The application of activity-based simulation techniques to model run- way operations at airports is described. The simulation tool used, STROBOSCOPE, is a discrete-event simulation system and program- ming language based on the three-phase activity scanning simulation paradigm. The model developed can be used as a tool to estimate runway capacity, delays, and double runway occupancy instances. The growth of air transportation services continues to outpace the ability to improve the capacity of the National Airport and Airspace System (NAAS) in the United States. According to recent FAA sta- tistics, the number of passengers traveling by air in the United States reached 643 million in 1998 (1). The number of enplanements is ex- pected to grow to 991 million by the year 2010 (1). The congestion at airports continues to grow and, for the past three decades, airport authorities have looked at various ways to efficiently operate aircraft on limited infrastructure resources. Some large airports, such as the Atlanta Hartsfield International (ATL) facility, handle more than 900,000 operations per year alone (1). It is recognized that the capacity of NAAS is a complex combi- nation of the collective capacities of airports, airspace, airlines and assets, and air traffic control. However, runway capacity dictated by large headways between aircraft operating in the vicinity of airports is, without a doubt, a critical component of the overall system that limits capacity. There are numerous tools and methods to estimate airport capacity and delay. During the past two decades, simulation and modeling techniques have become more popular to study aircraft runway operations at airports with the development of several macro- scopic and microscopic, fast-time simulation models. Macroscopic models such as the Airport Capacity Model (ACM), RDSIM, and DELAYS developed at the Massachusetts Institute of Technology and the Runway Capacity Model developed by the Logistics Management Institute (LMI) can be used to make policy decisions about the best runway operational practices at an airport. Microscopic models such as the Total Airport and Airspace Model (TAAM) and SIMMOD, the FAA airport and airspace simulation model, can handle detailed run- way operations but at an added computational and detailed user cost. Both macroscopic and microscopic models are described in good detail by Odoni et al. (2) and in various sources in the literature (3–5). Each of these models uses variations of the following model- ing paradigms: (a) analytical solutions to a queuing model (case of DELAYS), (b) capacity approximations based on time-space ap- proaches (Airport Capacity Model, LMI Runway Capacity Model), and (c) event-driven, discrete-event simulation (SIMMOD, TAAM, RDSIM). This paper describes a variation of these classical approaches to predict runway capacity and delays at airports using a general pur- pose, activity-based simulation system called STROBOSCOPE (6). STROBOSCOPE is a discrete-event simulation system and program- ming language based on the three-phase activity scanning simula- tion paradigm that has been used to model numerous construction engineering operations (6). PROBLEM STATEMENT This section describes the modeling of a runway whose details are provided here in their entirety so that the reader can re-create the modeling process and the results. Aircraft use runways to land and depart at an airport. When air- craft arrive in the vicinity of an airport, they wait for air traffic con- trollers to give them permission to land. After obtaining permission to land, aircraft enter a final approach corridor of specific length (usu- ally called common approach path), land, occupy the runway while decelerating, and finally exit to a taxiway. For aircraft departures, air- craft wait on a taxiway for instructions to enter the runway, acceler- ate, and take off. The common approach path in this simulation study is 15 km in length. As aircraft traverse this common approach path, they gener- ate wake turbulence. This turbulence must dissipate before another airplane can traverse the common approach path. The necessary dis- sipation time depends on the type of the aircraft that creates the tur- bulence and the prevailing atmospheric conditions in the vicinity of the airport, among other factors. Large airplanes create and tolerate more turbulence than smaller ones. The wake vortex phenomena, including simple models for use in airport capacity analysis, have been described in the literature ( 7–10). In the current NAAS, air traffic controllers separate aircraft using ground-based surveillance radars. To ensure that airplanes do not encounter wake turbulence beyond the one they can tolerate, air traf- fic controllers comply with prescribed minimum separation standards. The minimum separation distances between successive aircraft depend on the types of the aircraft and include a “buffer” distance that acts as a safety factor. This buffer distance is prescribed as 2100 m in length but is subject to measurement errors on the part of the air traffic con- troller. For this reason the actual buffer is assumed to be normally distributed with a mean of 2100 m and a standard deviation of 1260 m, but truncated two standard deviations at either side of the mean (this is actually a mixed distribution). The minimum distance between successive aircraft in the United States is shown in Table 1. When the trailing airplane is slower than the leading airplane, the minimum separation occurs when the Modeling Airside Airport Operations Using General-Purpose, Activity-Based, Discrete-Event Simulation Tools Julio C. Martinez, Antonio A. Trani, and Photios G. Ioannou J. C. Martinez and A. A. Trani, Virginia Tech, 200 Patton Hall, Blacksburg, VA 24061-0105. P. G. Ioannou, University of Michigan, 2354 G. G. Brown Bldg., Ann Arbor, MI 48109-2125. Transportation Research Record 1744 ■ 65 Paper No. 01-3476