A port expansion can move the bottleneck inland

Port investment often attracts attention because its assets are concentrated, observable and easy to illustrate. A loader, berth or crane can be photographed and its local rate can be tested. Inland interfaces may be dispersed across paths, loading points, border processes and maintenance arrangements. Visibility affects how a problem is described, but it does not determine what constrains the declared service. The system decision should compare effective annual capacities and the relevant time-based operating conditions before selecting the most tangible expansion.

The fictional base case makes this distinction deliberately clear. Rail is the annual binding stage while the loading asset has a slightly greater effective capacity. Increasing loading capability alone leaves the common flow unchanged. This is not a criticism of ports or a claim that inland assets are generally the constraint. It is a controlled counterexample to an asset-first inference. The paper's title describes a method of reasoning, not a geographical generalisation about where bottlenecks occur.

A port itself is a chain of functions. Marine access, berth occupation, cargo handling, storage, release and landside transfer can each limit a particular service. An expanded berth can be unusable when its approach, cargo supply or adjacent handling is constrained. More storage can be valuable for segregation without increasing the loader rate. A faster loader can reduce service time for a ready vessel without increasing annual corridor flow. These outcomes should be named separately so that a legitimate local benefit is neither exaggerated nor dismissed.

The investment question should therefore specify the margin being improved. Is the proposal intended to increase accepted annual quantity, reduce vessel exposure, improve recovery, serve a new cargo class or preserve an access obligation? The canonical model answers only the first question under its restricted annual assumptions. The other objectives require evidence and methods representing their mechanisms. A committee can approve a port intervention for a non-throughput reason, but it should not justify that decision with a throughput number that the model does not produce.

The base study calculates 11.90 Mt/year. The vessel-loading intervention increases the nominal loading input to 20.00 Mt/year while preserving its illustrative availability and the rest of the chain. Calculated flow remains 11.90 Mt/year. The unchanged Rail corridor still sets the annual limit. The case contains no capital cost or delivery schedule, so the result should be read as physical sensitivity, not a project recommendation.

Illustrative. The rail-only study increases nominal rail capacity to 20.00 Mt/year. The calculated flow becomes 12.00 Mt/year and Vessel loading becomes binding. The small change in delivered flow relative to the nominal intervention illustrates a near-adjacent constraint. An analyst who presented the added nominal rail rate as added corridor flow would overstate the result. The relevant quantity is what survives the minimum across all unchanged stages.

Illustrative. The coordinated study changes rail and loading nominal capacities together. It calculates 14.40 Mt/year and identifies Port stockyard as the next constraint. This case supplies a different intervention logic: joint relief can achieve a result that neither isolated visible expansion nor the original-stage improvement reaches alone. It still does not establish that simultaneous investment is preferable to a staged sequence. Timing, feasibility, incremental cost and learning value remain unmodelled.

The comparison preserves one cargo quantity and one demand assumption. In an actual proposal, an expansion may introduce a different commodity, customer or handling method. That can be a valid business change, but it is no longer the same intervention comparison. The analyst should write a revised service and cost basis, not attribute every change in output to added capacity. Controlled comparisons are valuable because they separate the mechanism under test from unrelated improvements that would otherwise make the conclusion impossible to challenge.

Moving a constraint can create new burdens. More rail delivery can require additional yard receipt, segregation and reclaim readiness. A greater loader rate can change stock drawdown and the relationship between rail batches and vessel windows. A port expansion that raises the annual bound may therefore need coordinated operating changes before that bound becomes usable. The committee should ask which adjacent interface becomes more demanding and who is responsible for preparing it.

The annual cost comparison reveals fixed-cost dilution, but it excludes the incremental obligations created by intervention. The base operating unit cost is 23.73 USD/t, while the coordinated case is 22.39 USD/t under unchanged cost inputs. This decline is mechanically produced by higher common flow. It is not a forecast tariff or an estimate of investment return. Any actual expansion appraisal must add the new capital and operating consequences rather than treat the illustrative denominator improvement as the entire economic benefit.

The constraint may also move outside the physical chain. Greater output can exceed contracted demand or require permissions that have not been secured. A port capable of loading more cargo is not useful if the customer cannot receive it or if a marine service is unavailable in the relevant window. The expanded service should be tested against eligible demand, access, quality and safety. The annual model's demand input is a bound chosen for the fictional comparison rather than evidence that incremental cargo can be sold or accepted.

A staged plan can use the next constraint as a learning point. Improve the original stage through a bounded operating measure, observe the response at the yard and loader, then revise the physical design if the next limit is confirmed. This sequence may reduce irreversible commitment, but it is not universally preferable. A construction window, shared mobilisation or service obligation can make coordinated works more efficient. The committee should compare staged and joint alternatives on explicit costs and risks, using the annual result as a constraint map rather than a complete investment selector.

The design brief should translate the intended benefit into interface conditions. If added loading capability is justified by shorter vessel service, define cargo readiness, access windows and the measurement boundary for service time. If it is justified by higher annual delivered quantity, define the inland replenishment and yard-release basis that must support it. If it is justified by resilience, define the interruption and recovery duty. These statements allow procurement and commissioning evidence to address the actual service rather than a generic equipment rate.

An acceptance test at the machine can verify local performance under the agreed conditions. A corridor acceptance test must also verify eligible cargo supply, coordinated receipt and loading, documentary release and destination transfer. Both are useful, but they establish different properties. The sponsor should preserve the local equipment warranty while adding a separate system test. Asking a supplier to guarantee uncontrolled inland service is likely to create an unworkable contract; omitting the system test leaves the sponsor without evidence that the investment serves its purpose.

The predicted stockyard constraint provides a concrete receiving test. Ask the fictional yard team to trace an eligible rail arrival through unloading, segregation, stock location and the next loading instruction under the proposed coordinated duty. A failure before unloading points to receiving access or handling; a failure after unloading points to usable storage, cargo compatibility or release to the loader. If both transitions succeed but formations cannot leave, the apparently downstream problem also consumes rail circulation. This decomposition turns a stockyard label into candidate duties that can be challenged. It avoids authorising a larger yard before determining whether the scarce function is storage volume, handling opportunity or resource release.

Separate the port recommendation into a flow proposition and a window proposition. The loading-only fictional case fails the annual-flow proposition because inland rail remains binding. It can still motivate a window study if faster loading would release a berth during a scarce access period. The coordinated case passes the model's annual-flow comparison but may fail that window study if eligible stock does not reach the loader in time. These conclusions can coexist without altering the frozen results. A port brief should identify which proposition justifies its design, the interface test supporting it and the incremental obligations omitted by the annual operating-cost allocation. A window benefit should never be counted again as extra annual tonnes without its own supporting mechanism.