Nicola Willis at Parliament. (File photo) Photo: RNZ / Samuel Rillstone
Analysis - A fortnight after being announced as Finance Minister, Nicola Willis was telling the nation about the dramatically escalating costs of the iReX project.
Less than a week later, she confirmed the cancellation of $750 million in funding proposed by her predecessor, officials scrambling to communicate that to their Korean counterparts.
Running on a platform of cutting wasteful spending would have incentivised Willis to scrap a project with ballooning costs.
Those costs were not helped by the timing of the election, and the previous government's focus on keeping options open for their successors.
Documents show officials were also actively encouraging the new coalition to shut the project down after years of secret mistrust of KiwiRail's information and objectives.
But ultimately, the new government's Cabinet made the decision to rescind funding, knowing full well it would likely mean going back to the drawing board - and without an alternative plan.
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.
Heading for the rocks: Officials' push for iReX's end
Shortly after being confirmed as ministers of the new coalition government, Willis and state-owned Enterprises minister Paul Goldsmith received a briefing from Treasury highlighting the urgent need to decide about funding for iReX.
"The project is at a critical juncture," the briefing said. It warned of rising costs that would only grow if a decision was delayed.
"The later a decision is made to cancel the ship contracts, the higher the break fee and sunk costs will likely be... Significant project milestones, for example steel cutting and keel laying, and associated payments due in early-2024, will significantly increase project break fees and potentially decrease optionality."
The officials said cancelling the previous government's offer of extra funding would likely mean the project being shut down and going back to the drawing board, but that it seemed like the best option - particularly given the coalition's promises to cut costs.
"The level of Crown support needs to be considered in the context of wider government objectives," the briefing said.
"Officials agree that Interislander investment is required to meet medium- to long-term service needs, if KiwiRail is to continue to provide this service. However, a fundamental reset on how to address ageing ferry and landside assets is required to provide a more affordable and commercially sustainable solution.
"We would recommend ministers, in line with existing Cabinet authority, reverse the in-principle decision from the then-Cabinet to increase the [funding].
"KiwiRail would be expected to exit its existing obligations under Project iReX."
Willis announced that decision two weeks later to the day.
Troubled waters: The growing mistrust of KiwiRail's information
The documents show officials did not quite trust the information from KiwiRail, particularly relating to the company's claims about the benefits of rail-enabled ships.
"KiwiRail's incentives are not likely to be aligned with delivering the most efficient Cook Strait crossing connection for passengers and freight. Rather KiwiRail has incentives to optimise its rail connection, which has had a significant impact on landside costs," that first briefing to Willis and Goldsmith said.
A train crossing the Rangitata Rail Bridge in April. Photo: Supplied/ KiwiRail
And, in a separate Ministry of Transport (MOT) briefing to their minister Simeon Brown around the same time:
"Key outstanding questions include: what landside investment would be required if KiwiRail (or another provider) wanted to provide a service like Bluebridge or the current Interislander (minus rail enablement)? Why this investment would not be provided by the ports as part of their ongoing role in maintaining and providing these assets? Why market prices for ferry services would not be expected to rise to the level required to support further investment?"
One document from Treasury officials shows they worried that KiwiRail "will view ministers eventually accepting the project largely in its current form as a fait accompli".
The mistrust was bad enough that the officials had considered - under Labour - tactics like having the government increase funding for the project without telling KiwiRail.
"An option to increase the tagged contingency to [$1b] as part of Budget 2023, but not communicate this to KiwiRail yet, was considered but is not recommended."
And when the new government arrived, they advised the ministers to make a decision without bothering to talk to KiwiRail about it first.
"We do not think ministers need to wait for further information from KiwiRail to make this decision."
Willis later said she had met with KiwiRail several times about the matter.
Officials' suspicion of KiwiRail's claims was nothing new. Back in June 2021 - before the ferry contract was fully signed off - officials said that while it was "responsible asset management" for the company to replace the ferries, they were concerned about the information they had been getting.
"Officials and [Treasury's contractors] AECOM continue to have several concerns with the information provided in the DBC (detailed business case) ... the focus going forward should be on how to manage and mitigate the remaining risks in the project."
By mid-March 2023, ahead of that year's Budget - Grant Robertson's last as finance minister - the concerns about KiwiRail's approach did not seem to have improved.
"KiwiRail has not considered, in detail, any scenario whereby the ship configuration (and contract) is amended, stating that renegotiating the contract would expose [project funding] to cost increase. It is not clear if this cost increase would be more than offset by reduced landside infrastructure costs."
Treasury then advised the shareholding ministers to send a letter to KiwiRail demanding further information.
"The letter is drafted in relatively strong language to reinforce to KiwiRail the importance of taking ministers' concerns seriously in providing a response, such that it is not assumed that further Crown funding will inevitably be provided," the report said.
The officials at the time considered further delays to the signing of port build contracts the best solution while they sought better information over the next couple of months before the Budget.
"This will delay KiwiRail's desired entry into landside infrastructure contracts, which we advise is acceptable and, given our assessment of the risks and associated costs, on balance preferable."
But by the time the coalition government came in, officials wanted KiwiRail to "go back to the drawing board".
"Focus on delivering a more cost-effective level of resilience, to remove rail enablement unless it is demonstrated to deliver marginal commercial benefits, and minimise, and transparently identify, connection costs (freight and road transport). KiwiRail should prepare a programme that has a positive NPV, is as commercial as possible, and minimises calls on shareholder funds."
All above board? What KiwiRail knew
KiwiRail, on the other hand, thought it had been providing plenty of information - much of it checked by independent analysts. A letter from chair David McLean to the ministers - with the heads of the Treasury and MOT copied in - shows they were taken aback to hear otherwise.
"We were not aware before our meeting last week of any concerns by Treasury about the veracity of any of the information we have provided, and we take very seriously the suggestion that it lacked veracity. I will be meeting with the Secretary of the Treasury next week and intend to raise our concerns about this advice directly with her," he said.
The letter, sent two days before Willis' public warning about increasing costs, was accompanied by details of the information KiwiRail had provided.
"For instance, 12 deep dive sessions were held with officials and AECOM on options with 120 files transferred providing detailed analysis."
The attachment also listed 25 deeper questions and responses, a list of 38 folders and documents provided, and another list of 31 external consultants brought on to verify and support KiwiRail's information.
Shopping for ships
Slides from a KiwiRail presentation also show they considered eight alternative options to the two large rail-enabled ships, but time and again chose two large rail-enabled ships as their best option. As a state-owned enterprise, KiwiRail is expected to be as profitable and efficient as a comparable business not owned by the Crown.
It estimated its preferred option would cost $1.889b in port upgrades, compared to for two large non-rail ships at $1.806b, or three medium rail ferries at $1.787b.
It considered the possibility of second-hand ships, but concluded "ongoing operating costs will be higher (three vs two, less fuel efficient being older, higher maintenance costs) and they will need replacing more quickly than new ships. Contract exit costs will also need to be considered".
Buying non-rail ships was considered less safe, would require double-handling for loading and unloading rail freight, and machinery like train carriages would need to be craned onto ships instead. KiwiRail said it may save about $100m in capital, and possibly more "by challenging terminal requirements/terminal design", but would also lead to a "significant increase" in running costs.
Willis has since hinted the government is leaning towards such an option with smaller ferries.
Smaller ships could also have saved about $101m in infrastructure costs, but KiwiRail said "this is more than offset by the higher cost the additional ships of $186m ... or $451m based on the likely market price of a change (effectively new) in the ship contracts". More ships - aiming for similar total capacity - would also mean higher running costs.
A "do minimum" approach with "materially worse" outcomes on sustainability, resilience, growth, guardianship of the Strait, customer experience, safety for customers and staff, and port relations, would still have cost $585m. This would have included building a new wharf, terminal building and other infrastructure in Picton and enhancing the existing facilities in Wellington for short-term second-hand ships.
Officials also thought the Wellington and Picton port companies likely to offer more funding if the ships were smaller and not rail enabled, advising that discussions with the ports should take place "as to what genuine landside development is required to support the Government's objectives, and who is best placed to fund and own that infrastructure".
But KiwiRail had been engaging directly with the ports - both of which are owned by councils - and did not think much of the chances of additional funding.
"There is zero likelihood MDC will enable Port Marlborough to invest any additional capital", they said. "Directors' assessment is that there is low-to-zero likelihood of any [CentrePort] commitment to more than [redacted] before 31 March 2023, and this should not be relied on".
KiwiRail also did not think it could operate simply use smaller ships without upgrading the port facilities, which were no longer fit for purpose.
"There are fundamental issues with current infrastructure. The existing wharf in Picton requires major upgrade which is not possible while continuing to operate the current fleet. Any rail ship would require replacement of the rail linkspans and nesting structures which are end of life. The Wellington wharf has remaining useful life, but with increasing maintenance costs and will require replacement within 15 years. In addition, any new ships of similar size to current will have reduced capacity due to new maritime standards and therefore would not meet medium-term growth needs," the slides said.
A difficult berth: Portside cost escalation
The documents also show disagreement and uncertainty from the ministries over the source of portside infrastructure cost blow-out. Treasury's advice to Willis said landside infrastructure accounted for about $937m of the increase, driven by the need for "larger ferries, with larger piles, stronger sea walls, and greater structural rigidity than estimated".
"The remaining cost escalation relates to increased KiwiRail project and financing costs as well as costs associated with bringing the fully operational date of the ferry services closer to the date when the ferries are expected to be delivered," it said. KiwiRail had merged some business groups and hired new managers for the project after Treasury called for stronger independent assurance and governance arrangements in 2021.
MOT in December had its own assessment of the wharf costs, saying it was "the combination of KiwiRail selecting both large and rail-enabled ferries, that have directly resulted in the land-side cost escalation".
KiwiRail in October 2022 identified four reasons for cost increases: inflation, new seismic codes in Picton, flood risk mitigation in Wellington, and the need for iwi engagement.
When it warned the following February that total costs were set to exceed $2.6b, it highlighted the cost-benefit analysis had decreased from a "very high" ratio of 4.4 down to a "high BCR" of 2.1.
This contrasts with officials' briefings, which focused on the project's "negative Net Present Value", meaning capital costs were higher than the long-term revenues KiwiRail could expect. This went from +$207m in 2021, to -$1.6b in December 2023.
But it does not take into account hard-to-estimate additional public benefits. Treasury noted in 2021 "it is difficult or not possible to measure some of the project's benefits against their costs".
The later advice to Willis also shows Treasury remained suspicious and discounted such benefits, saying there was "little evidence that additional investment ... to achieve secondary benefits delivers value for money."
"Officials, and our advisors AECOM, have considered the information provided by KiwiRail, including on the benefit cost ratio, but note that there remain material concerns with the assumptions used."
Treasury's review in March 2023 showed some of those involved in the project were concerned the port upgrades would be funded as part of the 'above rail' component of KiwiRail's revenue. Other aspects of the rail network - considered 'below rail' - are funded through the National Land Transport Fund.
"Some interviewees are concerned that this does not appropriately recognise the 'public good' element of the investment," Treasury wrote.
But officials were concerned about relying on the NLTF "when it is already constrained".
"Ultimately, these costs are likely to come back to the Crown," Treasury said.
Reversing in the dark: The unknown costs of going backwards
When Willis appeared on RNZ warning of the cost blowout of the project, she said the cost had gone "many times beyond what was initially scoped".
This is partially true: total project costs when the ship contract was signed were about $1.45b, increasing to just under $3b by the time it was cancelled. But what officials did not know - and could not know - was the cost of not continuing.
"The main risk of decreasing Crown funding and Project iReX stopping is the cost of the break fees, which cannot be estimated with precision," Treasury said, referring to the money HMD could extract if KiwiRail backed out of the ship-building contract.
"[KiwiRail] have indicated the likelihood of high break fees ... but KiwiRail have also acknowledged that ferry prices have increased significantly since they signed their fixed price contract, so it is conceivable the ferries could be sold on for a price that does not result is such high loses."
KiwiRail in December was open to reselling the finished ferries and it remains unclear whether the ship build contract has been fully exited or not.
Winding the rest of the project down would also have some costs.
"In the past KiwiRail has indicated the only viable option to unwind Project iReX would be to cancel the big ship contracts and walk away, with a negotiated break fee in excess of [redacted]," the Treasury briefing said.
"Some level of Crown support would be required to assist KiwiRail, [redacted] in exiting the project and pragmatically ending or making safe work that has been undertaken to date."
Willis has been adamant since cancelling the project the government remains committed to a "resilient, safe and reliable ferry service" but securing new ships would also require a new procurement process, contract and design - which KiwiRail thinks would add at least two years.
The government's subsequent decision to bring in an advisory group shows ministers were not fully confident they had all the facts. That group reported back to the government last month, but no solution has yet been identified.
Regardless of the advice, Willis, Goldsmith and the rest of Cabinet made the decision to cancel the additional funding for iReX, in full knowledge this would likely mean scrapping the project.
A Cabinet paper from December also highlights six risks, including:
- KiwiRail's reputation
- New Zealand's reputation and the relationship with Korea
- 'Negative commercial impact' on future ferry procurement
- Contract negotiation risk (including break fees)
- That KiwiRail might exit the Interislander business
- That KiwiRail may need to maintain the current Interislander fleet for longer
"KiwiRail would need to reconsider its strategy for the Interislander service, and there is the risk that further exploration will not yield viable options. KiwiRail advises there are very few ferries available for lease or purchase that would meet KiwiRail's requirements," the paper says.
Like Alexander the Great presented with the Gordian knot, Cabinet saw the tangled mess of KiwiRail's Interislander replacement project and decided it was simpler just to cut away.
What replaces it remains, for now, an open question - as does the final cost of that replacement.
