Abstract:
Facility requirements describe how the capacity of a facility should be adjusted over time to meet
the expected future demand levels. Practitioners use them to determine the strategic development
of airport passenger terminal facilities. The generation of facility requirements is extraordinarily
complex, since (i) airport strategic planning is subject to high levels of uncertainty due to the
extremely long planning horizons considered, and (ii) investments in infrastructure are subject
to irreversibility. This study presents a strategic capacity planning framework consisting of two
modules, by means of which stochastically optimal facility requirements for airport passenger
terminal facilities can be determined. The demand module is applied first. Its purpose is twofold:
on the one hand, to create annual aggregated demand scenarios of an airport by means of geometric
Brownian motion. On the other hand, to convert these scenarios into facility-specific design hour
loads with the help of linear regression models. Subsequently, the capacity expansion problem
module is used to determine conventional and flexible facility requirements that maximize the
net present value of an airport passenger terminal facility. For this purpose, both conventional
and flexible capacity expansion problem models, presented in the literature, are adapted to the
needs of airport strategic planning. Subsequently, they are solved with evolutionary optimization
algorithms. The framework is applied to a real-world planning example of the existing check-
in facilities at Zurich Airport. The aim of the planning example is to compare flexible facility
requirements with conventional facility requirements in terms of their economic value, and to
investigate how sensitive the proposed models are to variations in several input factors. The results
suggest that flexible facility requirements are generally more valuable than conventional facility
requirements. Moreover, the models applied in this study respond to changes in input factors in a
similar way to comparable models documented in the literature.