This article reviews the 10,000-metric-ton Russian nuclear submarine Kursk, with two nuclear reactors and 20 missiles aboard, had gone down in August 2000 following a pair of explosions approximately two minutes apart. Explosions had left the bow so deformed it was impossible to attach lifting devices, and there were serious doubts about the structural integrity of what remained. At Roslyakovo, the 21-tonne Giant was lifted clear of the water by two auxiliary pontoons. The pontoons, 100 meters long with a 15-meter beam, were sunk below the barge, winched into position beneath it, and then pumped out in order to raise the Giant-Kursk combination. This extra lift, in addition to deballasting of the drydock, provided the clearance the strand jacks needed to lower the sub on the stocks. Once the Kursk was positioned, the Giant 4 released its load and was withdrawn from the drydock.
On Oct. 7, 2001, the 10,000-metric-ton Russian nuclear submarine Kursk, with two nuclear reactors and 20 missiles aboard, began a slow ascent from its watery tomb 108 meters, or 350 feet, below the Barents Sea. Retrieving the Kursk was an emotional, as well as political, event for Russia; 118 crew members had died in the sunken submarine, some immediately, some after waiting more than 24 hours for a rescue that never came.
The Kursk had gone down in August 2000 following a pair of explosions approximately two minutes apart. Theories include a collision with another submarine or a defective torpedo exploding and detonating other munitions that destroyed the bow and sank the vessel. The real cause may never be known.
President Vladimir Putin promised the relatives he would bring the bodies home at any cost. But the Russians had neither the equipment nor the expertise to do the job, and so contracted with a Dutch company, Mammoet, which subcontracted the majority of the task to a joint venture, Mammoet Smit, formed specifically for the project.
Mammoet, based in Schiedam, is a worldwide specialist in extremely heavy lifting and transport. Its joint-venture partner, Smit International in Rotterdam, describes itself as a specialist in unconventional maritime services, including complex salvage operations and wreck removal. The sometime competitors had joined forces previously on the Oresund Bridge between Sweden and Denmark. For this job, Mammoet did the lifting, while Smit performed the underwater work.
“We had from mid-May, when the contract was let, to the end of September, beginning of October, when the storm season would begin,” said Larissa van Seumeren, a spokesperson for Mammoet. “In that time frame, we had to design, fabricate, and test equipment, modify a barge, contract for the building of auxiliary pontoons, and then execute the job.”
Usually, a project this size takes almost a year to complete. The shrunken schedule put virtually every element in the project on a critical path and it took 2,000 employees to pull it together.
“We knew we could do it,“ said Lars Wälder, a spokesperson for Smit International. “Smit had raised the Ehime Maru [the Japanese fishing boat that collided with a surfacing U.S. Navy submarine off Hawaii], and that was lying six times deeper than the Kursk. Here our biggest worry was the weather.”
Calm seas were essential; if storms prevailed, the sub could not be lifted.
Before the lift could begin, the bow had to be severed from the sub. Explosions had left the bow so deformed it was impossible to attach lifting devices, and there were serious doubts about the structural integrity of what remained. It the 1,000-tonne nose fell off during lift, the sudden shift in weight distribution could be catastrophic.
Normally, the damaged section would be cut off from the bottom up, but that method risked that near the end of the cut, the cutting tool might lift the entire vessel slightly before the two pieces broke apart and fell to the ground. Jarring nuclear reactors and unexploded munitions didn’t seem particularly advisable, so the team opted for a top-down approach.
The first compartment had contained the equivalent of 10 tonnes of TNT, and no one knew how many unexploded torpedoes were left. For safety’s sake, the cutting had to be done remotely and monitored via cameras mounted on remotely operated vehicles.
The cutting system used suction anchors, a well-known approach for deep sea anchoring and mooring. Each anchor consisted of two large-diameter cylinders open at the bottom and welded together by a frame. A suction pump on the top of the assembly was connected, through a manifold, to the inside of each cylinder. Reducing pressure inside the cylinders created a pressure differential between them and the sea that dug the anchors into the seabed.
A suction anchor was placed on each side of the sub approximately 20 meters from the hull, and the sawing element—a cable carrying a series of abrasive-covered drums—was strung across the bow between them. Each end of the sawing element was fastened to a hydraulic cylinder on an anchor. By extending and retracting the cylinders alternately, the system created a sawing motion that sliced through the bow.
Peggy Chalmers, a freelance writer based in Sunapee, N.H., holds an M.S. degree in mechanical engineering from Drexel University.
The cutting process clouded the sea with disturbed sand and cutting debris, so that every half-hour cutting was suspended to let the sea clear so the remotely operated vehicles could check progress. Surface operators could adjust the sawing element tension by drawing down on the suction anchors.
The cut was made at a downward angle 10 degrees off vertical so that the sub would lift free of the nose and not snag. Once the bow was removed, the opening in the hull was covered with a kapron net to prevent anything inside the ship from falling out.
“There were rumors that the Russians wanted to leave the nose behind because it contained some secret about how the accident happened,” Wälder said, “but we wanted to do it because it was badly damaged.”
The Russians have said they will retrieve the 20-meter-long nose this summer, but insist on doing it without outside help. According to divers working on the sub, nothing remains but broken steel.
Lifting From the Top
The Kursk was lifted from the top using cables attached to 26 metal gripper plugs mounted into the hull. Normally, cables would be wrapped under the hull, but the Russians were concerned about horizontal loading in the torpedo area. In addition, the 18-mm plates on the outer hull could not handle the loading. With the new approach, both the inner and outer hulls were pierced and the grippers installed between ribs at the inner hull. This provided a stronger lifting surface, and avoided the need to remove any equipment from the sub.
Designed jointly by Mammoet and Huisman-Itrec of Schiedam, each gripper had a pair of jaws that opened or closed when activated by an integral hydraulic cylinder. The jaws were lowered, in the folded position, through holes cut in the Kursk and then locked in place.
Each pair of jaws was individually profiled so that it would align with the curvature of the inner hull at its mounting point. In addition, steel pads, honeycombed with holes, were fastened to the jaw surfaces. Under load, the pads yielded up to 15 mm in order to compensate for any unevenness in the contact area.
The mounting holes were cut through the inner and outer hulls by a saturation diving team using a high-pressure (600-1,500 bar) water jet mixed with abrasives. The larger, outer holes were up to 2.5 meters in diameter to allow the divers to enter the space between the hulls and clear away cabling, pipes, or other obstacles. The team worked around the clock in six-hour shifts in water whose temperature hovered near freezing.
“Initially, cutting was difficult because of the rubberlike material on the sub surface used to reflect sonar,” Wälder said. “But once through that, the rest was easy.”
Determining Hole Locations
Due to the curvature of the Kursk, each mounting hole had to be cut differently so that the closed gripper could be lowered through it vertically. The hole locations were determined by the Russian Navy and the sub’s designer, Rubin Central Design Bureau for Marine Engineering, a government agency in St. Petersburg. Decisions were based on the sub’s interior design and the need to distribute lifting forces evenly along the hull.
To help the divers direct each gripper into place, a guidance cone was installed on the outside of the pressure huh around each mounting hole. Next, four 18-mm guy-wires were attached to a ring, lowered from the surface, and attached to a guidance cone. The wires then were tensioned to 15 tonnes, immediately activating a heave-compensation system to protect the wires from excessive stress due to wind and wave action on the barge.
The gripper was lowered to the Kursk by its own jack, while the guy-wires kept the gripper oriented and prevented it from rotating during the descent. Once the gripper was installed, the guy lines were raised and the process repeated for the next unit.
“Initially, we had some problems with installation,” Wälder said. “The first four or five grippers took us four to five days, and we began to worry because we had 26 to install. But things went smoothly once we got over the learning curve."
The semi-submersible pontoon salvage barge, Giant 4, was positioned over the Kursk by the Global Positioning System and held by an eight-point mooring system from four twin-drum winches on the main deck.
The 140 X 36-meter barge carried 26 strand lifting jacks, each capable of raising 900 tonnes. The jacks lifted the sub until it nested in specially constructed saddles mounted beneath the barge. The saddles were designed to handle the bending of the Kursk, as well as bending of the Giant’s hull under the load. A large recess cut in the bottom of the barge accommodated the sub’s conning tower, and load-bearing beams were installed on deck to strengthen the barge in the recess area.
Twenty-six holes, each 1 meter in diameter, were cut into the main deck and bottom hull of the barge and lined with insert rings. The lifting wires, with the grippers attached, were lowered vertically through these rings to the sub.
The muscle power came from the Mammoet-designed strand lifting jacks, a technique that the company had previously used. The strand lifting jacks mimic an individual pulling in a rope hand-over-hand. Each jack had a 54-strand cable bundle, and each strand consisted of seven twisted, high-capacity steel wires. The 54 strands were fed through top and bottom anchor heads that opened or closed to release or grip the cable bundle.
Initially, both anchor heads were closed. Then the bottom anchor head opened, releasing the cable, and a hydraulic cylinder pushed the top anchor head upward, drawing the cable bundle with it. At that point, the lower head closed, gripping the bundle, and the upper head opened so the cylinder could return it to the starting position. Then the cycle repeated. To lower the cable, the process was reversed. As the cable bundle was pulled out of the water, motorized spooling devices located above each strand jack took up the slack.
The 26 jacks were individually controlled by two linked computers that equalized the load among them and ensured that each cylinder had the same stroke when loaded, regardless of differences in oil flow and load. If a problem arose—perhaps the load was approaching a preset weight— the lifting would stop and an alarm would be raised.
Each strand lifting jack was mounted on a heave compensator—a gas suspension system fed by four nitrogen gas cylinders mounted vertically in a steel support frame. The stiffness of the system could be adjusted by increasing or decreasing the internal pressure. A ring main connected each unit to a storage bank of 16 buffer cylinders, providing a ready 250-bar backup.
While the Kursk had a static weight of only 9,500 tonnes, the hull had burrowed an average of 2 meters into the seabed and estimates to overcome the suction force ran as high as 35,000 tonnes.
“It is sort of like getting stuck with your boots in the mud,” van Seumeren explained. “It takes a lot of force to get free.”
Tension first was applied to the aft set of strand jacks, and as soon as the aft suction force decreased, the lift was transferred to the forward set. In the end, the suction forces proved much lower than predicted. The Kursk was easily freed from the mud and the lift proceeded at a cautious rate of 10 meters an hour.
A Long, Slow Tow
Eleven hours into the lift, the weather forecast turned ominous. Even though the Kursk was still 40 meters below its docking position with Giant 4, the decision was made to head toward shore. A tug started towing the barge while the Kursk continued its ascent. Five hours later, the Kursk docked with the saddle in the Giant 4 and the strand jacks were tensioned to prevent separation of the barge and the submarine during transport. By that time, the barge had already been towed 16 km. It would take another 36 hours of plodding at 4 km an hour to reach the Roslyakovo docks near Murmansk, some 195 km, or 120 miles, away.
At Roslyakovo, the 21-tonne Giant was lifted clear of the water by two auxiliary pontoons. The pontoons, 100 meters long with a 15-meter beam, were sunk below the barge, winched into position beneath it, and then pumped out in order to raise the Giant-Kursk combination. This extra lift, in addition to deballasting of the dry dock, provided the clearance the strand jacks needed to lower the sub on the stocks. Once the Kursk was positioned, the Giant 4 released its load and was withdrawn from the drydock.
It had taken five months of hard work, 16 hours of lifting, and 41 hours of towing, but the Kursk had finally come home.