Estimating Cumulative Traffic Loads, Volume II:
Traffic Data Assessment and Axle Load Projection for The Sites With Acceptable Axle Weight Data, Final Report for Phase 2

CHAPTER 2. PROCEDURES FOR TRAFFIC DATA ASSESSMENT AND PROJECTION

This chapter describes the procedures used to carry out the assessment of the quantity and quality of traffic data, and to estimate traffic loads for all years the LTPP sites were in service. The process of estimating traffic loads is referred to as traffic projection.

The assessment of the data quality and quantity was carried out for all LTPP sites. The projection of truck volumes was carried out for all sites with appropriate monitoring or historical truck volumes, and the projection of axle load spectra was done for all LTPP sites that had appropriate site-specific monitoring axle load data in the Level E release of the IMS database (first quarter of 2000).

The assessment of traffic data and the development of traffic projections have been carried out in two phases over the course of 30 months; they involved eight main activities (see figure 1).

Of the eight traffic data assessment and projection activities, three were carried out in Phase 1 and five in Phase 2. In this chapter, the activities are described in the order they are listed in figure 1 as steps 1 to 8. To enable the reader to follow the traffic data assessment and projection process better, a general outline of the eight activities is provided first, followed by a more detailed description of the activities.

The main traffic data assessment and projection activities carried out in Phase 1, and described in Phase 1 report were:^[1]

1. Preliminary assessment of LTPP traffic data - The objective of this activity was to obtain a basic understanding of the overall quantity and quality of traffic data. This understanding was necessary for the development of the traffic projection procedure.
2. Development of LTPP traffic projection procedure-The projection procedure developed in Phase 1 was designed to utilize fully all available historical and monitoring traffic data and to incorporate additional traffic information obtained from local agencies.
3. Validation of LTPP traffic projection procedure using case studies -The projection procedure was used to estimate axle loads for all in-service years of 12 LTPP sites with different amounts of historical and monitoring data. The objective was to validate the projection procedure using realistic examples.

The main traffic data assessment and traffic projection activities carried out in Phase 2, and described in here, were:

4. Development of LTPP Traffic Feedback and Resolution Package-The purpose of the LTPP Traffic Feedback and Resolution Package was to summarize and present historical and monitoring traffic data and the results of traffic data projections in a structured and user-friendly manner. The package, developed using a systematic procedure applied consistently to all LTPP sites, was the main means of communication and interaction with the participating agencies. The package contains traffic data summaries for the LTPP sites, easy-to-follow results of traffic projections, background information, and guidelines on how to assess traffic data and review traffic projections. During the development of the package, input and comments were received from the representatives of RCOs, members of the Expert Task Group (ETG) on Traffic Data Collection and Analysis, and others. To further evaluate suitability and ease of use, and to provide an opportunity for the RCOs to become familiar with the content and the purpose of the package, pilot studies involving four participating agencies were also carried out.

5. Preparation of the LTPP Traffic Feedback and Resolution Packages for all participating agencies -Altogether, 62 LTPP Traffic Feedback and Resolution Packages were produced (one for each participating agency). Each agency package addresses all LTPP traffic sites that belong to the agency, and each LTPP site was processed in terms of: (a) the assessment of the amount and quality of historical and monitoring traffic data, (b) the development of traffic projections, provided that appropriate traffic data were available, and (c) the assignment of confidence codes to the initial traffic projections.
   
6. Review of LTPP Traffic Feedback and Resolution Packages by RCOs-The responsibilities of RCOs include initial traffic data quality checks, as well as processing and storage of traffic data supplied by the participating agencies. Because of a long-standing involvement of the RCOs with traffic data issues on the local level and the knowledge of local traffic issues, the packages were sent to the RCOs for review prior to being sent to the participating agencies.

7. Review of Traffic Feedback and Resolution Packages by participating agencies-After the concerns of the RCOs were addressed, the updated packages were sent to the participating agencies for review. The involvement of the participating agencies has been crucial; many data problems cannot be resolved without local involvement.

8. Implementation of review comments received from participating agencies-To facilitate communication with the participating agencies on an individual section basis, a standardized questionnaire (Feedback and Data Resolution Sheet) for each LTPP site was developed, and the participating agencies were asked to complete it. Based on the responses we received from the participating agencies, the initial traffic projections were revised and the initial projection confidence codes were changed to reviewed projection confidence codes.

Step 1-Preliminary Assessment of LTPP Traffic Data

The Phase 1 report contains the results of the evaluation of LTPP traffic characteristics.^[1] The results include examples of spatial and temporal variation in truck volumes; comparison of trends in historical and monitoring annual truck volumes and annual ESALs; the evaluation of the potential of Class 9 vehicles for prediction of axle loads; and the evaluation of axle load distribution characteristics: The Phase 1 report contains a review of several previous studies that used LTPP traffic data to: characterize seasonal variation in truck volumes;^[8] evaluate the consequences of alternative sampling plans on the estimated annual traffic loads;^[9] and assess the potential of LTPP traffic data to develop models for predicting axle load distributions.^[10] The results of an exploratory analysis of trends in traffic data were presented in an interim report that preceded the Phase 1 report. ^[11]

The review of LTPP traffic data carried out during Phase 1 identified the following major issues, concerns, and recommendations regarding quality and quantity of traffic data and the projection of traffic loads.

• Trends in historical and monitoring data-The relationship between annual historical and monitoring ESALs, and between monitoring ESALs obtained for different years, investigated for all sites in the North Atlantic and North Central Regions, showed considerable variation. The differences between historical and monitoring ESALs, and between ESALs obtained for consecutive monitoring years, frequently exceeded 100 percent. Because truck volumes and axle loads on interstates and other major highways are expected to show relatively steady growth over the years, the observed differences indicated important data concerns requiring resolution.
• Involvement of participating agencies-Considering the available historical and monitoring data and the challenges involved in carrying out traffic projection, the involvement of participating agencies in the traffic projection process was identified as being essential for knowledge-based traffic data assessment and projection of traffic loads.
• Unavailability of monitoring data-Typically, only 3 or 4 years of monitoring axle load data are available for LTPP sites. Many sites do not have any axle load (WIM) or truck class distribution (AVC) data.
• Calibration of AVC and WIM equipment-AVC and WIM scale calibration concerns have not been studied in sufficient detail. For example, the CTDB does not contain any data on quadruple axles even though such axles regularly occur in several participating agencies, including Michigan and Ontario. For many sites, axle load spectra exhibited unexpectedly large year-to-year variations.
• Quality assurance -The QA techniques employed previously were characterized as an automated review to detect common equipment problems, rather than a comprehensive and detailed QA process. To ensure the integrity of the data used for traffic projections and for any subsequent analysis, the Phase 1 study recommended that a comprehensive QA process of underlying traffic data be carried out.
• Development of PLG -The data QA process would greatly benefit from the development of a knowledge base documenting typical values and ranges of traffic variables, particularly truck class and axle load distributions.

The knowledge gained during the preliminary assessment of the amount and quality of available historical and monitoring data was instrumental in the development of the LTPP traffic projection procedure.

Step 2-Development of LTPP Traffic Projection Procedure

This section outlines the LTPP traffic projection procedure. The detailed description of the procedure, including numerical examples in the form of case studies, is contained in the Phase 1 report.^[1] To satisfy the following design requirements, the projection procedure was developed to ensure:

• The ability to estimate annual axle loads for the majority of LTPP sites and for all years the sites were in service.
• Compatibility with the available historical and monitoring traffic data and their full utilization.
• Transparency (to be understandable to the users of projected traffic data) and modularity (for ease of future enhancements of the projection procedure, and for ease of updating the projections if more data become available).

The procedure for projecting axle load spectra for the individual LTPP sites includes the following activities:

• All available annual historical and monitoring data were used to establish a model for estimating annual truck volumes for all years the site was in service.
• The model for estimating annual truck volumes was used to obtain annual projection factors.
• A base annual axle load spectrum, representing a typical axle load spectrum for the site, was established.
• The projected annual axle load spectra for all in-service years were obtained by multiplying the base annual spectrum by the annual projection factors.

Mathematically, the LTPP traffic load projection procedure is based on the following formula:

Annual Axle Load Spectrum_y = Base Annual Spectrum * Annual Projections Factor _y

Where:

The above formula means that once the base annual spectrum was established, it was scaled using the annual projection factor corresponding to the annual truck volume for the given year. Consequently, it was assumed that the truck class distribution remained constant over the years and that only the amount of trucks changed.

Specific techniques, described in Phase 1 report, were developed to obtain base annual spectra for the five projection categories.^[1] The five projection categories, classified by the amount of available monitoring data, are defined in chapter 1 of this report. Annual axle load spectra were projected also for the years with monitoring (measured) annual axle load spectra. For these years, both the monitoring and the projected axle load spectra will be available in the IMS database.

To obtain cumulative axle loads, annual axle spectra could be summed as shown in equation 2.

Where:

Projections for Sites in Category 1 and 2

Category 1 and 2 sites have monitoring annual axle load spectra in the IMS database; these spectra were used to obtain the base annual spectrum. Traffic load projections carried out in Phase 2 were done for Category 1 and 2 sites only, and were in terms of annual axle load spectra. Typically, the base annual spectrum was an average of all acceptable annual spectra for the given site. Techniques used to establish base annual spectra for Category 1 and 2 sites are summarized as part of step 5 (figure 1) in the section titled "Development of Base Annual Spectra."

The Phase 1 report also contains the description of procedures developed for estimating base annual spectra for Category 3 and 4 sites, as well as numerical examples of traffic projections for these sites.^[1] Even though these procedures were not used in this report, they are outlined below because they are important for the understanding of the subsequent recommendation to develop the LTPP PLG (discussed in chapter 5 of this report).

Projections for Sites in Category 3

Category 3 includes sites with site-specific monitoring truck class distribution data but without acceptable site-specific monitoring axle load distribution data. Because these sites do not have annual axle load spectra in the database, the base annual spectrum must be estimated. According to the Phase 1 report, it is proposed to accomplish this by combining the following two inputs:^[1]

• Site-specific monitoring truck class distribution.
• Surrogate axle load spectra for individual truck types.

Site-Specific Monitoring Truck Class Distribution

If there are several years for which the site-specific monitoring truck class distributions are available, a typical truck volume distribution (called the base truck distribution) is established by plotting and assessing all annual truck class distributions using similar techniques as those used to obtain the base annual spectrum. The objective is to obtain the truck class distribution that best represents the given site.

Surrogate Axle Load Spectra for Individual Truck Types

The sources of surrogate axle load spectra include site-related data, regional data, and generic or typical data.

Projections for Sites in Category 4

Category 4 includes sites with annual truck volume data but without site-specific truck class and axle load distributions. Because these sites do not have annual axle load distribution in the database, the base annual axle load spectrum must be constructed. According to the Phase 1 report, it is proposed to accomplish this by combining the following three inputs:^[1]

• Site-specific total truck volume.
• Surrogate monitoring truck class distribution.
• Surrogate axle load spectra for individual truck types.

The sources of surrogate truck class and axle load distribution data include site-related data, regional data, and generic or typical data.

Step 3-Validation of the LTPP Projection Procedure Using Case Studies

The methodology for projecting axle load spectra was evaluated and demonstrated in Phase 1 using case studies for selected LTPP sites. Case studies are numerical example applications of the traffic projection procedure for actual LTPP sites. All together, 12 case studies were carried out (3 case studies in each projection category except Category 5). The following observations were based on the results of the case studies:^[1]

• The procedure for projecting traffic loads developed in the course of the Phase 1 study can be used to estimate axle load spectra for all LTPP sites.
• The involvement of participating agencies in the projection process is essential.
• The selection of truck class and axle load distributions required for Category 3 and 4 projections must be done judiciously and should be supported by a reference database summarizing characteristic truck class and axle load distribution data .
• The QA process would greatly benefit from developing a knowledge base or a catalog documenting typical or expected values and ranges of traffic variables, particularly axle load spectra for individual vehicle classes.

Step 4-Development of LTPP Traffic Feedback and Resolution Package

To meet the objectives of the Phase 2 study, it was necessary to work with all LTPP sites and with all participating agencies. The LTPP traffic database contains 890 unique traffic sites located in 62 agencies. Reflecting the objectives of the Phase 2 study, the objectives of the LTPP Traffic Feedback and Resolution Package were:

• To report back to the participating agencies the quantity and quality of the traffic data available for all LTPP traffic sites in their respective jurisdictions, using easy-to-understand graphic displays. Although the participating agencies had received traffic feedback reports in the past, the previous reports were limited in scope and did not contain all relevant traffic data for all historical and monitoring years, nor did they include long-term trends in traffic volumes and loads.
• To present traffic data in the format that would:
  • Facilitate the assessment of the quality and quantity of traffic data by the participating agencies, RCOs, and the project team.
  • Facilitate the understanding of the projection process, and the assessment of the initial traffic projections by the participating agencies and RCOs.

To efficiently assess traffic data and carry out traffic projections for all 890 LTPP sites, and to communicate the traffic projection results to participating agencies effectively, we developed a standardized package-the LTPP Traffic Feedback and Resolution Package (also referred to here as "the package"). The package combines graphic traffic data displays, information on how to interpret the data displays, and questionnaires addressing overall and site-specific traffic data issues. A separate package was prepared for each participating agency.

The rest of this section describes the content of the package and the involvement of the representatives of RCOs and others in its development and validation.

Content of the LTPP Traffic Feedback and Resolution Package

The package consisted of five items:

• Introductory letter.
• Outline of the LTPP traffic projection procedure.
• Initial overall feedback and resolution report.
• Initial site-specific feedback and resolution report.
• Site-specific reports for each LTPP site within the agency.

These five items are briefly described below. The package was quite bulky, particularly for agencies with many LTPP sites.

Introductory Letter

The introductory letter was prepared and signed by a representative of the RCO and was addressed to the State or Province LTPP contact engineer or other official. Typically, the letter outlined the purpose of the package and asked that a person within the agency who was familiar with the collection of traffic data for LTPP sites review the package and respond to the issues raised. Some letters also included additional information and comments regarding the traffic data availability for specific sites (e.g., if data submitted previously by the participating agencies were not yet included in the package).

Outline of the LTPP Traffic Projection Procedure

The outline of the LTPP traffic projection procedure was a 10-page report that explained the traffic projection procedure used to estimate traffic loads in straightforward language. The objective of the outline was to enable the reviewer to understand the relative importance of the different issues, potential discrepancies, and questions posed in the package. The outlined discussed:

• Objectives of the traffic projection procedure.
• Definition of key technical terms used in the outline.
• The need to estimate traffic loads in terms of axle load spectra. The outline also included a typical example of axle load spectra in graphic and tabular forms.
• Reasons for the involvement of participating agencies in the traffic estimation process.
• LTPP traffic projection procedure.

Initial Overall Feedback and Resolution Report

The initial overall feedback and resolution report was typically a 4- or 5-page report that contained a summary of traffic data assessment and traffic projection issues concerning more than one LTPP site. The report was called "initial" because it was concerned with the first traffic projections carried out for LTPP sites. The initial overall feedback and resolution report served two purposes:

• To summarize issues identified during the initial traffic data assessment and projection effort that required input from agency representatives.
• To seek additional traffic data and information from the agency representatives to improve traffic projections.

The issues and questions were grouped under the following headings:

• Overall review.
• Missing data.
• Location of sections.
• Traffic volumes.
• Vehicle classification-operation of AVC equipment.
• Axle weights-operation of WIM scales.

Overall review -The overall review included a table listing all LTPP sites for the agency and their corresponding LTPP experiment numbers and the initial projection confidence codes. The initial projection confidence codes were used to characterize the level of confidence associated with initial traffic projections. The assignment of the projection codes is described later ("Assignment of Initial Traffic Projection Codes" in step 5).

Missing data -This section provided a comprehensive listing of traffic data that were missing from the IMS database.

Location of sections -Particular attention was paid to the nearby sites located on the same highway, particularly if the sites were also located in the same direction of travel. The objective was to ascertain the existence of expected relationships between truck volumes and axle loads on related sites. Any potential discrepancies were brought to the attention of the agency representatives.

An example of the data assessment carried out under the heading of "Location of sections" is provided in table 3. As expected, nearby sites on the same highway in the same direction have similar truck volumes, truck percentages, and truck growth rates. The exception appears to be the two northbound sites on U.S. Route 93 in Arizona that have quite different truck percentages (15 versus 27) and quite different recent truck growth rates (1.0 versus 10.0 percent).

Traffic volumes -The objective of this part of the report was to emphasize the importance of trends in annual historical and monitoring truck volumes for the development of the traffic load projections. Sections with unexpected variation in annual truck volumes were identified and reviewers were asked to address truck volume discrepancies on a section-specific basis.

Vehicle classification-operation of AVC equipment-Typically, the main concern with vehicle classification was the number of vehicles that were not properly classified by the traffic monitoring equipment. These vehicles are identified in the IMS as Class 14 vehicles. Some agencies reported that the percentage of Class 14 vehicles in the total truck flow was, for the majority of sites and years, above 10 percent. Some agencies did not report any Class 14 vehicles. Questions posed to the agencies regarding vehicle classification were regarding: (a) distribution of Class 14 vehicles into "legitimate" vehicle classes; (b) reporting of vehicles that were not properly classified by AVC equipment; and (c) type of procedures used to ensure that the vehicles are properly classified.

Axle weights-operation of WIM scales-This segment of the overall report contained two parts.

The first part identified the LTPP sites with questionable axle load data and inquired about the procedures used to calibrate WIM scales and to review axle load data prior to their submission to RCOs. Questionable axle load distributions, for at least some sections and some monitoring years, were reported by all agencies that reported monitoring axle loads. Typically, the following questions were asked: What type of procedure is used to ensure that WIM scales are calibrated? Is WIM calibration done routinely? Do you have resources to review traffic data prior to their submission to LTPP? If so, what procedures do you use?

The second provided the results of analysis carried out to ascertain basic traffic loading patterns for all LTPP sites within the agency. The objective was to provide a summary of axle load characteristics and to identify sites that do not fit overall patterns. An example of such a summary is shown in tables 4 and 5.

The results in tables 4 and 5 indicate a similarity of traffic load characteristics with a few exceptions. For example, all TFs in table 4 are in the range of 0.9 to 1.3, except the TF of 0.6 for Mississippi site 1001. Consequently, the axle load data on this site need to be assessed to ascertain the reason for lower-than-expected axle loads.

Initial Site-Specific Feedback and Resolution Report

The purpose of this report was to:

• Describe the standardized displays of traffic data and the initial traffic projection results for individual LTPP sites.
• Provide guidelines for the assessment of traffic data and for the interpretation and evaluation of traffic projection results.

*Interstate; **1 kip = 4450 newtons

An example of the Initial Site-Specific Feedback and Resolution Report, prepared for the representatives of participating agencies assigned the task to review the initial traffic projections, is provided in figure 2. Specifically, figure 2 includes the entire site-specific feedback and resolution report prepared for the Mississippi Department of Transportation (DOT); the reports prepared for other agencies were similar.

The report in figure 2 refers to 26 site-specific reports (one for each of the 26 Mississippi LTPP sites that have traffic data in the IMS database). One of the 26 site-specific reports for Mississippi, report for site 285805 is presented in figures 3 through 10. To assist the reader in making the connection between the reports, figure 2 contains references to figures 3 through 10. The first data sheet of the site-specific report (figure 3) was on blue paper and is referred to as a "Blue Sheet."

Site-Specific Reports

The site-specific report for site 285805 (figures 3 through 10) consists of eight data sheets:

• Feedback and Data Resolution Sheet (figure 3).
• Site Map (figure 4).
• Annual Traffic Projection Sheet (figure 5).
• Projected AADT Truck Volumes (figure 6).
• Annual Vehicle Class Distribution (figure 7).
• Annual Load Spectra (figure 8).
• Average Annual Load Spectrum (figure 9).
• Projected Annual ESALs (figure 10).

The procedures used to develop site-specific reports are described in the Phase 1 report, and the content and purpose of the report are described in figure 2.^[1] Subsequent sections will provide a description of techniques used to develop the initial projections shown in figures 6 and 10, and to assign projection confidence codes.

The Blue Sheet in figure 3 is shown as having been completed by the Mississippi Department of Transportation (MDOT) to avoid the need to show separate examples of uncompleted and completed Blue Sheets. The completion of Blue Sheets by participating agencies will be discussed in step 7 (Review of LTPP Traffic Feedback and Resolution Packages by Participating Agencies).

Figure 4. Site map for site 285805.

Figure 5. Annual Traffic Projection Sheet for Site 285805.

Figure 6. Projected AADT volumes for site 285805.

Figure 7. Annual vehicle class distribution for site 285805.

Figure 8. Annual load spectra for site 285805.

Figure 9. Average annual load spectrum for site 285805.

Figure 10. Projected annual ESALs for site 285805.

Evaluation and Testing of LTPP Traffic Feedback and Resolution Package

The package described in the preceding section was the final version of the one that was distributed to the majority of the participating agencies. It was developed and refined over the course of several months through an interactive process involving reviews by interested parties and by representatives of several participating agencies. The two main review activities were:

• Review of the package by representatives of RCOs and by other interested parties.
• Review and enhancement of the package by pilot studies.

In addition, throughout the course of Phase 2 work, several small changes and enhancements to the package were instituted.

Review by Representatives of RCOs and Other Interested Parties

The important role of the RCOs in the projection process was recognized from the beginning of the study. Representatives of RCOs reviewed the first version of the LTPP Traffic Feedback and Resolution Package and subsequently reviewed agency-specific packages for all participating agencies in their respective regions.

The first package also was submitted to all members of the LTPP ETG on Traffic Data Collection and Analysis for review and comments, and a later version was presented during an ETG meeting in the spring of 2000. Reviews of an early version of the Package were also obtained from the representatives of the Technical Support Services Contractor and from the Minnesota Department of Transportation (Mn/DOT).

All reviewers provided valuable comments and suggestions on how to make the package more effective and user-friendly.

Review and Enhancement of the Package by Pilot Studies

During Phase 1 traffic projection work, it became apparent that it was necessary to involve representatives from the participating agencies in the traffic projection process because many data problems cannot be resolved without their involvement and help. The primary contact between the participating agencies and the LTPP program has been through RCOs. RCOs provide general technical support to the agencies regarding traffic data collection and analysis issues and are responsible for initial traffic data quality checks and for processing traffic data collected by the agencies. Consequently, the involvement of both the participating agencies and RCOs in traffic data assessment and projection is important. Pilot studies brought together the representatives of the participating agencies, RCOs, the project team, and others.

The main feature of the pilot studies was a 1-day meeting that was held at the participating agencies and attended by representatives of the participating agency (typically including an LTPP contact engineer, personnel responsible for traffic data management, field traffic data collection personnel, and others), RCOs (personnel responsible for traffic data), FHWA (representing the LTPP program, and FHWA Division Office), and two members of the project team.

The purpose of the pilot studies was to review and enhance the process of traffic data assessment and projection of traffic loads, including the content of the LTPP Traffic Feedback and Resolution Package, and to discuss specific issues concerning the agencies' LTPP sites. The pilot studies also provided an opportunity for the representatives of the RCOs to become better acquainted with the package and traffic data assessment and projection issues. For this reason, one pilot study took place in each of the four LTPP regions:

During the course of the meeting, attendees discussed the LTPP package prepared for the agency. Agency representatives provided comments on specific LTPP sections following the format of the Feedback and Data Resolution Sheet (figure 3).

The pilot studies resulted in several improvements to the package, and in better communication among all interested parties, particularly between the representatives of the RCOs and the project team.