The scrapping of the Interislander ferry replacement project was one of the first decisions the coalition made after coming to power, but some details have sailed under the proverbial radar.
Here, we explore some of those details.
RNZ is looking back at the handling of the 'iReX' Interislander replacement project, closely examining 44 documents that reveal the advice and information provided and how it led to decisions across multiple governments - culminating in the project's cancellation.
KiwiRail has been requesting funding for a replacement for years, in the knowledge the ships "should be renewed between 20 and 30 years of age".
It has three ships.
Aratere is about 25 years old, purpose-built by Spanish ship-builders in 1998. It went to a Singaporean shipworks in 2011 to undergo an extension which included cutting the ship in half and adding a section in the middle, then sewing it all up again. It is also the only ship in KiwiRail's fleet currently able to handle rail. This is also the ship that ran aground last month.
Kaiarahi is about the same age, built for a charter service in Turkey in 1998. It has been chartered by KiwiRail since 2013 when Aratere was out of service.
Kaitaki is about 28 years old, built for Irish Ferries in 1995. KiwiRail is the ship's sixth owner, having purchased it outright in 2017 after nine years of leasing it.
The company considers the ships to be coming to the end of their useful lives next year, 2025.
The port facilities are also near or at the end of their useful lives, with KiwiRail saying they are "life expired and sub-optimally configured due to a history of incremental short-term changes with no redundancy".
Despite the current government putting an emphasis on maintenance while it decides on how to proceed, a cabinet paper from December (on the decision to rescind funding) noted "maintenance will not defer replacement indefinitely".
The Interislander does about 3800 services a year, transporting about 850,000 passengers, 250,000 cars and up to $14 billion worth of freight. KiwiRail also expected freight volumes to increase 1.4 percent each year, partly as a result of increased use of rail.
KiwiRail considered all sorts of things to try to get its iReX project funded.
This included the possibility of selling "surplus land or individual lines of businesses" including Wellington Railway Station, but the company decided against this as unlikely to work.
"There is no significant land now held by KiwiRail that is not needed for current or projected rail operations. Sale of individual business units or the few assets which might not be needed, such as the Wellington Railway Station, would be problematic and/or unlikely to realise any material amounts," a briefing from February last year shows.
They also considered reallocating funding provided for other rail projects - but found most of the money had already been spent or set aside. They suggested one project that might have fit the bill:
"The Marsden Point Rail Link is the only rail project within the New Zealand Upgrade Programme that is not in contracted construction phase ... there is $345 million of funding which is currently unspent."
However, Treasury advised against allowing KiwiRail to reallocate the funding, because of separate ministerial governance structures.
Other options to boost funding for iReX included selling and/or leasing back ferries and rail assets, but the company concluded "a sale and leaseback for existing assets in KiwiRail's fleet would be unattractive to financiers due their age and condition".
Delays to decisions on building the port infrastructure meant the new ships would have to operate without the rail component until 2027 at the earliest.
Under a heading in a report in February, KiwiRail warned ministers of delays to the port rebuilds, in part due to the rising costs.
"Higher structural requirements and overall construction time mean the terminals programme will be 22 months late," it read.
It meant the new rail-enabled ships would be unable to use their rail-enabled capability until mid-2027 at the earliest.
"The delay in terminals infrastructure means the new ships would not be operational and would have to be mothballed until that time, resulting in holding costs until the facilities are ready," ministers were told in February.
As a result, they recommended in February 2023 a "reset" option - bringing the total project from $2.503b to $2.609b - to allow the ships to start being used earlier.
This would have meant reaching an agreement with Bluebridge to move to a new berth in Picton and taking over the Waitohi East berth in the meantime until the new facilities were built.
KiwiRail also looked into how it could provide a "descoped" option for the project using the $750m the previous Labour government put forward in funding just ahead of last year's election.
However, after two months of assessing the option, the company decided in November it would be "not viable".
Part of the reason for this was the passenger experience.
"It would mean no passenger boarding ramp. For safety reasons, passengers would need to board a bus, which would be driven onto the vehicle deck of the ferry, for them to then access the passenger areas of the ferry."
This and the lack of rail connection would also have significantly increased turnaround times for the ferries, meaning less rail freight and passenger capacity.
It would affect KiwiRail's ability to transfer rail equipment like train carriages between the North and South Islands too, and was "likely to impact the viability of the Christchurch to Picton rail line", KiwiRail's documentation said.
The port terminals for the ferries had also been lumped in with the "business" side of KiwiRail's funding model.
That's in contrast to the approach for funding the company's rail network and roads, which are paid for through the National Land Transport Fund.
Interviewees involved in the KiwiRail project told Treasury for its annual "gateway" review in March last year this did not seem like the right approach.
"Increased costs are currently showing a negative NPV for KiwiRail, but the Benefit Cost Ratio (BCR) is still showing a positive result," the report said.
"Interviewees suggested that this is further compounded by the current approach to categorising the terminals and other port facilities as 'above rail' when 'below rail' is funded through the national land transport fund and 'above rail' is funded by KiwiRail.
"Some interviewees are concerned that this does not appropriately recognise the 'public good' element of the investment in terminals as a critical link in a national network."
In May 2023, a KiwiRail presentation pointed out the ship purchase component would be commercially viable - it was the port upgrade that was not.
However, the ships being used do affect how the bits on land - the terminals, wharves and other infrastructure like roading to get cars on the ships - are set up.
"In isolation, replacing the ships is commercially viable without external funding, with ship costs being recouped via operating revenues over the ship's lives. However, the land-side infrastructure is also end-of-life and the high replacement cost (driven in part by higher standards and requirements) is commercially prohibitive. This is not unusual in major public infrastructure projects."
Treasury officials advised ministers against thinking of Interislander as the only Cook Strait connection.
"Care should be taken not to overstate the role that Interislander plays in resilience," they wrote in November.
The Ministry of Transport also warned the government's support for the Interislander could create a more unfair playing field for rival Cook Strait ferry operator Bluebridge - which performs four return sailings with two ships on weekdays - and suggested Bluebridge might buy a third ferry if iReX did not go ahead.
"This would 'almost certainly not occur' if KiwiRail proceeds with Project iReX as it would give it capacity advantages over Bluebridge given the benefits of KiwiRail operating large new ferries which have only entered the market due to a significant government subsidy," a ministry briefing said.
However, a KiwiRail presentation from May last year shows that even if Bluebridge bought a third ship, the on-land infrastructure would still need fixing up.
"Bluebridge may be able to add limited capacity (e.g. one more ship) in a monopoly setting, but without new (and costly) land side infrastructure, total capacity across the Strait would approximately halve - a significant problem for the country.
"In the event of a major service failure ~$10 billion worth of freight and some 750,000 people will be left waiting at Wellington and Picton, without an affordable way across the Strait."
KiwiRail also suggested it was possible for a third player to enter the market, but this would be a "high risk to NZ's freight supply chain".
"It's highly unlikely a new entrant could make the proposition work commercially."
Having been asked by the previous Labour government to find savings and cover as much of the cost of the project on its own books as possible, KiwiRail successfully cut $230m in costs.
This was partly through $164m of net savings through "value engineering", which means reworking the project to provide necessary functions at the lowest possible cost.
KiwiRail said bigger savings could be achieved, but "at the cost of revenue loss and higher operating costs".
Even after this cost saving, the company estimated the project was still $1.2b short.
That estimate also assumed final agreements for the port works could be completed between mid-August and September, that the final funding decision would not affect the ability to start works, and that no new geotechincal problems popped up.
One of Treasury's regular reviews in March 2023 warned ministers it would be useful to catalogue designs and other useful information from iReX in case it needed to go back to the drawing board.
" ... valuable design and programme management artefacts and outputs are being created through the VE (value engineering) processes leading to an acceptable minimum viable solution," the review said.
"If funding is delayed or not secured, then it will be critical that these are captured to inform any future exercise to re-baseline the programme and assess options."
With a priority rating of DO NOW, the review recommended records be kept, saying the information should "be methodically captured to be available to inform any future exercise (if required), to rebaseline the programme and assess options".
The slides from KiwiRail's presentation highlighted a question raised with the company of whether the already contracted-for ships could be completed then sold on to another buyer.
The company's response was that this was possible in theory, "but in practice no".
They would likely have needed hire more people to "increase the size of the ships team", and there was no guarantee the sale would recoup as much was spent on the ships to begin with.
Other reasons given were that this would delay the delivery of new ships by about two years, with a 30 percent increase in price.
"We see this option as largely impractical. Infrastructure upgrades still needed."
The Cabinet paper for the coalition's December decision to scrap the ferry project did not consider the effects on emissions in any detail.
The given reason? Because it was uncertain what other solution would be landed on.
"A Climate Implications of Policy Assessment is not required," the paper stated.
"The transition to more modern ferries that generate fewer emissions will differ based on the decision made and so may result in different emission profiles."