Thor Nitnirund: The Web Sling That Was Never on the Register

On 20 March 2025, the bulk carrier Thor Nitnirund was drifting in the Cook Strait, New Zealand, in rough weather. Cargo lashing chains had slipped to the side of the ship. The crew improvised a lifting system to retrieve them. A component of the system failed. A crew member was struck and sustained serious head injuries.

The TAIC (Transport Accident Investigation Commission, New Zealand) investigation found several contributing factors: no risk assessment, inadequate supervision, and an inadequate safety briefing. But the immediate cause was specific: a web sling that had not been inspected and was not listed in the vessel's Lifting Gear Register.

Nobody knew what condition it was in. Nobody had checked. Nobody had a system in place that would have required them to check.

What the investigation found

The pressure to act was real. Cargo lashing chains adrift in rough weather is a genuine problem — they represent a hazard to the crew, potential damage to the vessel, and cargo liability. The crew were not being reckless for its own sake. They were trying to solve a problem under time pressure, in difficult conditions, with the tools available.

The tools available included a web sling of unknown age, unknown condition, and unknown load history. It was not on the Lifting Gear Register. It had not been inspected at any recorded interval. There was no system in place that would have identified it as a piece of equipment requiring inspection — because the system didn't know it existed.

Beyond the sling itself, the operation had no formal risk assessment — even a brief, verbal dynamic risk assessment covering the hazards, the method, and a stop-work trigger. Supervision was inadequate. The safety briefing was inadequate. But those failures are, in some sense, downstream of the foundational one: equipment of unknown condition was used to support a load, with a person in the load path, and nothing in the safety management system caught it.

TAIC recommended actions

1. Every item of lifting equipment must be on the Lifting Gear Register — fixed and portable, including all slings (web, wire, chain), shackles, hooks, blocks, eye bolts, and spreader bars. If it can be used to lift a load, it is lifting gear. If it is lifting gear, it must be on the register with its SWL, last inspection date, condition record, and next inspection due date.
2. No lifting operation should proceed with equipment that is not on the register. If a sling cannot be found on the register, it should not be used. The correct response is to stop, identify or dispose of the unregistered item, and either register it with a proper inspection or remove it from the vessel.
3. Improvised lifting systems require the same risk controls as planned operations. The improvised nature of the operation did not reduce the hazard — it increased it. A dynamic risk assessment, even brief and verbal, is still required. Stop-work authority must be explicit and understood by all team members.
4. Supervision of non-routine deck operations must be active and present. Being available nearby is not supervision. The responsible officer must verify the equipment, the method, and the team's understanding before the operation begins, and must maintain direct situational awareness throughout.
5. When purpose-built equipment is not available for a required task, the correct response is to stop and plan — not to improvise. Accept the temporary consequence. Secure the situation as safely as possible. Recover properly with the right equipment when it is available.

Human element analysis

Improvisation under time pressure. Time pressure is one of the most consistent precursors to improvisation, and improvisation is one of the most consistent precursors to lifting casualties. The pressure to solve the problem — right now, with what's available — bypasses the systematic controls that exist precisely to catch this failure mode. The crew were not behaving irrationally. They were behaving the way people under pressure behave. The safety management system's job is to impose a pause that the pressure of the moment will not.

Invisible equipment. The web sling did not exist in the vessel's safety management system. It had no inspection history, no condition record, no SWL on file. It was, from the SMS's perspective, not there. This is a gap that audits and port state inspections can miss — an item not on the register doesn't appear in a register check. The only way to find it is a physical inventory of all lifting equipment onboard against the register.

The gap between supervision and delegation. The TAIC investigation found supervision was inadequate. Adequate supervision of a non-routine lifting operation requires the responsible officer to be present, to have verified the equipment and method before the operation starts, and to retain explicit authority and obligation to stop it. Delegating to a qualified bosun and remaining nearby is not the same thing.

How this pattern repeats

Aviation — American Airlines Flight 191 (1979): Maintenance crews at American Airlines had developed a shortcut to the approved engine-pylon removal procedure — saving 200 man-hours per aircraft. The shortcut introduced stress fractures that were invisible without the standard disassembly process. The modified procedure was not in the maintenance manual, not formally approved, and not visible to the safety management system until a pylon separated on takeoff. 273 people died. The parallel: equipment and procedures operating outside the formal system are invisible to the controls designed to catch failures in them.

Rail — Hatfield derailment (2000): A broken rail at Hatfield had been identified as requiring replacement. The replacement was delayed. The broken rail was known to maintenance teams but not systematically tracked in a way that made the risk visible to the people responsible for operations. When the rail failed under a high-speed train, four people died. The parallel: the hazard existed in physical reality but not in the management system — and what the system doesn't know, it can't control.

Offshore — Piper Alpha (1988): A permit-to-work system existed. On the night of the fire, the crew working on a condensate pump did not know that a pressure safety valve on the same system had been removed and blanked off — the information existed on a permit, but the permit was not communicated to the night shift taking over. Equipment state and permit state were not synchronised. 167 people died. The parallel: the safety system contained the relevant information, but it was not accessible to the people making decisions in real time.

Try the training scenarios

We've built five knowledge-check questions from the Thor Nitnirund investigation — covering Lifting Gear Register requirements, dynamic risk assessment, supervision obligations, stop-work authority, and improvisation under pressure. They're free to try on Crew Connect, built from the primary TAIC source material.

Available in Knowledge Checks, Safety module. For branching incident scenarios where your decisions change the outcome, try the Decision Simulator at crew-connect.net/simulator-training.