HIGH OCCUPANCY VEHICLE LANES - NOT ALWAYS MORE EFFECTIVE THAN GENERAL-PURPOSE LANES

The success of a high occupancy vehicle lane in motivating people to s hift to carpools and buses depends on maintaining a travel time differ ential between it and the adjacent general purpose lanes. This differe ntial, in turn, depends on the level of continuing delay on the genera l purpose lanes. Therefore, it is clear that a high occupancy vehicle lane that will motivate people to shift to high occupancy vehicles wil l not eliminate congestion. Consequently, it is not clear that constru cting a high occupancy vehicle lane will necessarily reduce delay more than construction of a general purpose lane. The objective of this re search is to determine the circumstances in which this would be the ca se. The hypothesis is that such circumstances would be quite limited, and this proves to be the case. The intended benefits of high occupanc y lanes are defined as reduced person-delay and reduced emissions. A m odel is developed to calculate these benefits for four alternatives: a dd a high occupancy vehicle lane, add a general purpose lane, convert an existing lane to a high occupancy vehicle lane, and do nothing. The model takes into account the initial conditions, the dynamic nature o f the travel time differential between the high occupancy vehicle lane and other lanes, and the uncertainty regarding the: extent to which p eople will shift modes. It combines queueing theory and mode choice th eory and provides a robust method for comparing alternatives using a s mall amount of easily observed data. Application of the model in typic al situations shows that with initial delays on the order of 15 min or more, adding a high occupancy vehicle lane would provide substantial reductions in delay and some reduction in emissions. However, in a wid e range of such situations, adding a general purpose lane would be eve n more effective. Only if the initial delay is long and the initial pr oportion of high occupancy vehicles falls in a rather narrow range, wo uld an added high occupancy vehicle lane be more effective. The propor tion of high occupancy vehicles must be such that it allows good utili zation of the high occupancy vehicle lane while maintaining a sufficie nt travel time differential to motivate a shift to buses or carpools. Adding a high occupancy vehicle lane to a three lane freeway will be m ore effective than adding a general purpose lane only if the initial m aximum delay is on the order of 35 min or more and the proportion of h igh occupancy vehicles is on the order of 20%. Federal policies encour age construction of high occupancy vehicle lanes and restrict funding for general purpose lanes in areas that have not attained air quality standards. The findings of this research suggest a need to reconsider these policies. (C) 1998 Elsevier Science Ltd. All rights reserved.