|

Calibrating Road Design Guides Using Risk-Based Reliability Analysis

Author(s): Mohamed Hussein, Tarek Sayed, Karim Ismail, Adinda Van Espen

Slidedeck Presentation Only (no paper submitted):

6C Hussein_Calibrating Road Design Guides Using Risk-Based Reliability Analysis

Abstract:

Existing geometric design guides provide deterministic design criteria for highway
elements which ignore the uncertainty associated with many design inputs and model
parameters. Reliability analysis has been advocated as an approach to account for this
uncertainty and to evaluate the risk associated with a particular design feature. This paper
discusses one important application of reliability analysis: the calibration of geometric design
models to yield consistent safety (risk) levels. In order to illustrate the concept, the paper
considered the case of restricted sight distance on horizontal curves due to existence of median
barriers. The paper provides calibrated design charts for the middle ordinate (M), defined as the
lateral distance between the edge of median barriers and the centerline of the inside traffic lane,
at different probability of non-compliance levels. The impact of several factors on the calibrated
design charts are discussed including the use of operating and design speeds and the target
probability of non-compliance.
The results show that the calibrated values of the middle ordinate (M) are generally lower than
those derived from the AASHTO design guide. The AASHTO design guide is so conservative
especially at high speeds and sharp curves. The results elucidate that significant reductions to
current design requirements could be obtained at reasonable reliability (risk) levels. The
calibrated design charts can offer designers dealing with highways with constricted right-of-way
an option to use lower middle ordinate values and enable them to estimate the safety
consequences of their decisions. Overall, the calibrated charts can aid the decision maker in
determining the safety implications of deviating from geometric design standards and
quantifying the safety level built in design values that are deemed acceptable.