Pavement management systems are experiencing transformational change as data are turned into information for better decision making and as our understanding of the underlying processes and mechanisms driving performance continues to advance. The breadth and depth of the presentations reflect this unique moment in the pavement management field.
Effect of Selected Factors in the Highway Performance Monitoring System on Pavement Condition
James Bryce, Amec Foster WheelerShow Abstract
Gonzalo Rada, Amec Foster Wheeler
Amy Simpson, Amec Foster Wheeler
This paper presents the results of investigating the relationship between many variables contained within the Federal Highway Administration’s (FHWA’s) Highway Performance Monitoring System (HPMS) and the reported condition of pavements. Given the rules developed by FHWA in support of the past two transportation funding bills, reporting accurate pavement condition for input into the HPMS has increased in importance. In light of this, FHWA undertook a study to assess the quality and representativeness of the data contained in the HPMS. This paper presents the results of the part of this study that assessed the relationship between pavement condition and climate, terrain, traffic, urban setting and reported segment length. First, the relationship between the overall pavement condition and these variables are presented to demonstrate their significance. Following this, an improved source for climate information that was used in this analysis is detailed. Finally, a logistic regression model that shows the relative importance of each variable on the predicted pavement condition is detailed. This analysis demonstrates the correlation between each of the reported variables, and the likelihood of a pavement being in a given condition. In addition, this analysis demonstrates a strong relationship between the lengths of the pavement segments and the overall condition of the pavement, showing that short segments tend to be in worse condition.
Effect of Measurement Variability on Overall Pavement Condition in the Highway Performance Monitoring System
James Bryce, Amec Foster WheelerShow Abstract
Gonzalo Rada, Amec Foster Wheeler
This paper presents the results of an investigation into the effect of measurement variability on the reported pavement condition within the Federal Highway Administration’s (FHWA’s) Highway Performance Monitoring System (HPMS). Given the rules developed by FHWA in support of the past two transportation funding bills, reporting accurate pavement condition for input into the HPMS has increased in importance. In light of this, FHWA undertook a study to assess the quality and representativeness of the data contained in the HPMS. This paper presents the results of the portion of the study that assessed the effect of variability in condition measurements on changes in three consecutive years of HPMS data. First, the changes in the condition metrics (i.e., cracking, rutting, faulting and roughness) over the three year time frame is presented. Following this, changes in the overall condition (i.e., good, fair and poor) were investigated, and an approach to predicting whether changes are due to maintenance is detailed. Finally, a logistic regression model that shows the relative importance of changes in each metric on predicting whether maintenance was performed is discussed. Ultimately, the results of the assessment demonstrate that, although significant variability exists in the HPMS data from year to year, this variability is offset by a large sample size. Individual pavement segments can show significant changes in the condition metrics from year-to-year, but the overall network condition is relatively stable.
Multiobjective Metaheuristics for Decision Making in Pavement Management
Lin Chen, Chang'an UniversityShow Abstract
Theunis Henning, University of Auckland
The use of optimization techniques has become a common functionality of Pavement Management software. Where the current applications’ technology mostly caters for single objective optimization, there is an increasing need to incorporate multiple objectives in the future. In the past, economic-based objectives for minimizing life-cycle cost were most commonly used. There is a growing need to include additional objectives in the analysis such as targeting a specific level of service and/or social outcomes from the maintenance planning perspective leading to multi-objective optimization problems (MOO). A significant number of heuristic multi-objective optimization techniques are available, arguably not all being applicable to the pavement management purpose. This paper investigates the efficiency and effectiveness of the most commonly used optimization techniques in pavement and asset management. The outcomes from testing of these techniques suggest a noticeable difference in terms of the quality of the resulting solution and a difference in the efficiency towards reaching the solution.
Maintenance Thresholds for Management of Highway Porous Pavements
Longjia Chu, Chang'an UniversityShow Abstract
Tien Fwa, Chang'an University
Currently there does not exist any sound engineering procedure for determining the maintenance threshold of highway porous pavements to meet their overall structural and functional service requirements. Many highway agencies apply the standard pavement condition and distress criteria developed for conventional asphalt pavements to manage porous pavements. This practice is not ideal because there are unique functional requirements of porous pavements that require different considerations in establishing appropriate additional maintenance thresholds. This study identifies the engineering basis for establishing additional maintenance thresholds for porous pavements, with the aim to ensure that porous pavements are effectively maintained to deliver satisfactorily the intended design functional level of service throughout their service life. The maintenance thresholds for the following two most common functional benefits of porous highway pavements are considered in this study: reduced tire-pavement noise, and improved wet-weather skid resistance. Equipment and techniques are now available to measure and monitor the performance of a porous pavement in these two functional aspects. They can be employed to develop maintenance thresholds for effective management of porous pavements. The method of On Board Sound Intensity (OBSI) for noise measurement is recommended for tire-pavement noise management. For skid resistance management, a more elaborate procedure involving flow permeability measurement and water film thickness estimation is proposed, followed by skid resistance analysis and prediction. The engineering basis and rationale for the proposed procedures are presented. Finally, based on measured data and calculated parameter values, the process of establishing maintenance thresholds is demonstrated with an example application.
Sensitivity Analysis of Performance Metrics to Different Parameters in Pavement Management System
Fengdi Guo, Massachusetts Institute of Technology (MIT)Show Abstract
Omar Swei, University of British Columbia
Jeremy Gregory, Massachusetts Institute of Technology (MIT)
Randolph Kirchain, Massachusetts Institute of Technology (MIT)
A pavement management system is a useful tool for departments of transportation to address the problems of limited budget and aging infrastructures. Previous research focuses mainly on the budget allocation process, trying to improve the optimization algorithms and consider the uncertainty of predictions for pavement deterioration. In any given pavement management system, there are usually many parameters. However, analysis has not been performed to determine the influence of different parameters on the pavement network performance. In this paper, the sensitivity of performance metrics to different parameters is explored based on the interstate pavement network in the U.S. state of Virginia by a probabilistic allocation network model developed at MIT. A statistical method is applied to conduct the sensitivity analysis. The sensitivity of performance metrics to different parameters is decided by p values, and the relative significance of different parameters is compared and ordered by z-score statistics.
A Case Study of Overweight Traffic on Pavement Costs: Application of Overweight Permit Database in Analyzing Impact of Shale Gas Development on Louisiana Roadways
Xiaohui Sun, Louisiana Transportation Research Center (LTRC)Show Abstract
Zhong Wu, Louisiana Transportation Research Center (LTRC)
Kevin Gaspard, Louisiana Department of Transportation and Development
Tyson Rupnow, Louisiana Department of Transportation and Development
Previous studies investigated the overall impact of the shale gas development on infrastructures without differentiating overweight trips from non-overweight ones. This may result in a difficulty in the damage cost recovery of those overweight trips through issuing overweight permits. In addition, the heavy truck trips were distributed either based on assumed origins/destinations with a limited number or simply based on the mileage percentages of different roadway classifications in the network. These assumptions may not reveal the actual situation. Therefore, this study aimed to overcome those disadvantages and estimate the impact of the shale gas related overweight truck trips on Louisiana roadways in the network level. RStudio was employed to extract and rearrange the overweight trips in the Haynesville area during 2006-2016 from the oversize/overweight database in a proper format. Network Analyst in the ArcGIS was utilized to assign these overweight trips directly on the roadway network according to the shortest path method. The vehicle miles travelled (VMT) according to roadway classifications were estimated subsequently, and it was found that there were 9.7 million shale-gas related overweight VMT during the period of 2008-2016, which would translate into a damage cost of 17 million USD. On average, the estimated damage cost due to the overweight trips in the construction of a single well would approximate to be $5,264 and the damage cost per overweight mile travelled would be roughly $1.74. Such cost results were found somewhat higher than the current permit fee practice of overweight trucks in Louisiana. Therefore, future study on a project level cost analysis is warranted.
Estimation of Axle Load Distribution for LTPP Sections Using Traffic and Geographic Data
Pedro Serigos, Amec Foster WheelerShow Abstract
James Bryce, Amec Foster Wheeler
Barbara Ostrom, Wood.
This paper presents the development of a methodology to estimate the axle load distribution as a function of traffic and geographical variables. This methodology was developed as part of a Federal Highway Administration research project which main objective consisted of the estimation of missing values for different traffic data elements from the Long-Term Pavement Performance (LTPP) program’s Pavement Performance Database. The traffic data variables used to estimate the axle load distribution for a pavement section were obtained from LTPP databases while the geographical data variables were generated through use of geographic information system software using metadata from different sources such as the National Transportation Data Atlas.
The proposed methodology uses site-specific data for estimating missing axle load distributions as opposed to the use of defaults national or regional values. The output of the analyses presented in this paper consist of a series of decision trees that can be divided in two groups. The first set of decision trees estimate the axle load spectra using two traffic variables: the Equivalent Single Axle Load, and the annual volume of trucks. The second set of decision trees address the cases for which the two traffic variables are not available and use two geographical variables: distance to a freight network, and distance to intermodal facilities.
Network-Level Pavement Maintenance Decision-Making Optimization Based on Comprehensive Ranking of Road Sections
Hongmei Li, Nanjing Institute of TechnologyShow Abstract
zhen shen, Nanjing Institute of Technology
Qiao Dong, Southeast University
The network level pavement maintenance decision-making is a multi-factor and multi-objective problem, it is usually determined based on limited factors and subjective judgment. This paper proposes an optimization model based on the ranking results of road sections and integer programming for network level highway pavement maintenance decision-making. Based on the Analytic Hierarchical Process (AHP) theory, this study developed a multi-index model capable of incorporating potential pavement maintenance related factors, considering their relative significance and generating an overall ranking for each road section. A total of five pavement maintenance related indices were considered in the study, including pavement overall quality, pavement structural capacity, pavement age, traffic level and road grade. The optimization is to maximize the overall ratio of benefit over cost in the analysis period through the method of integer programming. The computing work could be greatly reduced based on the ranking results of road sections in need of maintenance. A case study on the network level highway maintenance decision-making optimization in Jiangsu Province was conducted to illustrate the proposed procedure. The case study clearly demonstrated the applicability and rationality of the optimization model based on the comprehensive ranking of road sections. The results can be used as a guideline for highway agencies in their network level pavement maintenance decision-making process.
Development of a Cracking Performance Prediction Model for Replaced Concrete Slabs in California
Ashkan Saboori, University of California, DavisShow Abstract
Jeremy Lea, University of California, Davis
Venkata Kannekanti, University of California, Davis
Arash Saboori, University of California, Davis
John Harvey, University of California, Davis
Cracking and faulting are typical distresses in concrete pavements in California. These distresses at their initial stages cause poor ride quality and higher fuel consumption. With further progress, they lead to loss of structural capacity and serious safety issues. To maintain the road’s condition at an acceptable level, the timing of maintenance and preservation operations should be optimized. Slab replacement is a maintenance operation performed to improve the pavement condition by replacing cracked slabs with new ones. However, there is currently no performance prediction model for replaced slabs in the California pavement management system (PMS).
This paper addresses issues with collecting cracking data from the CA pavement network and developing performance prediction models for slab replacement. Cracking data for each project are collected from the California PMS and are used to develop survival and performance models. The models show that survival rate for thinner slabs (with a thickness less than 0.8 ft) drops drastically after seven years compared to thicker slabs that show low failure rates in the first twelve years of service. The performance prediction model is the probability of failure of replaced slabs versus time and is determined based on age, thickness, and traffic. The developed models will be used in the Caltrans Pavement Management System (PaveM) for recommending the best timing for future maintenance. The results will also be used in for conducting life cycle cost analysis (LCCA) and environmental life cycle assessment (LCA).
Investigating Cost-Effective Pavement Maintenance and Rehabilitation Strategies Through Life-Cycle Cost Analysis (LCCA) by Incorporating Variation in Performance Based on Material Types and Traffic Levels for Ontario Highways
Gulfam Jannat, University of WaterlooShow Abstract
Susan Tighe, University of Waterloo
With the increasing trend towards maintenance & rehabilitation (M&R) of pavements in pavement management systems (PMS), it is essential to make cost-effective use of the PMS M&R budget. As such, identification of associated cost-effective M&R treatments is not always simple in most PMS. Thus, a cost-effective pavement M&R approach is required to allocate the limited budget. In this study, a life cycle cost analysis (LCCA) is carried out for alternate pavement treatments incorporating the variation in pavement deterioration due to different traffic levels and material types. This variation is incorporated by considering the improvement in performance for different materials and also the variation in deterioration of condition for different materials along with different traffic levels. The LCCA is carried out for different rehabilitation options for a period of 40 years period. After comparing the net present worth (NPW) value of alternative treatment options, it reveals that the overlay of pavement with Dense Friction Course (DFC) asphalt layer is the most cost-effective choice in the case of higher AADT. On the other hand, overlay with Hot Laid-1 (HL-1) is a cost-effective treatment option for highway sections with lower AADT.
Although the results are related to the Ontario highway system, this can also be applied elsewhere with similar conditions. The outcome of the empirical investigations will result in the adoption of efficient road M&R programs for highways based on realistic performance predictions, which have significant impact on infrastructure asset management.
Key Words: LCCA, Net Present Worth, Maintenance & Rehabilitation
Cost–Benefit Estimate for Skid Resistance Improvements at the Network Level Using Markov Chains
Oscar Daniel Galvis Arce, University of Texas, AustinShow Abstract
Juan Porras-Alvarado, WSP
Zhanmin Zhang, University of Texas, Austin
Safety studies has proven that low values of skid resistance increase crash risk. For this reason, transportation agencies have established minimum skid thresholds to screen projects for further testing or skid resistance improvements. However, Benefit-Cost Ratio (BCR) analyses are not conducted to estimate the potential benefit of such minimum thresholds. Some agencies have conducted BCRs to evaluate skid improvement projects following a before-and-after analysis for a specific project, but BCRs have not been estimated at the network level. The objective of this paper is to provide a framework that estimates the BCR at the network level for defined minimum skid thresholds, where three main issues are addressed. First, a skid deterioration model at the network level is proposed, using a Markov Chain process. Second, the maintenance costs required to keep skid above a defined skid threshold for every pavement section in the network are estimated. Third, the expected economic benefits of crash reduction at the network level are estimated by comparing the expected crashes of the base scenario (no skid improvement) with the scenario of a minimum skid threshold in the network. Using the proposed methodological framework, a sample of highway sections that comprise 564 lane-miles in Texas is evaluated to demonstrate its applicability. As part of the findings from the numerical analysis, a curve between BRC and the minimum skid threshold was established. The established curves can be used by state and local agencies to evaluate the potential economic benefit of minimum skid policies at the network level.
GIS Integrated Methodology for Identifying Pavement Preservation Candidates in South Carolina
Franklin ReedShow Abstract
Jennifer Ogle, Clemson University
Bradley Putman, Clemson University
For the South Carolina Department of Transportation (SCDOT) to have improved abilities to implement pavement preservation, this research developed a process to identify the candidates for preservation from current SCDOT data. This procedure can be utilized to not only identify candidates, but through GIS, it can also provide the decision-maker with a visual representation of the proximity of candidates within a network, which can be useful when developing contracting plans or strategies for pavement preservation. Identifying the candidates can allow the SCDOT to allocate funding to appropriate counties or districts based on the need. It can also help track the overall progress of the pavement preservation program in increasing the number of lane-miles in good condition throughout a network.