Pavement life cycle assessment: from case study to machine learning modeling

Alam, Md Rakibul (2020) Pavement life cycle assessment: from case study to machine learning modeling. Masters thesis, Memorial University of Newfoundland.

[img] [English] PDF - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (4MB)


Climate change is a global challenge with long-term implications. Human activities are changing the global climate system, and the warming of the climate system is undeniable. According to a roadway construction study, the construction of the surface layer of an asphalt pavement alone generates a carbon footprint of 65.8 kg of CO₂ per km. Therefore, a sensible approach to study environmental impact from road pavement is crucial. Pavement life cycle assessment (LCA) is a comprehensive method to evaluate the environmental impacts of a pavement section. It features a cradle-to-grave approach assessing critical stages of the pavement’s life. Material production, initial construction, maintenance, use and end of life phases exist in an entire pavement life cycle. The thesis consists of three components, which started with finding the environmental impact for different pavement maintenance and rehabilitation (M&R) techniques in the maintenance phase. The second component evaluated the environmental impact due to pavement vehicle interaction (PVI) in the use phase. Finally, the goal of the third component was to develop a set of pavement LCA models. To evaluate environmental impact for four major M&R techniques: rout and sealing, patching, hot in-place recycling (HIR) and cold in-place recycling (CIR), initially a fractional factorial design approach was applied to determine which factors were significant. Considering those significant factors and other necessary data, a hypothetical LCA case study was performed for the city of St. John’s. It was found that the global warming potential (GWP) held the highest values among four M&R techniques. CIR technique produced the lowest percentage of GWP (83.87%), and for asphalt patching, the CO₂ emission resulted in the highest percentage (92.22%) which became the least suitable option. To understand the PVI effect, the required data and information are collected from the Long-Term Pavement Performance (LTPP) program. Out of 141 Canadian road sections, 22 sections were selected. Several climatic parameters, including annual precipitation, annual temperature, and annual freezing index data, were collected from these 22 sections and further processed for developing clusters using a hierarchical clustering approach. Finally, the Athena Pavement LCA tool was used to measure the environmental impact from the PVI effect for each cluster. It was found that cluster 2 (high annual precipitation, high annual freezing index, and medium annual temperature) experienced the highest rate of IRI increase and therefore, high GWP value. The LCA result also indicated a relatively higher GWP due to pavement roughness from heavy vehicle traffic compared with light vehicle traffic. For the PVI effect due to pavement deflection, cluster 4 (maximum vehicle load and the minimum subgrade stiffness) emitted the highest GWP among all the clusters. Pavement LCA tools require an extensive amount of data to estimate the environmental impact. In the first and second studies, all Canadian road pavement sections were not possible to consider because of the large quantity of time consumption for LCA of each section. Therefore, a database management software, Microsoft SQL Server Management Studio, was used for filtering and data manipulation of the LTPP database considering all Canadian road sections. The manipulated data were further used to develop the LCA models using machine learning algorithms: multiple linear regression, polynomial regression, decision tree regression and support vector regression. The models determined the significant contributors and quantified the CO₂ emission in pavement material production, initial construction, maintenance and use phase. Model validation was also performed. The study also revealed the contribution of Canadian provinces’ CO₂ emission. The proposed LCA models will help the decision-makers in the pavement management system.

Item Type: Thesis (Masters)
Item ID: 14652
Additional Information: Includes bibliographical references (pages 94-101).
Keywords: Pavement, LCA, Life cycle assessment, Machine learning, LCA model, Pavement vehicle interaction, Maintenance and rehabilitation, Environmental emission
Department(s): Engineering and Applied Science, Faculty of
Date: October 2020
Date Type: Submission
Digital Object Identifier (DOI):
Library of Congress Subject Heading: Pavements--Performance; Pavements, Asphalt--Environmental aspects.

Actions (login required)

View Item View Item


Downloads per month over the past year

View more statistics