Airports, optimised as a system

Airlines, air traffic control, the slot coordinator, the airport operator, ground handlers, border staff, government and local communities each hold different levers, and each measures success in its own way. Each stage of the journey through an airport is owned by someone different, so an investment that looks sound to one owner can make no difference to the system, or move a cost onto someone else.

Running the runway in mixed mode adds movements but removes the respite that runway alternation gives local communities. Declaring capacity close to the physical limit creates more slots but leaves no buffer, so small disruptions become long delays. Most capacity measures cut emissions per passenger but raise the total, so the same scenario reads as progress through one lens and failure through another.

• An illustrative hub. The airport is a generic two-runway hub. Its capacities are representative values rather than the declared capacity of any real airport.
• One peak hour. The model looks at the busiest hour only. It does not cover daily or seasonal patterns, night operations or recovery after disruption.
• A common unit. Converting each stage to passengers per hour makes stages comparable, but it simplifies stages whose real limits are movements, stand-hours or vehicles.
• Indicative delay. The delay curve shows how queues tend to build as a stage approaches full use. It is not a prediction of delay minutes.
• Cost bands. Lever costs and lead times are order-of-magnitude bands to support comparison, not estimates for a scheme.

Applied to a real airport, the same structure would be populated with that airport's declared capacities and operating rules, and tested with the people who run each stage before it informed a decision.