Melbourne Metro Tunnel was announced in 2016 with an original budget of $10.9 billion. It opened in December 2025 at a final cost of $13.48 billion — a 23% increase. Delivery ran through two alliances: the Cross Yarra Partnership (John Holland, Lendlease, Bouygues Construction) built the twin nine-kilometre tunnels and five new underground stations, while the Rail Network Alliance (John Holland, CPB Contractors, Alstom, AECOM) handled systems integration and rail infrastructure works.
Five new underground stations were delivered: Arden, Parkville, Anzac, Town Hall and State Library.
Delivery ran into September 2024 as the original contracted completion date, but actual opening slipped to December 2025. The Victorian Auditor-General confirmed by March 2024 that the project's contingency provisions had been insufficient for the risks it was carrying, and COVID-19 was cited as the primary cost driver behind the overall increase.
What this means for planners
Alliance contracting and risk sharing
Both delivery groups here were alliances, not fixed-price contracts. Under an alliance model, cost growth is shared between the alliance partners and the government client under a pre-agreed formula — it changes who absorbs a cost blowout, but it doesn't make the underlying schedule risk disappear. Planners working inside an alliance need to understand the commercial model as much as the network logic, because programme decisions (accelerate here, de-scope there) are also cost-sharing decisions.
Why COVID-19 was different from normal project risk
Most schedule risk registers model risks that are specific to a project — a subcontractor default, a design change, a weather event. COVID-19 was a systemic risk that hit labour availability, supply chains and site access simultaneously across every active project in the state. It's a reminder that a contingency allowance sized for project-specific risk will not cover a macro shock, and that programmes should carry a separate, larger allowance for correlated, market-wide disruption.
Contingency adequacy — the auditor's finding
The Victorian Auditor-General's finding that contingency was insufficient by March 2024 is a structural point, not a one-off: contingency set at the start of a multi-year mega-project is a forecast made years before delivery risk is fully known, and a fixed percentage-of-cost contingency rarely scales with how risk actually evolves over a programme this long.
Check schedule health → floatcheck.html