Guidelines for Probabilistic Pavement Life Cycle Cost Analysis
Susan Tighe
Abstract
Susan Tighe
Abstract
To select the most appropriate pavement design for a given situation, it is necessary to understand how the pavement properties and in-service conditions relate to performance and life cycle cost. A given design may be most appropriate on one type of road and least appropriate on another type of road. This design selection is further complicated by the advent of new design methodologies, materials, and construction delivery techniques. Life cycle economic analysis is an important tool for comparing alternative treatment strategies. A life cycle analysis can use a deterministic approach, which incorporates a single point value, or it can use a probabilistic approach, which includes a mean, variance, and probability distribution. The probabilistic approach is better suited to describing the uncertainty associated with engineering. The Canadian Strategic Highway Research Program Canadian Long-Term Pavement Performance database and data provided by the Ministry of Transportation of Ontario were used in this analysis. Most construction variables are generally believed to be best described by a normal distribution. However, a lognormal probability distribution is better suited to describing these variables. This best fit is based on both a mathematical examination and a comparison of similar variables such as stocks and real estate values. It is also shown that thickness is a probabilistic variable that should be combined with the cost and incorporated into pavement life cycle costing. Ignoring the lognormal nature of these variables introduces bias into a life cycle cost analysis and does not reflect the true overall cost.
OpenAlex reports 89 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
To select the most appropriate pavement design for a given situation, it is necessary to understand how the pavement properties and in-service conditions relate to performance and life cycle cost. A given design may be most appropriate on one type of road and least appropriate on another type of road. This design selection is further complicated by the advent of new design methodologies, materials, and construction delivery techniques. Life cycle economic analysis is an important tool for comparing alternative treatment strategies. A life cycle analysis can use a deterministic approach, which incorporates a single point value, or it can use a probabilistic approach, which includes a mean, variance, and probability distribution. The probabilistic approach is better suited to describing the uncertainty associated with engineering. The Canadian Strategic Highway Research Program Canadian Long-Term Pavement Performance database and data provided by the Ministry of Transportation of Ontario were used in this analysis. Most construction variables are generally believed to be best described by a normal distribution. However, a lognormal probability distribution is better suited to describing these variables. This best fit is based on both a mathematical examination and a comparison of similar variables such as stocks and real estate values. It is also shown that thickness is a probabilistic variable that should be combined with the cost and incorporated into pavement life cycle costing. Ignoring the lognormal nature of these variables introduces bias into a life cycle cost analysis and does not reflect the true overall cost.
Key concepts: Probabilistic logic, Life-cycle cost analysis, Pavement engineering, Probabilistic analysis of algorithms, Probability distribution, Activity-based costing, Variance (accounting), Log-normal distribution