement Award, declaring it among the outstanding engineering feats of the 20th century.

Explorer was based on “the sound application of engineering principles. That’s what Graham did,” Canby says with admiration. “The mass and size – that could be done again. But no one builds ships that large (anymore). This was a total one-off.”

Related Topics