Least-cost ground holding strategies with departure and arrival delay uncertainties

Abstract

Ground Holding Programs (GDP) were introduced in the early eighties to lower the air traffic controllers' workload when airports are congested and aircraft have o wait in the air before they can land. The underlying idea is that it is cheaper (i.e. lower operating costs for the airline) and safer to wait on the ground at the origin airport than wait in the air at the destination airport when both departure and arrival airports have uncertain capacities. In the future free flow management environment flights subject to a GDP will be assigned, only a controlled time of arrival (CTA), i.e., target time by which each flight should arrive at the terminal airspace of its airport of destination. It will then be up to the airlines to determine the times at which their flights should leave the gates at their airports of origin, in order to meet the CTAs at the airport of destination. In other words, the FAA will no longer issue estimated departure clearance times or controlled times of departure for flights subject to GDPs. The objective of this research is to assist the airlines in developing a methodology for determining for themselves the optimum gate departure time for a flight aiming at meeting a given CTA. The proposed Departure and Arrival stochastic model (D/A model) takes into account the two main sources of uncertainty namely departure and arrival delay uncertainties. This model builds on and improves considerably some earlier research that dealt only with the uncertainty about delay on arrival. A closed form solution for the optimum ground time is found without making any assumption on the distribution of delays for both the linear and nonlinear cost of delay cases. Finally, numerical computations comparing the performance of the different strategies at hand are presented and the proposed strategy outperforms all others for all cases.