New Zealand / Transport

'Serious concerns' raised about safety of Pūhoi-Ōkahu highway north of Auckland

14:41 pm on 15 September 2020

Serious safety concerns have been highlighted in the design of two highway viaducts being built north of Auckland.

Photo: 123RF File photo

Documents show at one stage, a leading engineering firm warned of a critical risk of "multiple fatalities" among construction workers, but it later dropped this concern.

However, the Transport Agency said the project on State Highway One at Pūhoi and Ōkahu had been fully checked out, the design was strong enough and there were no unaddressed safety concerns.

The 15-to-20m-high, four-lane, 300m-long viaducts north of Auckland use a slender deck panel design across wide spans that is unusual in this country.

Timeline of safety concerns on the Puhoi-Warkworth highway project Photo: RNZ/Vinay Ranchhod

The documents released under the Official Information Act show that calculations and computer modelling by two experienced engineers laid out "serious concerns" the concrete panels might fail in either of two ways: buckling quickly during construction, or cracking gradually as traffic runs over them.

They show that later, the panels passed load testing - though they deflected (bent) more than expected - and the firm the two engineers worked for dropped its concern about immediate fatal failure.

However, there is very little in the OIA record about the second major concern, of long-term cracking reducing the durability of viaducts meant to last 100 years.

Veil drawn back

The documents show who was worried, and just what was done in the lead-up to the seven-tonne, up-to-10m-long panels being hoisted up onto the viaduct girders - an install which began in July and continues now.

Such visibility is unusual: the Transport Agency and its contractors, as a rule, strive to keep major project problems hush-hush.

The agency said in June that the safety and durability concerns were raised by a "former employee" of a project consultant.

However, the documents show it was a current employee at the time - an unnamed engineer with 30 years' experience.

He went directly to NZTA in early March; that move probably put him in breach of contracted confidentiality clauses.

He said in emails to the agency that the design was "fraught with danger".

His concerns got backing from a second unnamed veteran engineer who worked with him at the multinational consulting engineering firm Mott MacDonald.

The firm, twice, in March and April, told the project's managers of its "serious concerns"; two dozen times its 17-page report uses the word "failure" regarding steel trusses embedded in the deck panels.

There are also three pages of calculations.

The trusses are there only to take the weight of wet concrete poured on top of the panels during construction.

Lots of the trusses were already on-site, but not installed, at the time the warnings were made.

With the project behind schedule, Mott MacDonald got an "urgent" request from the builders to approve the steel trusses, emails said.

The whistle-blowing engineer repeatedly offered to come up with alternative designs, while noting he expected resistance due to the investment already made in the trusses.

The transport agency said in June that, in response to the concerns, two independent reviews were done, one for it, one for the alliance, NX2.

NX2 is behind the public-private partnership project (PPP) that aims to eventually connect Auckland to Warkworth with a new highway.

Both reviews came to the same conclusion: the design and construction were "fit for purpose", NZTA said.

But the OIA documents show only one review was done.

It is only one-page long - two lengthy paragraphs - and there are no calculations attached.

In it, the reviewer, a specialist bridge engineer with a PhD, suggested the viaduct design, while new to New Zealand, was routinely used overseas and "should be fine".

But he was inconclusive about safety.

"Without understanding the design basis of the deck slabs and reviewing the design, I cannot comment definitively on whether they are safe in the temporary or permanent state," the reviewer told NZTA on 21 May.

The OIA documents suggest the reviewer was not sent Mott MacDonald's full report to consider.

The Transport Agency has insisted that a second review was done (though not, it amended, for the NX2 alliance, but for the construction joint venture of Fletcher and Acciona, a building firm from Spain).

However, the document NZTA refers to is, in fact, Mott MacDonald's 27 March report raising the concerns.

"There are no outstanding or unaddressed safety issues relating to the design of the viaducts," the agency said repeated statements last week.

"The same system is used routinely in Australia and around the world for bridge construction," it said, referring to an example.

But Mott MacDonald's engineers noted the overseas examples it had been shown "generally ... have much smaller spans" than Ōkahu and Pūhoi.

The agency stressed how many checks had been made along the way, during planning but also doing surveys during construction.

Mott MacDonald said Acciona had used an analysis of the deck design that "does not represent the true nature of how concrete will perform".

Engineering opinions often do vary.

Panels pass load testing

Other checks have been done.

Two independent engineers in March approved the panels against Australasian standards but recommended load testing.

Mott MacDonald also recommended load tests, and to go further, with a review of the entire truss design.

In June the deck panels were tested pre-install, with seven tests of four lots of panels.

Loaded with up to three times the expected weight, there was no sign of failure.

"Based on the sample set of panels tested, there is an appropriate factor of safety in the temporary design case," Mott MacDonald said in July after it reviewed the tests.

The firm dropped its concerns about short-term panel failure.

By then, according to emails in the OIA, the engineer who first raised the alarm had been asked to leave Mott MacDonald, and had done so.

RNZ has approached the firm for comment but had no response.

Contracts usually forbid contractors and consultants speaking to the media.

The load tests did yield a surprise: the panels deflected (bent) more than twice as much (26mm) as allowed for (10mm).

Mott MacDonald suggested on 6 July this could affect the panels' long-term performance.

It alerted the viaduct's structural designers, Beca and Tonkin and Taylor, to this.

Within 36 hours, the designers issued a one-line email saying the deflection was okay, and would not reduce the strength of the deck panels once they were all joined up.

The Transport Agency said three successful concrete pours topping off the panels on the girders since July proved the deck panels were up to standard.

Important quotes to note

  • "Heavily welded reinforcement is not permitted in bridge decks ... it is strongly recommended that NX2 prove the resilience and demonstrate sufficient margins of safety." - whistleblowing engineer (unnamed), 17 March
  • "The possible failure combined with the extreme consequence of failure, which is multiple fatalities, represents a critical risk which must be fully addressed before further construction of the design deck system proceeds." - Mott MacDonald, 27 March
  • "This system is routinely used in Australia and around the world and there are several proprietary brands ... which you can Google if you want to find out more." - Independent reviewer to NZTA on 21 May
  • "...there was no failure of the precast panels during loading" - Mott MacDonald, 6 July
  • "The project confirms that there are no outstanding or unaddressed safety issues relating to the design of the viaducts, and categorically rejects any suggestion to the contrary." - NZTA, 11 September