Golden Ray — Anatomy of a Stability Capsize

The Golden Ray, a 656-foot pure car and truck carrier with more than 4,100 vehicles aboard, capsized in St. Simons Sound less than 40 minutes after leaving Brunswick, Georgia on 8 September 2019. The NTSB found the cause was not weather, grounding or fire but a data error: the chief officer mis-entered ballast quantities into the stability program, and nothing in the operator's system caught it before the ship sailed.

What happened

A routine outbound transit became a total loss inside an hour.

• The 656-ft vehicle carrier sailed from the Port of Brunswick with 4,100-plus vehicles, 23 crew and a pilot aboard.
• Less than 40 minutes out, during a 68° turn to starboard in St. Simons Sound, the ship heeled rapidly to port.
• It reached a 60° heel to port in under a minute despite counter-measures, then grounded outside the channel.
• Two crew were seriously injured; four engineering crew were trapped in the engine spaces for nearly 40 hours before rescue. All 24 aboard survived.

The cause: a stability calculation no one checked

The NTSB's probable cause was the chief officer's error entering ballast quantities into the stability calculation program, which gave him an incorrect determination of the vessel's stability and left the Golden Ray with an insufficient righting arm to counter the forces of the turn. At the moment of capsize the ship's metacentric height was 5.8 feet — below the 8.3 feet required and below the roughly 8 feet the chief officer believed he had. The report found the vessel did not meet international stability standards at departure. Contributing to the loss: the operator's safety-management system had no effective procedure to verify a stability calculation before sailing.

Why the loading interface keeps producing car-carrier casualties

The Golden Ray is a stability casualty, but its shape is familiar: the risk concentrated at the interface before the voyage, not on it. As with the Höegh Xiamen — where the NTSB found the cause in loading-time battery handling — the failure here was a pre-departure human input that no independent check caught. Car carriers load and discharge fast, with stability and cargo condition changing at every port call, so the window where one unverified number becomes a total loss is at the terminal and the departure calculation, not mid-ocean. The common thread across car-carrier losses is not the hazard type; it is whether the data at the interface was right, and whether anyone checked it.

The severity tail: when removal dwarfs the hull

The insurance shape of the Golden Ray is the same lesson underwriters draw from fire casualties. The hull was declared a total loss at $62.5 million and the cargo at $142 million — but the wreck removal alone was reported at about $842 million and climbing, with total salvage and environmental cleanup exceeding $1 billion. Removal ran to more than thirteen times the hull value. The heavy-lift VB-10000 cut the wreck into eight sections over two years, with the final section lifted on 25 October 2021. Whether the trigger is a battery in thermal runaway or a mis-entered ballast figure, the hull is the small number on a car-carrier casualty; the wreck, cargo and environmental tail is what sets the exposure.

What it means for owners, class and underwriters

The fix the NTSB pointed to is unglamorous and auditable: an independent second check of the departure stability calculation, written into the safety-management system rather than left to one officer's data entry. For owners and class, that is the same principle detection brings to the fire problem — verified, reproducible data at the point where car-carrier risk is actually decided. For underwriters, the Golden Ray is another data point that a car-carrier casualty is priced on the tail, not the hull, and that the cheapest control is usually a check that confirms the pre-departure numbers are real.

Sources

• NTSB — Marine Accident Report MAR-21/03, "Capsizing of Roll-on/Roll-off Vehicle Carrier Golden Ray, St. Simons Sound, Brunswick, Georgia, September 8, 2019" (investigation DCA19FM048): probable cause (ballast-entry error; SMS verification gap), GM 5.8 ft vs 8.3 ft required, sequence of the capsize, $62.5M hull / $142M cargo — ntsb.gov.
• NTSB press release NR20210914 — "Inaccurate Stability Calculations Caused Capsizing of Vehicle Carrier Golden Ray" (14 September 2021).
• The Brunswick News / TradeWinds / gCaptain — wreck-removal cost ~$842M (rising), total salvage + cleanup above $1 billion; VB-10000 eight-section cut; final section removed 25 October 2021.
• [VERIFY: total salvage/cleanup quantum reported in the press ranged and kept rising; treat the >$1B figure as a reported estimate, not a settled final number.]