Modelling Airport Capacity Constraints in Air Travellers' Airport Choice
WINNER OF The Neil Mansfield Award
M Gelhausen, German Aerospace Center (DLR), DE
The purpose of this paper is to develop an enhanced passengers? airport choice model which allows for limited capacity as capacity constraints are becoming even more important in the future and a capacity shortage at one or more airports in an airport sys
Over the next 20 years air transport demand (measured in revenue passenger kilometres, RPKs) is forecasted by the aircraft industry to increase by about 5% per year worldwide and between 4% and 5% per year in Europe (Airbus 2008; Boeing 2007). Thus RPKs may double within the next 15 to 18 years. Eurocontrol (2008) expects the number of flights to increase at an annual rate of 2.2% to 3.5% in Europe until 2030, depending on the future development of various political, environmental and economical factors. Here, the growth factor for 2030 in relation to 2007 lies within a range of 1.7 to 2.2. Capacity constraints already exist today at many airports and are becoming increasingly more important for the future development of air transport demand and supply. Efforts to ease constraints, in particular runway expansions to accommodate higher levels of demand for aircraft movements, take some time until they are actually implemented.
Capacity constraints include not only limited physical infrastructure like runways and terminals but also administrative restrictions like night curfews, noise & emission budgets or noise & emission limits, which all restrict the overall level of air travel demand an airport is potentially able to serve. If available airport capacity lies below the present or future demand potential of a particular airport, the airport choice of individual air travellers will be affected and will thus differ from a no-capacity-constraints case. Here demand potential of an airport is defined as the number of air travellers who choose a particular airport without capacity restraint. However, airport choice varies considerably when travellers are faced with capacity constraints, and thus depends on the gap between demand potential of an airport and the demand at capacity level. Thus it would seem appropriate to incorporate the impact of capacity constraints in a systematic and coherent way when planning studies on future airport choice.
Air travellers? first choice of a departure airport may not necessarily be a realistic one in a capacity-limited airport environment where demand exceeds supply at some airports. Therefore some air travellers will opt for second choice flight offers from other airports. The existence of sufficient supply at every airport, however, is a basic assumption of many passengers? airport choice models.
The purpose of this paper is to develop an enhanced passengers? airport choice model which allows for limited capacity as capacity constraints are becoming even more important in the future and a capacity shortage at one or more airports in an airport system significantly affects individual choice behaviour. The modelling approach aims for a high degree of consistency.
The model is based on the principal of individual utility maximisation and constitutes a refined discrete choice approach. Nonlinear programming is employed to formulate a capacity constrained discrete choice model which is solved by Simulated Annealing. Game theory is employed to ensure a solution of the problem, which represents a stable equilibrium in an interactive decision environment with different parties involved.
Since the model is based on the principle of individual utility maximisation and simulates individual choice behaviour, it is possible to study the effects of limited airport capacity on a microscopic level. Therefore, detailed conclusions on how a particular capacity shortage changes individual airport choice behaviour are possible. This paper shows that a capacity shortage at one airport changes choice behaviour substantially in relation to an unconstrained scenario and its effects reach beyond the constrained airport. The methodology proposed in this paper is general in nature and therefore applicable to other problems of travel behaviour analysis and discrete choice theory. Furthermore, the methodology enables to turn an already existing unconstrained discrete choice model into a model with capacity limits.
Association for European Transport