Port Economics
modelling Level 3

Port Economics

A port is a node with two sets of edges: maritime connections to global shipping lanes, and landside connections to the domestic hinterland. Its competitive position depends on both. Canada's four major ports — Vancouver, Montreal, Prince Rupert, and Halifax — occupy different positions in this geography, and the competition between them for hinterland capture is reshaping Canadian freight flows.

Prerequisites: Throughput functions, queuing theory, inventory in transit, hinterland capture

Updated 12 min read

1. The Question

What determines how much traffic a port captures, and from which geographic hinterland? Why is Prince Rupert — a small city of 12,000 people on the northern British Columbia coast — one of the fastest-growing container ports in North America? And why do shippers in Saskatchewan sometimes choose Vancouver over Montreal for routing goods to European markets, even though Montreal is geographically closer?

These questions belong to the economics of port competition, which is a branch of transport economics with a specific mathematical structure. Ports compete not on price alone — container handling rates are relatively transparent and competitive — but on the total generalised cost of moving a container from its origin to its destination, including the maritime leg, port handling time, and the landside rail or road movement to or from the interior. The port that minimises total generalised cost for a given origin-destination pair wins that flow. The geographic locus of shippers indifferent between two competing ports — where the total costs are exactly equal — is the port’s competitive boundary, and the region inside that boundary is its hinterland.

2. Canada’s Port Hierarchy

Canada’s container port system is organised around four major gateways, each serving a distinct geographic catchment and connected to different maritime trade lanes.

Vancouver: Roberts Bank, Centerm, Vanterm

Vancouver handled approximately 3.3 million TEUs in 2023, making it by far Canada’s largest container port. The port complex is actually three separate terminals on the southern British Columbia coast: the Roberts Bank terminal complex (including the Deltaport and Westshore bulk terminals), Centerm, and Vanterm. Together they handle Asian trade dominated by consumer electronics, manufactured goods, vehicles, and machinery inbound, and Prairie grain, potash, coal, lumber, and other bulk and break-bulk commodities outbound.

Vancouver’s landside connectivity is provided by both CN and CP, giving it the competitive advantage of rail network redundancy that Prince Rupert lacks. Two competing railways create pricing competition on the landside leg, and two networks provide routing alternatives when one is disrupted. Vancouver’s disadvantage is its geography: it sits at the southern end of a coastal mountain corridor that constrains rail capacity and creates bottlenecks when volume peaks. It is also further from Shanghai and other major Asian ports than Prince Rupert — a disadvantage that compounds with container shipping rates on longer voyages.

Prince Rupert: Fairview Container Terminal

Prince Rupert handled 870,000 TEUs in 2023, and the trajectory is steeply upward. The Fairview Container Terminal opened in 2007 with initial capacity of 500,000 TEUs and has been expanded several times. A Phase 2 expansion approved in 2023 will bring capacity to approximately 1.8 million TEUs when complete.

The economic case for Prince Rupert rests on a specific arithmetic. The port is 2.4 sailing days closer to Shanghai than Vancouver. On a containership with an operating cost of approximately USD$1,800 per day, 2.4 days of sailing equals USD$4,320 saved per voyage. A Panamax vessel carrying approximately 4,000 TEUs would yield a per-TEU maritime cost advantage of roughly USD$1.08 — or approximately CAD$1.50 at typical exchange rates. For certain commodity flows where time value is not paramount, this maritime cost advantage more than offsets Prince Rupert’s slightly longer landside rail transit to major inland destinations.

CN’s direct rail connection from Prince Rupert to Chicago runs approximately 4.5 days compared to 4 days from Vancouver — a 0.5-day landside disadvantage that partially but not completely offsets the 2.4-day maritime advantage. For most Pacific container flows, Prince Rupert’s net total-cost advantage is positive and growing as terminal capacity expands and CN’s service frequency increases.

Montreal: Inland Port on the St. Lawrence

Montreal handled 1.57 million TEUs in 2023, making it Canada’s second-largest container port. Its distinctive characteristic is its inland location — the port sits at kilometre 1,600 of the St. Lawrence Seaway, approximately 1,600 kilometres from open ocean. Deep-draught ocean vessels can reach Montreal directly, making it a true inland ocean port.

Montreal serves Ontario and Quebec for European trade, and provides a competitive alternative to Halifax for Atlantic Canada shippers. Its rail connections run westward to Toronto and beyond via both CN and CP, and the port’s proximity to the Ontario manufacturing base — considerably closer than Halifax — gives it a landside cost advantage for central Canadian origins and destinations.

Halifax: Atlantic Gateway and European Proximity

Halifax handled 535,000 TEUs in 2023 from two terminals: Halterm and Fairview Cove. Halifax’s competitive advantage is proximity to Europe: it is approximately 2 days closer to northern European ports than Montreal, and many days closer than Pacific Gateway ports. For Atlantic Canadian shippers sending goods to Europe, Halifax is the obvious choice. For Ontario and Prairie shippers, the calculus is more complex: the shorter maritime leg to Europe must be weighed against the longer landside rail movement from interior origins to Halifax.

3. Throughput Economics

Berth productivity

The fundamental metric of port productivity is the number of container moves per crane-hour. A container move is the loading or unloading of one TEU from a vessel. World-class productivity at major Asian ports (Busan, Shanghai, Singapore) runs at 35 to 40 moves per crane-hour. North American ports typically operate at lower rates, partly due to labour agreements, partly due to older equipment, and partly due to different vessel size mixes.

Vancouver averages approximately 28 moves per crane-hour across its terminals. Prince Rupert’s Fairview terminal, with newer equipment and a more streamlined labour agreement, achieves approximately 32. The difference may sound small, but at scale it is significant: at 28 vs 32 moves per crane-hour, across multiple cranes operating around the clock, the productivity differential translates into meaningful differences in vessel turnaround time and annual throughput capacity.

The throughput function

Port throughput capacity — the maximum number of TEUs a port can handle in a year — is determined by the product of several factors:

\[T_{max} = n_{berths} \times n_{cranes} \times m_{ph} \times h_{op}\]

where $n_{berths}$ is the number of berths capable of handling container vessels, $n_{cranes}$ is the number of ship-to-shore cranes per berth (typically 2-4), $m_{ph}$ is the moves per crane-hour, and $h_{op}$ is the annual operating hours (typically around 7,000-8,000 for a round-the-clock operation with maintenance downtime). Container dwell time — the average time a container spends in the port between vessel unloading and landside pickup — also affects effective throughput: high dwell reduces yard space available for incoming containers and constrains overall throughput below the theoretical maximum.

4. Queuing Theory and Port Congestion

The M/M/1 model applied to ports

Ship arrivals at a port can be modelled as a Poisson process with rate $\lambda$ (ships per day), where the Poisson assumption captures the random, approximately memoryless nature of vessel arrivals under typical operating conditions. Vessel service (berth occupancy time) follows an approximately exponential distribution with rate $\mu$ (vessels processed per day per berth). The single-server M/M/1 queuing model — or its multi-server M/M/c extension for ports with multiple berths — provides the foundational analytics.

The key quantity is the server utilisation:

\[\rho = \frac{\lambda}{\mu}\]

When $\rho < 1$, the port can handle the incoming vessel traffic on average and the queue remains bounded. The expected waiting time in the queue before entering service is:

\[W_q = \frac{\rho}{\mu(1 - \rho)}\]

This expression has a crucial property: as $\rho \to 1$, $W_q \to \infty$. Congestion grows explosively as utilisation approaches 100%. A port running at 80% utilisation has manageable queues; the same port running at 90% has queues roughly twice as long; at 95%, roughly five times as long. The non-linearity of queuing congestion is why ports invest in capacity well before they appear to be “full” — and why the supply chain crisis of 2021-22 was so severe when it hit.

The Vancouver 2021-22 crisis

During the 2021-22 global supply chain disruption, the Port of Vancouver experienced a utilisation rate approaching $\rho \approx 0.92$ — close enough to saturation that the queuing formula predicts severe congestion. Average vessel waiting time before berth access rose from the normal 1-2 days to 8-12 days. Container dwell times in the port yard extended from a normal 4-5 days to 12-15 days, further reducing effective throughput and creating a feedback loop: high dwell used yard space, reducing crane productivity, increasing vessel wait times, which further delayed container pickup.

The crisis illustrated both the non-linearity of queuing congestion and the systemic nature of port bottlenecks. The proximate cause was a surge in import demand following pandemic restrictions. But the structural cause was that Pacific Gateway infrastructure had been running at high utilisation for several years, with limited buffer capacity. Prince Rupert, with lower initial utilisation and newer facilities, experienced considerably less disruption than Vancouver during the same period — a demonstration that geographic redundancy in port infrastructure has measurable economic value when the primary gateway becomes saturated.

Prince Rupert grew nearly 190% from 2010 to 2023, compared to Vancouver’s 65%. Unlike Vancouver, Prince Rupert saw minimal throughput decline during COVID, partly because its lower pre-crisis utilisation left more buffer capacity.

6. Port Hinterland Geography

The geographic logic of hinterland competition: Vancouver and Prince Rupert compete for Prairie and Ontario freight to Asia; Montreal and Halifax compete for eastern Canadian freight to Europe. The hinterland boundary between Pacific and Atlantic gateways for grain exports from Saskatchewan shifts depending on the relative cost of CN rail to the two coasts.

7. Asia-Pacific Transit Time to Chicago

Prince Rupert’s 2-day ocean advantage over Vancouver is partially offset by its 0.5-day longer inland rail to Chicago. Net advantage to Prince Rupert: approximately 1.5 days total, with an additional maritime cost saving from fewer vessel operating days.

8. The Hinterland Competition Model

The hinterland boundary between two competing ports $A$ and $B$ is the set of origin points where the total generalised cost of routing through port $A$ equals the total generalised cost of routing through port $B$:

\[C_A(o) = C_B(o)\]

where $C_A(o)$ is the total cost (maritime + port handling + landside transport) for a shipper at origin $o$ routing through port $A$. The boundary is a curve in geographic space; all origins on one side prefer port $A$, all origins on the other prefer port $B$.

For Vancouver vs Prince Rupert, the boundary for Asia-Pacific container flows sits somewhere in the Canadian prairies, shifting east as Prince Rupert’s capacity expands and CN’s service frequency improves. As the boundary moves east, Prince Rupert captures more Prairie origins that previously defaulted to Vancouver. This is the commercial geography of port competition: not a sudden shift but a gradual boundary migration driven by infrastructure investment and service quality improvement.

The hinterland concept also explains why port infrastructure investment is strategic: a new terminal or improved rail service at port $A$ shifts the boundary in port $A$’s favour, capturing hinterland from port $B$ and generating revenue to justify the investment. Port competition is, in this sense, a geographic game played with infrastructure as the primary instrument.

References

Port of Halifax. 2024. “Steady 2024 Cargo Results at Port of Halifax.” Halifax: Halifax Port Authority. https://www.porthalifax.ca/steady-2024-cargo-results-at-port-of-halifax/

Port of Montreal. 2024. Historical, Cumulative and Detailed Statistics. Montreal: Montreal Port Authority. https://www.port-montreal.com/en/goods/overview/statistics

Prince Rupert Port Authority. 2024. “Port of Prince Rupert Handles 23.5 Million Tonnes of Cargo in 2023, a 5 Percent Annual Decline.” Prince Rupert: Prince Rupert Port Authority. https://www.rupertport.com/port-of-prince-rupert-handles-23-5-million-tonnes-of-cargo-in-2023-a-5-percent-annual-decline/

Prince Rupert Port Authority. 2024. 2023 Annual Report. Prince Rupert: Prince Rupert Port Authority. https://2023.rupertport.com/

Transport Canada. 2024. Transportation in Canada 2023. Ottawa: Transport Canada. https://tc.canada.ca/en/corporate-services/transparency/corporate-management-reporting/transportation-canada-annual-reports/transportation-canada-2023

Vancouver Fraser Port Authority. 2024. “Strong Canadian Exports Support Record 2023 Trade Volumes Through the Port of Vancouver.” Vancouver: Vancouver Fraser Port Authority. https://www.portvancouver.com/article/strong-canadian-exports-support-record-2023-trade-volumes-through-port-vancouver

Vancouver Fraser Port Authority. 2024. 2023 Statistics Overview. Vancouver: Vancouver Fraser Port Authority. https://www.portvancouver.com/about/reports-and-resources

References