Risk-based framework for accommodating uncertainty in highway geometric design

Abstract

The development of highway standard design models involves various assumptions regarding design inputs and the road environment. This paper suggests an improvement to the treatment of uncertainty in design inputs by replacing the current deterministic approach with a reliability-based framework. Reliability theory deals with the propagation of quantified variability in design inputs throughout the design process. In such a framework, each design output corresponds to a theoretical probability of noncompliance to design requirements. These probabilities can be used to assess and compare the a priori safety level associated with various design scenarios. This paper proposes that such a priori safety level of standard design outputs should be consistent and close to a prespecified target level. A set of methods is proposed to determine a target value for design safety. A general framework for calibrating standard design models is presented. To demonstrate the concept, the paper presents an application of the calibration framework to the standard design model of crest vertical curves. Calibrated design charts are constructed to yield a consistent design safety level.