An empirical study to determine freight travel time at a major port

This paper examines the reliability measures of freight travel time on urban arterials that provide access to an international seaport. The findings indicate that the reliability index calculated by the median of travel time, which is less sensitive to extreme values in a highly skewed distribution, is more appropriate. This paper also examines several statistical distributions of travel time to determine the best fit to the data of freight trips. The results of goodness-of-fit tests indicate that the log-logistic is the best statistical function for freight travel time during the midday off-peak period. However, the lognormal distribution represents a better fit to arterials with heavily congested traffic during peak periods. Additionally, travel time prediction models identify the relationships between travel time, speeds and other factors that affect travel time reliability. The analysis suggests that incident-induced delays and speed fluctuations primarily contributed to the unreliability of freight movement on the urban arterials.