From 1995 an HOV lane is in operation in N-V1 Madrid Corridor with a limit of 2+ occupants. It has improved speed for HOVs and the reliability of bus services in the corridor. At present there are some congestion problems during morning peak period, while the traffic flow is growing some 4% every year. Therefore Madrid Transport Authority is considering different strategies in the HOV lane to avoid future congestion in the facility, such as to raise the cut off limit to 3+. To test the effects of such possibility it has been adjusted a discrete choice model to the corridor situation: it means that the modal choices could be car-solo driver, car 2 occupants, car-3+ occupants, and bus. To feed that model it has been necessary first to do a 1,700 household CATI survey on user behaviour and preferences and among them to select a 98 person sample to perform an SP survey. Answers have been customised providing public transport and cost alternatives in each face to face interview. Three variables have been used in the exercise: trip time, trip cost and "inconvenience" which includes the effect of getting a third occupant, change timetable, etc. This HOV lane is the only experience in Europe in operation for several years. It provides choice patterns for European standards where the public transport mode is a real option. This is a big difference with the American experiences, such as trip length or urban density, etc. O-D matrix and network has been adapted to the situation and implemented in emme/2 model. Results from each run have been adapted according to the corresponding results from the ALOGIT choice model. The forecasted O-D matrix for the year 2005 has been used to test different strategies in the long term managing of the HOV lane. Finally the results allow evaluation of long term policies in terms of trip time, traffic flow, congestion and pollutant emissions and to compare the results among different scenarios.
Abstract