Travel time variations over time and routes: endogenous congestion with degradable capacities

Abstract

ABSTRACTTo investigate how the travel time variability develops over time and routes, in relation to mean travel time, this paper analytically explores the departures with day-to-day capacity variations. Congestion is treated as an endogenous phenomenon. An either three-regime or two-regime solution case is defined, dependent on a combined effect of different capacity variations and cost parameter values. Rigorous formulations and condition of each solution situation are provided. A more complete trip making process is modelled as well considering joint route and departure time decisions of travellers in response to uncertainty. Examples with different capacity variations and cost parameters are presented with solutions graphically shown, for both single- and parallel-bottleneck models. It appears that larger capacity variation yields higher travel cost to travellers due to higher early schedule delay cost and unreliability cost. Travel time unreliability cost is not neglectable in transport economics appraisal. In a long-time period, travel time variability is increasing over time, while mean travel time is decreasing at the same time. There is no evidence on the proportional relations between travel time variability and mean travel time over time and routes.