Incorporating Delay Effects into Airport Runway Pavement Management Systems

This paper presents an approach to address pavement management decision problems at airports with multiple runways. It relaxes some of the underlying assumptions made in previous studies, and explicitly considers time requirement for runway reconstruction, deterioration dependence on traffic levels, and the growth of traffic demand over time. A finite-horizon dynamic program is formulated to investigate the interplays among M&R action time, functional interdependence between runways, and traffic growth. Results from computational studies reveal these interplays, in particular the trade-off between present M&R action and delay cost and long-term benefits brought by significantly upgrading pavement condition through reconstructing runways. Sensitivity analyses suggest that baseline demand, demand growth rate, and the parameter differentiating traffic-dependent transition probabilities significantly affect optimal M&R decisions and total expected cost-to-go. 1 Graduate Student Researcher, Department of Civil and Environmental Engineering, Univ. of California – Berkeley, 107E McLaughlin Hall, CA, 94720-1720. E-mail: [email protected] 2 Professor, Department of Civil and Environmental Engineering, Univ. of California – Berkeley, 110 McLaughlin Hall, Berkeley, CA 94720-1720. E-mail: [email protected] 2 CE Database subject headings: Infrastructure; Pavement management; Airport and airfield runways; Traffic delay Introduction Airport runway infrastructure management refers to the process of making decisions concerning the maintenance and repair (M&R) actions to apply to runways. The objective of making these decisions is to ensure both the functional and operational performance of runways, taking into account agency cost, user costs and M&R associated delay cost. The latter stems from the functional interdependence of runways, and the fact that a certain amount of time is required for M&R actions to be completed. In road pavement management research, functional interdependence has become a research topic of increasing importance. One stream of studies is related to scheduling planned M&R activities to minimize traffic delays. Fwa et al. (1998) employ genetic algorithms to solve a 20-section road network problem. Chang et al. (2000) adopt dynamic traffic assignment models and a Tabu search methodology to minimize traffic delays caused by work zone combinations. Simulation has been used to address similar problems (Sanford-Bernhardt and McNeil 2004). Wang et al. (2002) develop a hybrid scheduling methodology combining genetic algorithms and microscopic traffic simulation. A second stream of studies accounts for functional interdependence in the M&R planning process by jointly minimizing agency and user delay cost. Adey et al. (2003) 3 determine the optimal policy that minimizes the total cost on a transportation network consisting of five bridges, taking into consideration the functional role and performance of the bridges in the network and the consequences if an adequate level of service is not provided. Hajdin and Lindenmann (2007) present a network optimization model that yields minimum overall agency and user cost through intervention bundles on complementary sections in a road corridor. Durango and Sarutipand (2007) formulate quadratic programs to study both substitutable and complementary structures in transportation infrastructure networks, and demonstrate that optimality requires coordinating M&R interventions for clusters of facilities in transportation systems. Among practical projects where delay is of particular concern, researchers find that the delay elements are surprisingly large in the overall cost. For example, when developing work zone models, Moavenzadeh (1971) and Butler (1973) find that delay costs alone overwhelm the agency costs associated with pavement reconstruction and rehabilitation. HDM III, a widely used Highway design and Maintenance Model, shows that the user costs represent more than 85% of the total discounted project costs over a 20-year period (Watanatada et al., 1988). In the airport runway pavement context, functional interdependence refers to traffic re-allocation due to M&R actions and associated runway closure. This interdependence has been qualitatively recognized in McNerney and Harrison (1995). At Dallas/Fort Worth International Airport the Air Transport Association estimated the cost of closing runway 18R/36L in 1990 to range between $110,000 and $131,000 per day, due to airline delays (Lary et al. 1991). The substantial cost notwithstanding, efforts to