Death of a sailor: A system of failure

By Geoff Cumming

Navy sailor Byron Solomon died at sea during a training exercise on the HMNZS Canterbury. Photo / Supplied
Navy sailor Byron Solomon died at sea during a training exercise on the HMNZS Canterbury. Photo / Supplied

A coroner's ruling that the death of Byron Solomon on HMNZS Canterbury was preventable leaves his family still seeking accountability. Geoff Cumming reports.

When, on June 12, 2007, Prime Minister Helen Clark cut the cake at a Melbourne shipyard to mark the handover of the Navy's newest ship, the top brass sparkle was more than ceremonial. Not only was HMNZS Canterbury the biggest ship in the fleet, the multi-role transporter meant New Zealand's Army, Air Force and Navy could undertake combined sea-land missions under their own steam for the first time.

But the $177 million ship, based on a roll-on roll-off ferry, was riddled with flaws and would need many millions spent on a "get well" programme. Despite concerns that the Canterbury had not been sufficiently tested and its equipment did not meet requirements, the Ministry of Defence had formally taken possession two weeks earlier at the Williamstown yard of main contractor Tenix. The Prime Minister had set a date for the New Zealand launch and there was pressure within the Navy to take delivery to boost flagging morale.

The Canterbury was the first of seven new ships for a Navy realigned to patrol our economic zone and undertake humanitarian as much as military missions.

Four months later, that haste contributed to the first death of a Navy sailor on active duty in 37 years. Byron Solomon, aged 22, drowned when trapped in a Rhib (rigid hulled inflatable) which overturned during launching, with the Canterbury under way off Cape Reinga.

The joint Ministry of Defence/NZ Defence Force team which oversaw the project took a pounding in subsequent inquiries. The telling quote came from the Coles Inquiry: "[The project] has been managed to get the ship into service as soon as possible and has been characterised by shortcomings in project management and governance, and collective wishful thinking."

That it took just shy of four years to complete an inquest into Solomon's death tells another story: a pass-the-buck game that dogged Coroner Brandt Shortland's efforts to explain what happened.

The Solomon family endured two years of pre-inquest meetings as lawyers representing the Navy/MoD, shipbuilders Tenix and Lloyd's Register (who certified the ship as meeting international standards) argued over admissibility of evidence. Finally, last August, Shortland ringfenced issues of design and acquisition.

"What has emerged is a matrix of legal arguments and implied accusations from involved parties resulting in finger-pointing over the many facets of the acquisition process of the HMNZS Canterbury," he wrote. The wrangling centred on design and contractual disputes, technical expertise and compliance issues.

"It is my view these legitimate issues should be debated or argued in either a Royal Commission of Inquiry or some other form of litigation."

At the inquest in February much of the testimony was "handed up" and not read out in court. Byron's father Bill Solomon, who was unable to cross-examine several key witnesses, complained publicly that a large chunk of evidence had been shut out.

Examination of the "handed-up" evidence of naval staff and expert witnesses - and reports and documents from previous inquiries - suggests responsibility for Solomon's death extends beyond the Ministry of Defence and the Navy to include Tenix, Lloyd's Register and the Government of the day.

It was a systemic failure, made worse by a reluctance to pursue those responsible. Prime contractor Tenix oversaw the installation of components which did not meet the Navy's specifications, parts which may individually have met safety requirements but which were incompatible as an integrated system. One component, the self-righting system, was installed despite being subject to a manufacturer's recall.

Surveyor Lloyds Register failed to pick up gaps and errors in certification and was in the dark about some of the Navy's requirements. It signed off components which were not fit for purpose, according to the Navy's expert witness.

The MoD/Defence Force team, under pressure to get the ship delivered, and under-staffed in some technical areas, did not adequately oversee or enforce the contract, subsequent inquiries found.

The rescue-boat launch system selected by Tenix was just one of many serious defects not picked up. It was supposed to be capable of launching a Rhib "fast rescue" boat in moderate seas with the Canterbury making up to 10 knots.

Under the system, the Rhib was lowered into the sea by a hoist wire but remained secured to the ship by a boat rope until the crew released it from the hoist.

But at 10.33am on October 5, 2007, under clear skies and in a slight swell off Cape Reinga, the system went tragically wrong (see graphic).

Once the boat was in the water a quick-release shackle that connected the boat rope opened prematurely, transferring the force of the water to the hoist wire which was still attached to the Rhib.

It took just 11 seconds for the Rhib to capsize. Two crewmen were thrown out, went under the Rhib and surfaced behind the ship.

Byron Solomon and Leading Seaman Dwayne Pakinga, who were struggling to remove an O-ring which connected the Rhib's lifting strops to the hoist, were caught between the strops and a side pontoon as the boat flipped.

With the Rhib being dragged along by the Canterbury, Pakinga managed to get his head above water and pull himself out, bruised and battered. But Solomon was trapped. His automatic lifejacket had inflated and hindered his escape.

On the ship, efforts to reverse the winch to hoist the boat back up failed. The davit operator tried to play out more wire to create some slack but "it started smoking and then it wouldn't work anymore".

Three officers defied standing orders to jump from the moving ship to try to extract Solomon. Petty Officer Mark Taylor eventually found the self-righting device and pulled the cord - it came away in his hands. After 20 minutes, three crew managed to pull Solomon from under the Rhib but it was too late. Taylor subsequently received a bravery medal.

Later inquiries found the O-ring which connected the lifting strops to the hoist wire was oversized and incompatible with the release mechanism. The strops themselves were a hazard and the boat's self-righting device was defective. And the davit used to hoist and lower the Rhib was not designed to operate at the speed the Canterbury was travelling.

In his inquest statement, Commander William Craig, deputy inspector general of the NZ Defence Force, said Tenix failed to recognise the risk of capsize if the boat rope came away with the Rhib still hooked up to the davit wire.

As Solomon's parents stressed, the same premature opening of the bow shackle had occurred the previous day while launching the Canterbury's other Rhib. Incidents with the shackle went back nine years. "Yet no alarm bells rang to say 'we've got an issue'," Bill Solomon said in his inquest statement. "Fate had already given the Navy plenty of warning that the system ... was flawed and nobody did anything."

Bill Solomon, a company manager and consultant, conscious of an employer's health and safety obligtions, found the Navy's reluctance to address identifiable hazards inexcusable.

"Einstein said the definition of insanity was to keep doing the same thing and expect a different result."

So much time pressure was brought to bear that normal planning stages were omitted, Solomon told the hearing.

A design review should have taken place well before the ship was commissioned. The review which followed the drowning identified 24 hazards which could have been identified beforehand.

Mediation with Tenix (now BAE Systems) over warranty issues for the seven Project Protector ships resulted in an $85 million payment - most of it for defects on the Canterbury, says a joint MoD/Defence Force statement.

The deal was a "holistic solution" which settled the issues "without apportioning any amount to any particular claim". No resolution of the certification issues is being pursued with Lloyds Register.

Defence Minister Wayne Mapp does not believe a Royal Commission is justified - which leaves the family weighing up their legal options.

Solomon's mother Jayne Carkeek: "They've all contributed and at this stage they haven't been made accountable - and that's what we want as a family."

The MoD/Defence Force team's lax oversight and lack of technical expertise were highlighted in previous inquiries. But who was responsible for the flawed components and why weren't they picked up during surveying, certification and testing?


Many of the launch system components had not been through proper certification or been tested as an integrated unit, the military inquiries which followed the drowning found.

The ship was built at the Merwede shipyard in Rotterdam and sailed to Melbourne in August 2006 for military fit-out. The Coles Inquiry found that the ship left Rotterdam "before planned acceptance of all essential seagoing functions and the material configuration, functionality and condition of the ship".

Lloyd's Register was contracted by Merwede to undertake classification and certification, verifying that components and systems complied with international safety regulations.

Commander William Craig, deputy inspector general of the Defence Force, in his inquest statement said accurate certification was paramount because it formed the baseline for the Navy's safety case. Under cross-examination, he said the safety case recognised potential dangers in the launch system but there was no time to make changes before acceptance. There was pressure "from the Government down" to get the ship commissioned.

Commander Craig said Tenix was late in delivering an "operational safety case" which meant risks were not picked up at the design stage.

But it was logical for the Navy to assume the selected boats and their integration with the ship were designed to comply with safety rules and that any design issues had been worked through before delivery. After all, he said, the ship had Lloyd's certification.

The ship's second in command, Lieutenant Commander Matt Wray, said it was assumed that the boats and launch equipment had been studied, tested and checked before the ship left Holland.

The Navy's court of inquiry which followed the incident and a subsequent certification audit found neither Rhib was properly certified "either as a rescue boat or as a fast rescue boat".

The distinction is crucial: a fast rescue boat has a launch system with higher specs, capable of launch with the ship making above five knots and in sea conditions which the Navy had specified. But an independent design review, led by Peter Jenkins of UK maritime consultancy Burness Corlett-Three Quays, found the only certification was a certificate of compliance for a rescue boat, based on a prototype Rhib.

Jenkins said Lloyd's Register was responsible for ensuring the two fast rescue boats had valid compliance certificates.

He found the Rhib was "quite unacceptable for its intended use either as a rescue boat or as a fast rescue boat," mainly due to the lifting strop arrangement.

Lieutenant Commander Andrew Curlewis, who provided technical advice to the court of inquiry, in his inquest statement said the lack of certification for the davit winch and the rescue boats were significant concerns. No certification could be found for the O-ring, and the one provided made manual extraction from the hook mouth more difficult.

Three months after the drowning, Lloyd's carried out a certification audit which found multiple omissions and errors. The audit notes that classification does not guarantee fitness for purpose and frequently omits small components where there is good experience of their performance. "For example, strops and other loose gear for rescue boats are not included in [safety at sea] requirements."

But the document rather contradicts itself when it identifies the need for certification for the rescue boats, the davits, davit fall wires and winches, lifting strops, O-ring and shackles - and confirmation that the system was capable of launching the boats "with the vessel heeled either way to 20 degrees".

The report suggests Lloyd's was unaware of the Navy's specification for fast rescue boats. It says contractual arrangements between Tenix and Merwede contained confidentiality requirements preventing direct communication between Lloyd's and the NZ Navy/MoD team. While Lloyd's was made aware of some aspects of the Navy's specifications, "they were not flowed down through the contracting chain."

In a statement issued to the Weekend Herald, Lloyd's said the ship was certified to standards used for commercial shipping "in line with the RNZN's procurement strategy". For the Rhibs, the standards were those designed for merchant ships to recover people in an emergency. Use of the Rhibs for naval boarding purposes was a more demanding scenario which "would not have been considered in the certification of the Rhib".


Expert witness Peter Jenkins said there was no guarantee that a certified fast rescue boat, launching appliance and hook release mechanism would perform satisfactorily as an integrated system. "For this reason, it is essential that a set of trials is carried out on the installed system to verify that it would perform up to requirements."

Director of naval construction Michael Fraser said the Navy later learned that the lifting gear "may have been excluded altogether from the testing and certification of the boat".

Jenkins said if installation and pre-acceptance trials had been carried out, "the deficiencies in the installation and boat type would have been identified and the accident prevented".

Captain Dean McDougall was the Navy's inspector general at the time, with oversight of the new fleet's introduction. He said there was concern that a number of systems were being accepted via design review, which he understood to be only a drawing and specifications check rather than physically checking that the system met contract requirements.

Wray said Tenix was reluctant to let personnel complete any extra training (outside of Tenix-organised courses) on the rescue boats and other sub-systems until the Navy had officially taken delivery. He believed this related to health and safety laws.


Michael Fraser, in his inquest statement, said his approval as the Navy's design acceptance authority was not sought for the Canterbury acquisition.

Lieutenant Commander John Deere was verification and acceptance manager, seconded to the Ministry of Defence. He said it was Tenix's role to plan and manage the verification and acceptance programme. These responsibilities included verifying that components and systems met the Navy's specifications.

Deere said the programme was witnessed by the MoD "as priorities dictate and resources permit" - not all testing was witnessed.

The 2008 Coles Inquiry found that the Navy/MoD project team was way behind on verification and acceptance due to staff shortages.

But with so many faults noted and so many elements untested, why take delivery on May 31, 2007?

"Project staff made it clear to the project manager and project director that the technical issues would present a significant challenge if the ship were accepted," the Coles Review states.

"But the project director [Gary Collier] recommended acceptance, expressing concern that delay could lead to demoralisation and people leaving the service. The imperative reported by the Chief of the Navy to Chief of Defence Force and Secretary of Defence was of the need to get the Canterbury into service."

Inquest statements show that pressure was also felt from the political sphere. "There was certainly lots of interest in getting Canterbury as soon as possible," Wray said. "There was also pressure from the Ministry of Defence to accept the ship, and the Prime Minister had set a date to commission HMNZS Canterbury."

Once it had the ship, the Navy's culture took over. Wray said it was assumed that everything had been planned, designed and checked before the ship left Holland. The ship had a Lloyd's inspection and had been used by professional mariners. "So we had a strong belief that this was our lot so let's make it work."

Wray said the O-ring/hook release issue was being "dealt with in the background" in post-delivery discussions with Lloyd's Register but the Navy continued to work with the system in the meantime. "We believed that this was the design that had been approved and used [pre-acceptance] by professional mariners. So we made it work."

- NZ Herald

Get the news delivered straight to your inbox

Receive the day’s news, sport and entertainment in our daily email newsletter


© Copyright 2017, NZME. Publishing Limited

Assembled by: (static) on production bpcf02 at 28 May 2017 19:29:49 Processing Time: 653ms