Pavement Performance Measures and Forecasting and The Effects of Maintenance and Rehabilitation Strategy on Treatment Effectiveness (Revised)

EXECUTIVE SUMMARY

The efforts of the research team in this study focused on the following two objectives of the Federal Highway Administration (FHWA) Long-Term Pavement Performance (LTPP) Data Analysis Program—Expanded Strategic Plan:⁽¹⁾

• Objective 7: Quantification of the performance impact of specific design features and its problem statement “Common Characteristics of Good and Poorly Performing AC and PCC Pavements.”
  
  
• Objective 9: Comprehensive use of LTPP to improve the management of pavement assets.

These two objectives were addressed and accomplished through the development of the following products:

• Two sets of pavement performance measures (good, fair, and poor) and (very good, good, fair, poor, and very poor) that were used to establish dual pavement condition rating systems based on pavement functional and structural performance over time.
  
  
• Efficient pavement performance quantification and prediction methodologies for pavement maintenance and rehabilitation treatment options.
  
  
• Tools for the evaluation of the roles of pavement preservation/maintenance and rehabilitation in the design of long-life pavement.

This study had the following specific objectives:

• Define pavement performance in a way that supports the selection of cost-effective pavement treatment strategy.
  
  
• Provide better estimates of pavement treatment effectiveness and the role of pavement treatments in the pavement’s service lifecycle.
  
  
• Develop pavement performance prediction methodologies that are applicable to the pavement condition and distress data collected before and after the application of treatments or series of treatments.
  
  
• Analyze whether falling weight deflectometer (FWD) data can be used to indicate impending surface defects.
  
  
• Make recommendations for subsequent studies regarding the impacts and/or selection of pavement maintenance, preservation, and rehabilitation treatment options and strategies and their impacts on the pavement service life.

The dual pavement condition rating systems (described in chapter 3) are based on proposed ranges of the remaining functional period (RFP) and the remaining structural period (RSP). The research team used RFP, RSP, and LTPP-measured time-series pavement condition and distresses to develop efficient pavement performance quantification and prediction methodologies (see chapters 5 and 6). The methodologies were then used to study the benefits of various pavement treatments. The team calculated the pavement treatment benefits using various approaches, the LTPP data, and databases obtained from three State transportation departments—the Colorado Department of Transportation (CDOT), Louisiana Department of Transportation and Development (LADOTD, and the Washington Department of Transportation (WSDOT). The resulting benefits were used to analyze the impacts of pavement design factors (such as asphalt and concrete thickness and base drainage) on pavement performance. Further, the research team calculated and scrutinized the weighted average benefits of treatments applied to all pavement types in each climatic region.

To define the pavement deterioration curve with a reasonable level of certainty, all functional and structural data analyses were based on three or more time-series data points (all available data from the LTPP or State databases). The team found that the LTPP database contains fewer than three International Roughness Index (IRI) and/or distress data points over time for some of the LTPP test sections (see chapter 4 for details). Indeed, in several cases, only one data point had been collected between the applications of consecutive treatments. To increase the number of test sections that could be analyzed, the research team developed the following two procedures (see chapters 5 and 6):

• For asphalt concrete (AC) overlay or mill-and-fill treatments, one data point of 0.0039 inches (0.1 mm) at 0.01 years after construction was added to the rut depth and cracking data. This addition facilitated the analyses of most test sections having only two time-series data points.
  
  
• A novel probabilistic approach was developed using the LTPP data for the estimation of RFP or RSP of pavement sections where only one IRI and/or one distress data point was available in the database. The method was referred to as the One Record Condition State Estimate (ORCSE) method (see chapter 5).

During the study, the research team conducted two sets of analyses. In one set, the LTPP inventory and pavement condition and distress data from LTPP Standard Data Release 28.0 (2014) were used. Results of the analyses included RFP and RSP and the treatment benefits expressed in various terms. In the second set, the inventory and pavement condition and distress data that were measured along various pavement projects by three State transportation departments (CDOT, LADOTD, and WSDOT) were requested, received, reviewed for compatibility with the LTPP data, and analyzed. For each treated pavement project, the treatment benefits were calculated using the same parameters as those used for the LTPP data analyses. The team then compared the results of the analyses of the LTPP and State data. Chapter 8 details the objectives of the comparison, which included the following:

• Determine whether the LTPP data are representative of the State data.
  
  
• Assess whether the dual pavement condition rating systems developed in this study are also applicable to the State data.
  
  
• Evaluate whether the developed treatment benefit measures are applicable to the State data.

In this study, the measured pavement deflection data along flexible and rigid pavement test sections were also studied to determine whether pavement deflection could be used as an indicator of future conditions or surface distresses. Because the deflection data were measured at different times and temperatures, the data for flexible pavement test sections were adjusted to a standard temperature of 70 ºF (21 ºC) using the Asphalt Institute (AI) and other procedures. It was determined that existing procedures were not accurate. Therefore, a new global temperature correction procedure applicable to all deflections measured by the FWD sensors in the four climatic regions was developed (see chapter 7). The impact of this global procedure on the backcalculated layer moduli in flexible pavement was also assessed.

Chapter 9 summarizes the numerous conclusions the research team reached based on the results of the analyses. The following milestone conclusions were reached:

• The dual pavement condition rating systems developed in this study and based on ranges in RFP and RSP were applicable to both the LTPP and State data.
  
  
• For any data collection cycle, the pavement condition and distress data varied from one test section to the next. Hence, the measured pavement condition and distress data along some control sections were not representative of the data along the corresponding test sections.
  
  
• The flexible pavement treatment benefits calculated from the LTPP data indicated that drainable bases had substantial positive impact on pavement performance in the wet-freeze (WF) region, and drainable bases did not add benefits in the other three climatic regions (wet-no-freeze (WNF), dry-freeze (DF), and dry-no-freeze (DNF)).
  
  
• Comparable test sections located in the four climatic regions did not perform the same with regard to IRI; rut depth; or alligator, longitudinal, or transverse cracking. Sections located in the WF region performed the worst compared with test sections located in the other three climatic regions.
  
  
• The impact on pavement performance in the WF region could be lessened by increasing the thickness of the AC and/or improving drainage.
  
  
• Thin overlay treatment had little to no impact on pavement performance in terms of alligator, longitudinal, and transverse cracking. Reflective cracking appeared in a short time period after the application of thin overlay.
  
  
• The condition and distress of the pavement sections before treatment affected the treatment longevity. The worse the pavement condition and distress was before treatment, the shorter the expected service period of the treatment was.
  
  
• The LTPP Seasonal Monitoring Program (SMP) deflection data did not support the AI temperature correction procedure.
  
  
• A new and innovative algorithm was developed to adjust the measured deflection data to a standard temperature of 70 ºF (21 ºC). The new algorithm applied to all deflection sensors and in all climatic regions.
  
  
• No consistent trends in the pavement deflection were observed over time. Hence, inclusion of deflection data in the algorithm of the dual pavement condition rating systems was not appropriate.
  
  
• Deflection data measured using the FWD could be used neither as an indicator of future pavement condition or distress nor to develop threshold values for the analysis of the pavement RFP and RSP.
  
  
• The methodologies used in the analyses of the LTPP data were also successfully applied to the State data. These methodologies were computerized using MATLAB® computer programs and Microsoft® Excel spreadsheets for formatting and organizing the results.
  
  
• The major difference between the LTPP and the State data was that, for each pavement condition and distress type, the LTTP database contained one data point per test section for every data collection cycle. In contrast, for a given pavement project, the State databases contained as many data points as the number of 0.1-mi (0.16-km)-long pavement segments along the project.
  
  
• The LTPP-measured IRI and distress data were similar and representative of the State-measured IRI and distress data in terms of magnitude and variability.
  
  
• The treatment benefits calculated using the LTPP data were parallel to the benefits of similar treatments calculated using the State data.
  
  
• The treatment benefits calculated using the LTPP data could be used as benchmarks for State transportation departments to check the performance of their pavement treatments and to assist them in conducting lifecycle cost analyses.

Based on the results of these analyses and the conclusions, the research team makes the following recommendations:

• Adopt the new dual pavement condition rating systems based on ranges in RFP and RSP as national standard measures to classify pavement condition and performance in a way that supports the selection of cost-effective pavement treatment strategy.
  
  
• Embrace the pavement performance prediction methodologies as national standard methodologies to unify and standardize the assessment of pavement performance and pavement treatment effectiveness.
  
  
• Conduct further research studies to incorporate the new dual pavement condition rating systems and the pavement performance prediction methodologies in the lifecycle cost analyses to optimize short- and long-term pavement treatment strategies.
  
  
• Conduct studies to determine the factors causing variability in the measured pavement condition and distress data and develop procedures to minimize their effects.
  
  
• Measure, at minimum, three sets of pavement condition and distress data over time before treatment.
  
  
• Initiate studies to establish automated quality control and assurance procedures for data collection and storage that minimize the impact of subjective factors. These studies may include the newly developed self-powered wireless macro-sensors that can be embedded in pavement and transportation infrastructures)
  
  
• Fund a study to determine the most efficient data collection frequency based on treatment types (such as thin and thick overlay, chip seal, and so forth) and their expected service periods.