The two models under examination in this paper include one based on work by Oliver & Samuel and one based on work by Koopman. The former method examines the airport airside queues from a wholly deterministic standpoint using the actual flight times from tower data and average service time for arrival rate and service rate, respectively. The latter method assumes a Poisson arrival rate and defines lower and upper bounds for delays based on a deterministic service rate and a Poisson service rate, respectively. Both models are executed using data from the airport tower at Schiphol Airport in Amsterdam, Holland. Airport capacity is treated as a variable in each of twelve runs which comprise the experiment. Demand is input in 60-minute, 30-minute and 15-minute time intervals for each of four profiles selected from the tower data. The results concur with expectations in that the Oliver 7 Samuel model consistantly predicts delays below the Koopman band. The correlation between the two models improves the greater the peaking in the demand profile. If some of the delay ignored by the Oliver & Samual model could be replaced artificially, the correlation would be excellent. The results are surprising from the standpoint that 30-minute time intervals produce better correlation than either the 60-minute or the 15-minute intervals. While the former case is expected, the latter is not and, in conjunction with other evidence, seems to indicate that spurious delays have entered the Koopman methodology as the time interval is reduced from 30 to 15 minutes. Additional research is required to further substantiate this claim.

  • Corporate Authors:

    Massachusetts Institute of Technology

    Department of Civil Engineering, 77 Massachusetts Avenue
    Cambridge, MA  United States  02139
  • Authors:
    • Scalea, J C
  • Publication Date: 1976-6

Media Info

  • Pagination: 78 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00155947
  • Record Type: Publication
  • Source Agency: Massachusetts Institute of Technology
  • Report/Paper Numbers: MS Thesis
  • Files: TRIS
  • Created Date: Sep 20 1977 12:00AM