Modeling the Dynamics of Propagated Flight Delay

Abstract

The air transportation industry is becoming of greater importance to the world’s economy, since a growing part of the global population is able to afford air travel. However, the current network is limited in its’ capacity and therefore, more and more delay is experienced within daily operations. Research has shown that flightdelays have a high economic impact on the air transport industry. Therefore, the understanding of delay and how it propagates within a network is thus of relevance for the industry. A lot of research has been performed on identifying the causes of delay, as well as their impact on stakeholders’ costs and passengers’ satisfaction. However, it is of equal importance to characterize the propagation of delays within a network of airports thus identifying the role and the sphere of influence airports have with respect to the network. Based on this a research objective has been formulated as follows, The research objective of this thesis assignment is to investigate whether different sources of primary delay induce contrasting dynamics of propagated delay within a network of airportssimulated by a stochastic queuing network model. Based on this research objective, a new model has been proposed, which uses queues to simulate the different processes aircraft experience during the day, within the context of the United States National Aviation System. Each airport within the network has been represented by three queuing systems, which simulate the arrival, turnaround and departure of aircraft. Based on an empirical database of the US domestic market, the queuing parameters have been determined, which enables the simulation of both local queuing delay as well as propagated delay airports received from other airports. To fulfill the research objective, three experiments have been designed, which will be used to test the performance, accuracy and capabilities of the model. First, six days have been selectedwith differentweatherconditions. Then, a day of the week aggregation is performed to test the difference between different days of the week. Finally, a case study has been performed to see the influence of zero, one, and five airports underlow IFR conditions. Based on the results of the simulation model, it can be concluded that themodel is capable of simulating the relationships between airports and their delay sources with the identification of which airports are delaygenerators and which airports are delay receivers. Furthermore, the case study showed that it makes a big difference if one, five or zero airports are affected by capacity limitations. In the scenario with five airports it even resulted in a network-wide effect with propagation of delay from the East coast until the West coast of the United States. Altogether, this project demonstrated that with a relatively simple queuing model, the dynamics of propagated delay could be simulated within a network of airports, but is less capable tomimic the behavior under extreme conditions. At the same time, the model has shown to be able to provide more information on the propagation of delay and its’ source. Moreover, this study showed the identification of the different natural roles airports have within the network.