COMMUNITY BRIEFING NOTE 02
Do the passengers fit on the ferries?
PTA forecasts millions of additional public-transport trips from the ferry expansion. This note tests those forecasts against something rather more basic: the physical capacity of the proposed ferry service.
Damien Smith · 6 September 2026
METRONET on Swan Ferry Expansion Business Case
The issue
PTA forecasts 3.1 million additional public-transport trips a year by 2030, rising to 7.2 million by 2055 for a 20-minute ferry service. For a 30-minute service, the forecast is still 7.1 million additional trips by 2055.[i]
It is unclear how those trips are caused by the proposed ferry.
The ferry capacity limit
The new ferries carry about 100 passengers.
To test the effect of capacity, I modelled service frequencies between 20 and 30 minutes and average vessel utilisation between 10% and 25%, allowing for passengers occupying more than one ferry leg.
That produces approximately 270,000 to 1.0 million ferry passenger journeys a year.
Against this sits PTA’s forecast of up to 7.2 million additional public-transport trips a year.
There may be an explanation through transfers, induced travel or wider network effects. But the published business case does not show the bridge between the number of people who can physically use the ferries and benefits on that scale.
Does STEM constrain demand to available capacity?
PTA says its patronage forecasts were produced using the Strategic Transport Evaluation Model (STEM).
This is important because STEM is a strategic, network-level model.
A 2009 National Transport Modelling Working Group review identified public-transport capacity as an unresolved issue in Perth STEM. It recommended analysis of capacity and demand and said that, if public-transport crowding were to be implemented, appropriate crowding functions would need to be developed.[ii]
That was seventeen years ago and STEM may have changed substantially since then.
But for this project the question is fundamental:
Was the ferry patronage generated by STEM constrained by the number of passengers the proposed ferry service can actually carry?
The published business case does not say.
The published figures are difficult to reconcile
PTA’s published CBA is based on a 20-minute service.i
The same business case says the charging system is being designed around 25-minute frequency.
The current operational application says the service is timetabled at 25 minutes but can be scaled to 30 minutes, and specifically notes that the 30-minute timetable does not rely on the proposed additional speed exemptions.[iii]
Yet PTA’s forecast additional trips fall only from 7.2 million at 20 minutes to 7.1 million at 30 minutes, despite the substantial reduction in service frequency. That needs explaining.
Effect on the business case
Community Briefing Note 01 showed that applying Treasury’s 7% discount rate reduced my reconstructed BCR from PTA’s published 1.7 to about 1.16.
When passenger-related benefits are also constrained by the physical capacity and utilisation of the ferry service, my optimistic case produces a BCR of approximately 0.28.
That case assumes a 20-minute service and 25% average utilisation. Actual utilisation could be materially lower, particularly because the ferry largely competes with existing Transperth bus and rail services rather than serving an entirely new transport market. Lower utilisation or a less frequent service reduces the BCR further.
The 0.28 includes no allowance for loss of recreation or amenity at Matilda Bay.
The counterfactual
My lawyer friends would call this the counterfactual, or the “but for” test. Put simply:
What would public-transport patronage be but for the ferry?
If PTA says the ferry causes up to 7.2 million additional public-transport trips each year, what exactly is different between the with-ferry and without-ferry cases?
How many of those additional trips use the ferry? How many are former motorists? How many occur elsewhere on the public-transport network — and why do they occur because the ferry exists?
Those are the trips that underpin much of the claimed economic benefit.
Two questions for PTA
1. Was the STEM modelling used for the business case constrained by the actual passenger-carrying capacity of the ferry service? If so, what vessel capacity, service frequency, crowding and average utilisation assumptions were used?
2. What is PTA’s counterfactual?
What are forecast public-transport trips in the without-ferry case, and how does the ferry cause the additional trips claimed in the business case? Please identify actual ferry journeys, mode shift from cars and additional trips elsewhere on the network.
About the author
Prepared by Damien Smith, Perth naval architect and Director of Damien Smith Design Pty Ltd.
Disclosures: Damien Smith Design was a tenderer for the Swan River ferry project. Damien is also on the Mounts Bay Sailing Club Management Committee and is an Australian Labor Party member. The views expressed are his own.
[i] . PTA, METRONET on Swan Ferry Service Expansion — Business Case Summary 2025, pp. 10–15. PTA states that STEM was used to model ferry demand and mode shift and reports 7.2 million additional annual public-transport trips in 2055 at 20-minute frequency and 7.1 million at 30 minutes.
[ii] National Transport Modelling Working Group, Critical Review of Transport Modelling Tools (Implementation Options), Sinclair Knight Merz, 2009, s 4.8.2, p. 67. The review recommended analysis of public-transport capacity and demand in Perth STEM and contemplated crowding functions being developed if capacity-constrained public-transport modelling were implemented.
[iii] PTA, Ferry Operations Development Application, 21 August 2026.