In September 2013, a salvage team began rotating a 114,500 GT cruise ship back to vertical off the coast of Giglio Island, where it had rested on its side against a submerged rock ledge for twenty months. The mechanics of the rotation are documented well enough. What is less visible in any photograph of the operation is that the structural calculation behind it could not be fully checked before it started.

The Costa Concordia had capsized in January 2012 and settled at roughly 65 degrees. Steel sponsons — buoyancy caissons, partially ballasted to help control the rotation — were attached to the exposed port side of the hull. A platform structure was built on the seabed beneath the vessel to receive it as it came upright, since no natural surface existed there to take the load. Strand jacks, chains and anchoring systems applied a slow, controlled pulling force, working with the sponsons' ballast to bring the hull through its arc. All of that had been modelled, tested and reviewed long before the rotation began.

114,500 GT
Hull tonnage
65°
Starting list angle
20 months
Time partially submerged
≈19–20 hrs
Rotation duration
Sept 2013
Parbuckling completed
The condition — a load case that kept changing
A structure that would not hold still
A vessel that had spent nearly two years partially submerged, exposed to tidal movement, storms and continuous seawater ingress, does not hold still as an engineering object. Compartments that had flooded did not flood uniformly; furniture, fittings and debris inside the hull had shifted with every storm that passed through the Tyrrhenian Sea; corrosion had been working on the steel throughout. The mass distribution engineers were calculating against was an estimate of a state that had been degrading and rearranging itself for twenty months.
The response — verification during the event
Instrumented as closely as it was pulled
The hull was instrumented and its movement and structural response were monitored continuously as the arc progressed, giving the engineering team a live read on whether the structure was behaving inside the range they had modelled. This is a different kind of verification than the kind used on a new-build structure, where a calculation can be checked against known material and geometry before anyone commits to it. Here, verification had to happen during the operation itself.
The margin — slow enough to intervene
A rotation designed to be stoppable
Rather than pursue more precision in the initial model — a limit that a degrading, partially-submerged structure could never fully remove — the engineering answer was a monitoring regime capable of catching the gap between prediction and reality in near enough real time to intervene, and a rotation slow enough that intervention remained physically possible throughout.

Engineers were not solving for a known structure under known load. They were solving for a structure whose load case was itself the unknown.

Schematic cross-section illustrating the parbuckling principle used to right the Costa Concordia: buoyancy sponsons attached to the hull, strand jacks and chains applying a pulling force, and an underwater platform built to receive the hull.
Schematic illustration — not to scale, not a technical drawing of the actual vessel or salvage equipment. Diagram: EE&HL Network, 2026.
Rows of strand jack towers on the salvage platform beside the capsized Costa Concordia at Isola del Giglio, shortly before the September 2013 parbuckling operation.
Strand jack towers mounted on the salvage platform ahead of the Costa Concordia parbuckling, Isola del Giglio, 18 August 2013. Resized from the original for web display.
Photo: Disma Ballabio / Wikimedia Commons, CC BY 3.0.

It is worth sitting with what that meant for the team responsible for the reaction forces at each sponson. They committed to a structural design knowing the input to their own calculation would keep shifting for the entire duration of the operation. The vessel came upright over roughly nineteen to twenty hours, was refloated the following July, and was towed to Genoa for dismantling; total costs for the wreck-removal project are widely reported at well over a billion euros. The harder figure to state cleanly is the one nobody could have written down in advance — how much of the reaction-force calculation at each sponson was, by necessity, an estimate the operation itself was designed to correct in progress rather than confirm beforehand.

For related context on structural monitoring under uncertainty, see the EE&HL case study on Troll A. For the wider sector context, see Heavy Lift Engineering Offshore.

Related case studies: Troll A · Bullwinkle · Gullfaks C

Sources: gCaptain — contemporaneous operational reporting on the Titan-Micoperi consortium and rotation timeline; Marine Log — sponson and strand-jack operational detail; Wikipedia — general background, alongside the primary reporting above. Total project cost figures vary by source and by what is included (defueling, salvage, and dismantling are sometimes reported separately); reported totals range from approximately €1.5–2 billion.

Frequently Asked Questions

Parbuckling is a salvage technique that rotates a capsized vessel back upright using controlled pulling forces combined with buoyancy. On the Costa Concordia, steel sponsons (buoyancy caissons, partially ballasted) were attached to the exposed port side of the hull, and strand jacks, chains and anchoring systems applied a slow, controlled pulling force that worked together with the sponsons' ballast to bring the hull through its arc onto a platform structure built on the seabed to receive it.

The Costa Concordia had spent nearly two years partially submerged and exposed to tidal movement, storms and continuous seawater ingress before the rotation began in September 2013. Flooded compartments had not flooded uniformly, loose fittings and debris had shifted with every storm, and corrosion had been acting on the steel throughout. The mass distribution engineers were calculating against was an estimate of a condition that had been degrading and rearranging itself for twenty months, rather than a fixed, known structure.

The hull was instrumented and its movement and structural response were monitored continuously as the rotation progressed, giving the engineering team a live read on whether the structure was behaving inside the range they had modelled. Because the load case could not be fully verified beforehand, the rotation was carried out slowly enough that the operation could be paused or adjusted if the readings drifted outside the modelled range.

The wreck-removal contract was awarded to a joint venture of Titan Salvage (a subsidiary of Crowley Maritime) and the Italian firm Micoperi. The parbuckling rotation began on 16 September 2013 and was completed after roughly nineteen to twenty hours of continuous work on 17 September 2013, bringing the 114,500 GT hull from approximately 65 degrees off vertical to upright.

After the September 2013 rotation, the hull was left resting on the underwater platform while further sponsons were fitted to the previously submerged starboard side. It was refloated the following July using buoyancy from the sponsons and was towed to Genoa for dismantling. Total costs for the wreck-removal project are widely reported at well over a billion euros.