Reliability-Based Geometric Design Approach for Selecting Basic Number of Freeway Lanes

Abstract

Reliability-based geometric design analysis is more suitable for explicitly addressing the level of variability and randomness associated with design inputs than is a more deterministic design approach. Reliability analysis was used to estimate the probability distribution of operational performance that might result from basic decisions made to achieve a design level of service (LOS) for number of lanes on a freeway. The concept was demonstrated with data from I-15 and I-80 in Utah. To account for uncertainty in the design inputs, statistical distributions were developed, and reliability analysis was carried out with Monte Carlo simulation. The outcome of this probabilistic analysis was a distribution of vehicle density for a given number of lanes during the design hour. The main benefit of reliability analysis is that it enables designers to explicitly consider uncertainties in their decision making and to illustrate specific values of the distributions that correspond with the target LOS (e.g., the 65th through 85th percentile density levels correspond to those in the design LOS). The results demonstrate how uncertainty in estimates of the percentage of daily traffic in the design hour, the directional distribution, the percentage of heavy vehicles, and the free-flow speed significantly contribute to variation in vehicle density on a freeway.