U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

 
TECHBRIEF
This techbrief is an archived publication and may contain dated technical, contact, and link information
Back to Publication List        
Publication Number:  FHWA-HRT-16-046    Date:  August 2016
Publication Number: FHWA-HRT-16-046
Date: August 2016

 

Long-Term Pavement Performance Program—Pavement Performance Measures and Forecasting and The Effects of Maintenance and Rehabilitation Strategy on Treatment Effectiveness

PDF Version (766 KB)

PDF files can be viewed with the Acrobat® Reader®

 

FHWA Publication No.: FHWA-HRT-16-046

FHWA Contact: Yan “Jane” Jiang, HRDI-30, (202) 493-3149, jane.jiang@dot.gov

This document is a technical summary of the Federal Highway Administration Long-Term Pavement Performance Program report, Pavement Performance Measures and Forecasting and the Effects of Maintenance and Rehabilitation Strategy on Treatment Effectiveness (FHWA-HRT-16-047)

INTRODUCTION AND OBJECTIVE

This TechBrief presents the methodologies and procedures used by the research team in the analyses of the Long-Term Pavement Performance (LTPP) data to develop and implement pavement performance measures and to analyze treatment effectiveness. The LTPP study focused on using data from the various LTPP experiments to define pavement performance in a way that supports the selection of cost-effective pavement treatment strategy and to better estimate pavement treatment effectiveness and the role of pavement treatments in the pavement’s lifecycle. This TechBrief includes a description and examples of the dual pavement condition rating systems, LTPP data analyses results, and application of the analyses to datasets from three State transportation departments.

DUAL PAVEMENT CONDITION RATING SYSTEMS

To address the objectives of this study, comprehensive dual pavement condition rating systems were developed based on pavement conditions and rates of deterioration. An accurate pavement condition rating system best represents pavement behavior when it is based on both current and future pavement conditions. The main benefit of including the estimation of future conditions is the ability of pavement managers to plan, budget, and create cost-effective long-term pavement treatment strategies to preserve the pavement network. Pavement condition ratings based on current conditions alone only allow decisions to be made for the given data collection cycle and do not support lifecycle cost analyses (LCCA).

Balanced and comprehensive dual pavement condition rating systems were developed based on two types of pavement conditions: functional and structural. The functional rating is based on ride quality (International Roughness Index (IRI)) and safety (rut depth) and is expressed by the remaining functional period (RFP). (See the section entitled Definitions of RFP and RSP.) The structural rating is based on cracking and rut depth or faulting and is expressed by the remaining structural period (RSP). The RFP and RSP could be expanded by the road authority to include other pavement conditions and distress types such as skid resistance for the RFP, block cracking (for the RSP of flexible pavements), or corner breaks (for the RSP of rigid pavements). In this study, two pavement condition measures (IRI and rut depth) were used to calculate the RFP and four pavement distress types (alligator, longitudinal, and transverse cracking and rut depth or faulting) to calculate the RSP.

Definitions of RFP and RSP

The RFP is defined as the shortest time period measured in years from the time of the last data collection cycle to the time when a functional condition (e.g., IRI, rut depth, or other) reaches its corresponding prespecified threshold value. For a given pavement section and when supported by the available data, two or more RFP values can be calculated: one based on IRI, one on rut depth, one on skid resistance, and so forth. The shortest of the RFP values is assigned to the pavement section in question to flag the section for potential treatment actions. It should be noted that the measured pavement condition and distress data must be retained in the database and used to facilitate the selection of treatment types. The RFP is depicted in figure 1.

Figure 1. Graph. RFP condition states (CSs) for three- and five-level scales. This figure consists of a graph that displays IRI as a function of elapsed time. The y-axis is labeled “IRI (inch/mi)” and the x-axis is labeled “Elapsed time (years).” A legend below the graph indicates five different conditions named “Very good,” “Good,” “Fair,” “Poor,” and “Very poor” designated by patterned boxes. These are diagonal crisscross line pattern, diagonal line pattern, horizontal dashed line pattern, vertical dashed line pattern, and horizontal solid line pattern. Data is plotted as a solid line where for approximate elapsed time (x-axis) values 0, 5, 10, 15, and 20 years the IRI (y-axis) values are approximately 50, 76, 92, 125, and 171 inch/mi, respectively. A horizontal dashed line is located at approximate IRI value 171 inch/mi. Two solid lines, one horizontal and the other vertical, intersect each other at approximate elapsed time (x-axis) value 16 years and approximate IRI (y-axis) value 130 inch/mi. Good, fair, and poor are indicated for approximate ranges of elapsed time (x-axis) values 0 to 12, 12 to 16, and 16 to 20 years at approximate IRI (y-axis) value 200 inch/mi. Very good, good, fair, poor, and very poor are indicated for approximate ranges of elapsed time (x-axis) values 0 to 7, 7 to 12, 12 to 16, 16 to 18, and 18 to 20 years at approximate IRI (y-axis) value 220 inch/mi. Vertical lines extend from the dividing lines between each condition to the plotted data. Good, fair, and poor are indicated for approximate ranges of IRI (y-axis) values 50 to 105, 105 to 130, and 130 to 171 for approximate elapsed time (x-axis) value 23 years. Very good, good, fair, poor, and very poor are indicated for approximate ranges of IRI (y-axis) values 50 to 70, 70 to 105, 105 to 130, 130 to 150, and 150 to 171 for approximate elapsed time (x-axis) value of 25 years. Horizontal lines extend from the dividing lines between each condition to the plotted data.
1 inch/mi = 0.01588 m/km.

Figure 1. Graph. RFP condition states (CSs) for three- and five-level scales.

The RSP is defined as the shortest time period measured in years from the time of the last data collection to the time when a structural distress reaches its corresponding prespecified threshold value. For a given pavement section and when supported by the available data, two or more RSP values should be calculated: one for transverse, longitudinal, alligator, edge, and block cracking and one for either rut depth (for flexible pavements) or faulting (for rigid pavements). The shortest of the RSP values is assigned to the pavement section in question to flag the section for potential treatment actions. Once again, the measured distress data must be retained in the database to facilitate the selection of treatment types. The RSP is depicted in figure 2.

Figure 2. Graph. RSP CSs for three- and five-level scales. This figure consists of a graph that displays alligator cracking as a function of elapsed time. The y-axis is labeled “Alligator cracking (ft-squared/mile)” and the x-axis is labeled “Elapsed time (years).” The same conditions provided in the legend of Figure 1 are present in the graph. Data is plotted as a solid line where for approximate elapsed time (x-axis) values 0, 5, 10, 15, 20, 25, and 30 years the alligator cracking (y-axis) values are approximately 0; 0; 250; 2,000; 3,100; 3,100; and 3,100 ft2 -squared/mi, respectively. A horizontal dashed line is located at approximate alligator cracking value 3,100 ft2/mi. Good, fair, and poor are indicated for approximate ranges of elapsed time (x-axis) values 0 to 15, 15 to 19, and 19 to 23 at approximate alligator cracking (y-axis) value 4,000 ft2/mi. Very good, good, fair, poor, and very poor are indicated for approximate ranges of elapsed time (x-axis) values 0 to 10, 10 to 15, 15 to 19, 19 to 21, and 21 to 23 at approximate alligator cracking (y-axis) value 4,500 ft-squared/mile. Vertical lines extend from the dividing lines between each condition to the plotted data. Good, fair, and poor are indicated for approximate ranges of alligator cracking (y-axis) values 0 to 2,000, 2,000 to 3,000, and 3,000 to 3,100 at approximate elapsed time (x-axis) value 24 years. Very good, good, fair, poor, and very poor are indicated for approximate ranges of alligator cracking (y-axis) value 0 to 250, 250 to 2,000, 2,000 to 3,000, 3,000 to 3,050, and 3,050 to 3,100 for approximate elapsed time (x-axis) value of 26 years. Horizontal lines extend from the dividing lines between each condition to the plotted data.
1 ft2/mi = 0.05806 m2/km.

Figure 2. Graph. RSP CSs for three- and five-level scales.

Rating Scale

Two rating scales could be used for each of the RFP and RSP. One scale is based on three levels, and the other scale is based on five levels. For each scale, the pavement condition states (CSs) can be expressed either in descriptive terms or in numeric terms that represent the RFP and the RSP. For example, for the three-level scale, the CS of a given pavement section may be expressed in one of the following categories as illustrated in this list and in table 1:

The dual rating systems could be used to select treatment categories at the network level. For example, preservation treatments should generally be applied to pavement sections with fair or better CSs. Heavy preservation treatment, or more likely rehabilitation, should generally be applied to pavement sections having poor RSP CSs. The treatment selection should be verified at the project level.

The five-category rating scale included in table 2 provides finer and additional data analyses.

Table 1. Pavement condition rating based on three CSs.
CS RFP Range (Year) RSP Range (Year) Cost per Lane-Mile (Lane-Km) $ times 105
Code Pattern or Color Descriptor
RFP RSP
1 Horizontal solid line or red Poor < 4 < 4 5 (3.1) 10 (6.2)
2 Horizontal dashed line or yellow Fair 4 to < 8 4 to < 8 2.2 (1.4) 5.5 (3.4)
3 Diagonal crisscross line or green Good > 8 > 8 0.3 (0.2) 1.0 (0.6)

 

Table 2. Pavement condition rating based on five CSs.
CS RFP Range (Year) RSP Range (Year) Cost per Lane-Mile (Lane-Km) $ times 105
Code Pattern or Color Descriptor
RFP RSP
1a Horizontal solid line or red Very poor < 2 < 2 5 (3.1) 10 (6.2)
1b Vertical dashed line or pink Poor 2 to < 4 2 to < 4 3 (1.9) 7 (4.4)
2 Horizontal dashed line or yellow Fair 4 to < 8 4 to < 8 1.5 (0.9) 4 (2.5)
3a Diagonal line or light green Good 8 to < 13 8 to < 13 0.5 (0.3) 1.5 (0.9)
3b Diagonal crisscross line or green Very good > 13 > 13 0.1 (0.1) 0.5 (0.3)

 

The three- and five-category rating scales are also depicted in figure 1 and figure 2 for RFP and RSP, respectively. Note that for both pavement rating scales, the main reason for using the same ranges in years for RFP and RSP for each rating category is for ease of communication. The CS numbers could be used for programming purposes, the color code for mapping, the descriptive (poor, fair, and good) for communication with the public and legislators, and the RFP and RSP scales for planning and for LCCA. The latter could be better accomplished if the road agencies used their own cost data to assign an average pavement preservation cost per lane mile for each RFP range and for each RSP range (similar to the conceptual cost included in table 1 and
table 2). In general, the cost of preserving pavement sections based on short RFP is much lower than the cost of preservation based on short RSP.

The main advantage of the RFP and RSP is that the value of each should decrease 1 year for every calendar year. Although the RFP and RSP are calculated using nonlinear functions to model the pavement condition and distress as a function of time (i.e., exponential function for IRI, power function for rut depth, and logistic (S-shaped curve) for cracking), the resulting RFP and RSP are linear functions of time. Each should decrease by 1 year for every calendar year. Nevertheless, the RFP and RSP can be considered forecasting tools that can be used to establish cost-effective strategies that address planning, budgeting, and contracting pavement preservation activities at the proper time.

ONE RECORD CONDITION STATE ESTIMATE (ORCSE)

As stated earlier, for a given pavement section, the RFP and/or the RSP values can be considered the forecasting tool to flag pavement sections that require attention. The two values are based on the measured time-dependent pavement condition and distress data. The data are fitted to the proper mathematical function, which is used to forecast the time at which the pavement section in question will reach the threshold value. However, for many pavement sections, the pavement management database does not contain the required minimum three data points to model the pavement condition and distress as a function of time. This scenario is especially true for newly constructed or rehabilitated pavement sections. To address the problem, the LTPP database was used to develop a forecasting system called ORCSE based on a single measured data point.(1) In the development, the LTPP data from Specific Pavement Study (SPS)-1 were analyzed, and the RFP and RSP for each test section were calculated. The calculated RFP and RSP values and one individual measured data point were modeled using the Epanechnikov Kernel probability density function model. The model provided the closest fit to the observed data while generating secondary benefits of the identification of possible sub-probability groups or divergent behaviors within the larger SPS-1 generalized sample. Table 3 lists CS probabilities based on ranges of measured IRI for SPS-1 test sections before treatment (BT). For older SPS-1 sections, the ORCSE results were within 5 percent of those using three or more data points. Therefore, the ORCSE is a reliable method for estimating RFP and RSP while time series data are under collection.

Table 3. ORCSE model table example from LTPP SPS-1 BT evaluation.
IRI Range (inches/mi) Probability of a CS or RFP Level for Selected IRI Ranges (percent)
CS 1 or RFP
< 2 years
CS 2 or RFP
2 to < 4 years
CS 3 or RFP
4 to < 8 years
CS 4 or RFP
8 to < 13 years
CS 5 or RFP
> 13 years
.25–.50 0 0 0 3 97
.50–.75 0 0 3 14 83
.75–1.00 0 0 4 19 77
1.00–1.25 0 0 9 39 52
1.25–1.50 0 3 19 66 11
1.50–1.75 0 12 49 37 1
1.75–2.00 3 28 59 10 0
2.00–2.25 13 61 22 3 0
2.25–2.50 55 36 6 3 0
2.50–2.70 88 8 1 2 0
1 inch/mi = 0.02 m/km.

 

LTPP DATA ANALYSES

Pavement condition and distress data as well as inventory and treatment data were obtained from the LTPP Standard Data Release 28. The available data for test sections from SPS-1 through SPS-7, General Pavement Studies (GPS)-6, GPS-7, and GPS-9 experiments were modeled both before and after the application of treatments. For the total of 1,555 test sections (see table 4), sufficient data were available and were analyzed before treatment (BT) and after treatment (AT) in 1,182 instances; sufficient data were available and were analyzed BT in only 1,544 instances, and sufficient data were available and were analyzed AT in only 2,691 instances. For each treatment application, the IRI data were modeled using an exponential function, rut depth data using a power function, and cracking data using a logistic (S-shaped) function.

Table 4. Summary of treatments applied to SPS-1 through SPS-7 and GPS-6, -7, and -9.
Number of Test Sections Number of Treated Sections Number of Treatment Applications Pavement Distress/ Condition Number of Treatment Applications Analyzed
BT and AT BT Only AT Only
1,555 1,301 2,674 Cracking 278 463 925
IRI 468 558 911
Rut depth 394 453 747
Faulting 42 70 108
Total 1,182 1,544 2,691

 

The conclusions and recommendations drawn from these analyses are presented in the last section of this TechBrief. For each treatment application, the RFP and RSP values were calculated based on the threshold values. The following should be noted:

Application of Methodology to State Data

Similar analyses to those performed on the LTPP data were performed on pavement condition and distress data measured by three State transportation departments (Colorado, Louisiana, and Washington) along various pavement projects of their respective pavement networks. The data collected and stored by the State transportation departments are similar to the LTPP data with few exceptions. These data include the pavement segment length, definition of some pavement distresses and units of measurement, pavement structure or cross-section, and other inventory data. The three State transportation departments collect and store data for each 0.1-mi (0.16‑km)‑long pavement segment along their respective networks. For each pavement project and for each data collection cycle, the number of data points available in the database is the same as the number of consecutive 0.1-mi (0.16-km)-long pavement segments along the project. The analyzed pavement projects were subjected to various treatments, including single chip seal, thin and thick HMA overlays, and thin and thick mill-and-fill treatments.

Nevertheless, results of the analyses indicate the following:

The important implication of the findings is that the analysis methodologies described in this study can be used by the various State transportation departments. Furthermore, the RFP and RSP concept and the CSs apply equally to the State and the LTPP data.

Finally, the research team strongly recommends that treatment benefit benchmarks obtained from the analyses of the LTPP data be adopted and used by the American Association of State Highway and Transportation Officials (AASHTO) and the State transportation departments. This would standardize the designations of the treatment benefits prior to the LCCA.

Deflection Data Analyses

The RFP and RSP algorithms are primarily based on the measured time-series pavement surface condition and distress data and their corresponding threshold values. Hence, the distress (such as cracking) must be visible from the pavement surface to be counted. During the development of the RFP and RSP concepts, it was envisioned that the pavement deflection data could be used to indicate impending surface distress and be a part of the RSP algorithm. Such algorithms would empower State transportation departments to take corrective actions prior to the manifestation of surface defects. To incorporate deflection into the RSP algorithm, a deflection threshold value must be developed for each pavement section. To investigate the potential for the development of deflection threshold values, the measured falling weight deflectometer (FWD) deflection data of various LTPP seasonal monitoring program (SMP) and other test sections were analyzed.

The main purpose of the analyses of the deflection data is to identify relationships, if any, between the measured pavement deflection and the measured pavement condition data to determine whether the deflection data can be used to estimate the optimum time for pavement preservation. To conduct the analyses, the measured pavement surface deflections of some flexible pavement test sections were examined. Figure 3 shows the impacts of the measured pavement surface temperature on the pavement surface deflection measured at deflection sensors 1, 2, 4, and 7 (0, 8, 18, and 60 inches (0, 203, 457, and 1,524 mm) from the load) of the FWD. As was expected, the data indicate that the effects of temperature decrease with increasing distance from the center of the FWD load plate. The measured deflections were then adjusted to the standard temperature of 70 ºF (21 ºC) using existing temperature adjustment models, such as the Asphalt Institute (AI) and BELLS models.(2,3) None of the SMP site results agreed with the measured deflection data at the standard temperature of 70 ºF (21 ºC). Therefore, a global temperature adjustment procedure was developed using the LTPP measured deflection data along with various SMP test sections. The new procedure is applicable to all deflection sensors and in all climatic regions.

Figure 3. Graph. Peak measured pavement deflection at sensors 1, 2, 4, and 7 versus pavement surface temperature for LTPP test section 010101, F3. This figure consists of a graph that displays average peak deflection as a function of average pavement surface temperature. The y-axis is labeled “Average peak deflection (micron)” and the x-axis is labeled “Average pavement surface temperature (°C).” Circles, triangles, diamonds, and squares represent Sensor 1, Sensor 2, Sensor 4, and Sensor 7, respectively. A straight line is fit through each set. For average pavement surface temperature (x-axis) values 10, 20, 30, 40, and 50 degrees Celsius; these data sets have the following approximate average peak deflection (y-axis) values; set one circles (205, 245, 280, 320, and 355 microns); set two triangles (190, 205, 225, 250, and 275 microns); set three diamonds (145, 150, 155, 160, and 165 microns); set four squares (50, 50, 50, 50, and 50 microns).
oC = (5/9)(oF–32o).
1 mil = 25.4 micron.

Figure 3. Graph. Peak measured pavement deflection at sensors 1, 2, 4, and 7 versus pavement surface temperature for LTPP test section 010101, F3.

The AI method and the new procedure were then used to adjust the measured deflection data over time to the standard temperature of 70 ºF (21 ºC). Figure 4 depicts the percent error between the temperature-adjusted deflection data and the measured deflection data at 70 ºF (21 ºC) at the same SMP test site. It can be seen that the new procedure has much smaller errors than the AI procedure.

After adjusting the measured deflection data to 70 ºF (21 ºC), the adjusted data of one test section were then plotted against the pavement distress data that were measured at the same SMP test section. It was found that the measured and temperature-adjusted deflection did not correlate with the measured cracking data. In fact, the deflection data did not show any consistent pattern against time. Therefore, the data cannot be used to develop a deflection threshold value and cannot be included in the RSP algorithm. However, the deflection data can be used to estimate the moduli of the pavement layers and to design pavement treatments.

Figure 4. Graph. Percent errors of the temperature-adjusted deflection data using the new procedure and the AI procedure. This figure consists of a bar graph that displays error in temperature adjusted deflection as a function of deflection sensor number. The y-axis is labeled “Error in temperature adjusted deflection (percent)” and the x-axis is labeled “Deflection sensor number.” Horizontal lines, square pattern, dots, and checkered pattern represent Global_11°C, AI_11°C, Global_30°C, and AI_30°C, respectively. For deflection sensor number (x-axis) values 1, 2, 3, 4, 5, 6, and 7 the error in temperature adjusted deflection (percent) (y-axis) values read: set one Global_11°C are approximately 3, 0.3, 0.5, 1.5, 1.8, 7.8, and 2.2 percent, respectively; set two AI_11°C are approximately 10, 10, 12, 15, 18.3, 17.5, and 19.5 percent; set three Global_30°C are approximately 0.3, 4.5, 5, 4.5, 3.5, 0.2, and 0.1percent; set four AI_30°C are approximately 2.2, 1.9, 3.5, 6, 7.5, 10, 13.5 percent.
1 ºF = [(5/9) x (ºF-32)] ºC.

Figure 4. Graph. Percent errors of the temperature-adjusted deflection data using the new procedure and the AI procedure.

CONCLUSIONS

Based on the results of the analyses, conclusions and recommendations were drawn that cover a range of study-related topics.

Pavement Performance Measures

The researchers drew the following conclusions related to pavement performance measures:

Flexible Pavements

Concerning flexible pavements, the researchers drew the following conclusions:

ORCSE Method

The researchers concluded the following:

Rigid and Composite Pavements

Concerning rigid and composite pavements, the researchers concluded the following:

Deflection

In the area of deflection, the researchers drew the following conclusions:

State Data

The State data indicate the following:

RECOMMENDATIONS

Performance Measures

Based on the results of the LTPP and State data analyses and the conclusions listed above, the following actions are strongly recommended:

Flexible Pavements

Concerning flexible pavements, the following actions are strongly recommended:

ORCSE Method

The researchers recommend that the ORCSE method be expanded to other pavement conditions and distresses and applied to a wider range of LTPP and State transportation department data to further verify its successful prediction of pavement CSs.

Deflection

In the area of deflection, the researchers recommend the following:

State Data

Concerning the State data, the following are recommended:

Future Studies

Future studies should address the following recommendations:

REFERENCES

  1. FHWA. (2014). LTPP Pavement Performance Database Release Notes, Standard Data Release 28, Federal Highway Administration, Washington, DC, obtained from: https://infopave.fhwa.dot.gov/Data/StandardDataRelease, last accessed August 1, 2015.

  2. Asphalt Institute. (1983). Asphalt Overlays for Highway and Street Rehabilitation, Manual Series No. 17 (MS-17), Lexington, KY.

  3. Lukanen, E.O., Stubstad, R., and Briggs, R. (2000). Temperature Predictions and Adjustment Factors for Asphalt Pavement, Report No. FHWA-RD-98-085, Federal Highway Administration, Washington, DC.

Researchers—This study was performed by Michigan State University Department of Civil & Environmental Engineering in East Lansing, MI, and NTH Consultants, Ltd.,
in Lansing, MI.

Distribution—This TechBrief is being distributed according to a standard distribution. Direct distribution is being made to the Divisions and Resource Center.

Availability—This TechBrief may be obtained from FHWA Product Distribution Center by e-mail to report.center@dot.gov, fax to (814) 239-2156, phone to (814) 239-1160, or online at https://www.fhwa.dot.gov/research/.

Key Words—Pavement performance; pavement rehabilitation; LTPP data; data analysis; good, fair, and poor; remaining service life; remaining functional period; remaining structural period.

Notice—This document is disseminated under the sponsorship of the U.S. Department of Transportation in the interest of information exchange. The U.S. Government assumes no liability for the use of the information contained in this document. The U.S. Government does not endorse products or manufacturers. Trademarks or manufacturers’ names appear in this report only because they are considered essential to the objective of the document.

Quality Assurance Statement—The Federal Highway Administration (FHWA) provides high-quality information to serve the Government, industry, and public in a manner that promotes public understanding. Standards and policies are used to ensure and to maximize the quality, objectivity, utility, and integrity of its information. Periodically, FHWA reviews quality issues and adjusts its programs and processes to ensure continuous quality improvement.

 

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101