Click for next page ( 40


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 39
classes and over the two directions of travel.8 TrafLoad sets the monthly distribution factors for the Level 3A site equal to the corresponding factors for the associated site. Similarly, the hourly distribution factors for the Level 3A site are set equal to the ones for the associated site, if they exist.9 Level 3B Sites Level 3B sites are similar to Level 3A sites except that they have no associated Level 1 or Level 2 site. Instead, users are required to assign each Level 3B site to one of 17 Truck Traffic Clas- sification (TTC) groups that have been defined by the Pavement Design Guide team.10 Table 3.3 lists the 17 TTCs along with the criteria used to distinguish among them. The Pavement Design Guide software uses the TTCs as the basis for disaggregating estimated AADTT into the standard FHWA VCs. Pavement designs developed by the Pavement Design Guide software for Level 3B classifica- tion sites require the use of load spectra for the standard FHWA VCs. These load spectra nor- mally will be developed by TrafLoad. If these are not supplied by TrafLoad, the Pavement Design Guide software will use a set of default load spectra that have been developed from national data. 3.6 Forecasts The Pavement Design Guide software requires forecasts of linear or exponential rates of change in the AADTi over the design life of the pavement. A simple procedure for estimating 8 There are two potential improvements to the current TrafLoad procedure for handling Level 3A sites. One improvement would require TrafLoad to be modified to produce estimates of overall AADT for Level 1 and Level 2 sites (instead of just AADTi for truck classes). If TrafLoad has an overall AADT value for each associated site, then, for Level 3A sites, TrafLoad would require estimates only of AADT (but not percent trucks), since the percentage of trucks could be assumed to be the same as at the asso- ciated site. An alternative improvement would entail implementing a somewhat more sophisticated (and more demanding) algorithm for analyzing Level 3A sites. This algorithm would require that total traffic be counted at the 3A site at the same time as it is being counted at the associated site, with the counts at the 3A site being obtained by direction and, if practical, for a period of at least 48 hours. TrafLoad would then estimate AADTi for the Level 3A site, by direction, by using the volume counts at the two sites as the basis for scaling the corresponding estimates of AADTi for the associated site. 9 As discussed above, TrafLoad does not create hourly distribution factors for all Level 2B sites. Accord- ingly, TrafLoad does not create hourly distribution factors for some Level 3A sites that are associated with Level 2B sites. 10 ERES Consultants and FUGRO-BRE, Determination of Traffic Information and Data for Pavement Struc- tural Design and Evaluation, NCHRP Project 1-37A, Interim Report, December 1999, pp. 3953. 2-39

OCR for page 39
Table 3.3 Truck Traffic Classification Groups Percentage of AADTT in Key VCs TTC Description VC 9 VC 5 VC 13 VC 4 1 Major single-trailer truck route (type I) > 70 < 15 <3 - 2 Major single-trailer truck route (type II) 60 - 70 < 25 <3 - 3 Major single- and multi-trailer truck route (type I) 60 - 70 5- 30 3 - 12 - 4 Major single-trailer truck route (type III) 50 - 60 8- 30 0 - 7.5 - 5 Major single- and multi-trailer truck route (type II) 50 - 60 8 - 30 > 7.5 - 6 Intermediate light and single-trailer truck route 40 - 50 15 - 40 <6 - (type I) 7 Major mixed truck route (type I) 40 - 50 15 - 35 6 - 11 - 8 Major multi-trailer truck route (type I) 40 - 50 9 - 25 > 11 - 9 Intermediate light and single-trailer truck route 30 - 40 20 - 45 <3 - (type II) 10 Major mixed truck route (type II) 30 - 40 25 - 40 3-8 - 11 Major multi-trailer truck route (type II) 30 - 40 20 - 45 >8 - 12 Intermediate light and single-trailer truck route 20 - 30 25 - 50 0-8 - (type III) 13 Major mixed truck route (type III) 20 - 30 30 - 40 >8 - 14 Major light truck route (type I) < 20 40 - 70 <3 - 15 Major light truck route (type II) < 20 45 - 65 3-7 - 16 Major light and multi-trailer truck route 7 - 17 Major bus route - - - > 35% Source: ERES Consultants and FUGRO-BRE, Determination of Traffic Information and Data for Pavement Structural Design and Evaluation, NCHRP Project 1-37A, Interim Report, December 1999, pp. 40 and 42. these rates of change is presented below. Some more sophisticated forecasting procedures are discussed in Part 1, Appendix A. A Simple Procedure A simple procedure for forecasting the rates of change in traffic volumes for the design lane or design direction at any particular project site is presented below. In the procedure, the rates of change are referred to as "growth rates" to emphasize that, for the purpose of pavement design, traffic growth is of primary interest. However, the procedure may also be applied to sites at which traffic is expected to decline. The procedure consists of six steps: 2-40

OCR for page 39
1. Distinguish two groups of VCs: single-unit trucks and buses (FHWA Classes 47); and combina- tions (Classes 813).11 The distinction between the two VC groups permits the development of separate growth rates for single-unit trucks (which are used almost exclusively to serve the local economy) and combinations (whose usage responds to a much wider range of influences). 2. Identify all Level 1A sites for which estimates of AADTi have been developed for at least 4 years and that are believed to have historic rates of growth in the volume of heavy vehicles that are similar to those at the project site. 3. Associate the project site with one or more Level 1A sites identified in Step 2. Only Level 1A sites are used for this purpose because the AADTi estimates developed for these sites are likely to achieve a much greater level of consistency over time than estimates developed for other sites. 4. Use regression to estimate either linear growth rates or exponential growth rates for each Level 1A site for each VC group.12 In choosing between the two types of growth, a simple option is to choose the type that is believed to best describe expected future growth in truck traffic at the project site--linear growth if it is believed that the annual increase in this traffic is not likely to grow and exponential growth if this annual increase is expected to grow. (The Pavement Design Guide software has no provision for sites at which the annual increase is expected to decline over time. For such sites, linear growth should be assumed.) If this option is used, the same type of growth should be assumed for both VC groups (single- unit trucks and combinations). A slightly more complex option is to choose the type of growth that best fits the historic data at the Level 1A sites and then to modify the type of growth in Step 6. This option is discussed further in the latter part of the next subsection. If this option is used, the type of growth used in the regressions need not be limited to linear or exponential, and the regres- sions for single-unit trucks can be one type of growth and those for combinations can use a different type of growth. 5. For each VC group, average the growth rates obtained in Step 4 for the associated Level 1A sites. 6. Judgmentally adjust the growth rates on the basis of a review of macroeconomic and site-specific factors. The Step 6 review should consider any identifiable factors suggesting that future growth in heavy-vehicle traffic at the target site is likely to differ from past growth at the Level 1A sites. Factors to be considered include Expected changes in macroeconomic trends, 11 The vehicle-class groupings recommended here are the only ones handled by the current version of TrafLoad. 12 Regression capabilities are available in most computer spreadsheets as well as in many other types of software. 2-41

OCR for page 39
Planned and recently completed facilities that may affect the generation of truck trips, and Planned and recently completed highway projects that may affect truck routings. This last category includes both new and upgraded feeder routes and new and upgraded par- allel facilities. An interesting example of the effect of upgrading is the recent conversion of New Mexico SR-44 from two lanes to a four-lane divided highway (and its redesignation as US-550). The upgraded facility has attracted a significant amount of truck traffic heading northwest from Albuquerque that formerly used several other Interstate and U.S. highways. The facility is now feeding an increased number of trucks onto roads heading north and west from the Farmington area. Procedures that may be used for making the required adjustments are discussed in the next subsection. Adjusting the Forecast The adjustments made in Step 6 may be made directly to the (linear or exponential) growth rates developed in Step 5. Alternatively, it may be helpful to plot the Step 5 results in a spread- sheet and to use the plot as an aid in making the adjustments. As an example, assume that, for a site of interest, AADT of combination trucks in the base year is estimated to be 1,000 and the forecast growth rate is estimated to be 3 percent per year. The solid line in Figure 3.3 represents this forecast over a 20-year period. Forecast AADT for com- bination trucks at the end of this period is 1,754. Assume that the analyst believes that some downward adjustment of this forecast to a 2-percent annual growth rate may be appropriate. The dotted line in Figure 3.3 shows this alternative fore- Figure 3.3 Three Alternative Forecasts AADTCT 2,000 3% per Year 1,800 2% per Year 1,600 Linear 1,400 1,200 1,000 800 600 400 200 0 to to+20 Year 2-42

OCR for page 39
cast, producing a forecast-year value of AADT for combination trucks of 1,457. The forecast-year volume is 17 percent lower, but the total volume over the entire 20-year design period has only been reduced by 9.6 percent (from 9.81 million to 8.87 million). A third forecast assumes the same truck volumes in the base year and in the forecast year as those of the second forecast, but the third forecast assumes linear growth. This third fore- cast is also shown in Figure 3.3. This forecast assumes somewhat larger increases in traffic vol- umes than does the second forecast in the early years but smaller increases in the later years. The total volume over the 20-year design period is 8.97 million, slightly higher than the total volume produced by the second forecast. In the above example, the initial forecast (Steps 4 and 5) was developed under the assumption of exponential growth, but consideration was given to substituting linear growth. Such a pos- sible substitution is always an option, subject to two restrictions on the selected form for the forecast: It must be either linear or exponential and It must be the same for both VC groups.13 For the purpose of pavement design, the most important traffic volume estimates are the total numbers of trucks and buses, by VC, expected over the design period. For the three forecasts discussed above, these estimates are represented graphically by the areas under the three curves in Figure 3.3. In developing these estimates, it should be recognized that the most important information consists of the base-year estimates of existing AADTi. Forecasts of over- all growth in traffic volume (which are necessarily more speculative) are less significant--in the example, a 65-percent increase in overall growth (from an increase of 457 combinations per day for the two lower curves to an increase of 754 for the highest curve) increases the total number of combinations over 20 years by only about 10 percent. Using TrafLoad Although TrafLoad does not develop traffic forecasts, it does provide the user with substan- tial flexibility for entering forecasts. These options are summarized in Table 3.4. It allows the user to provide a linear or exponential growth rate developed in Step 5 of the above proce- dure. Alternatively, modified forecasts produced in Step 6 can be described in terms of the overall change in traffic volume over the forecast period or in terms of the volume forecast for the end of the period. For Level 1, 2, and 3A sites, TrafLoad allows separate specification of forecasts for two VC groups: single-unit vehicles and combinations. However, it does not currently allow separate forecasts for individual VCs. Exponential growth rates specified for one of the two VC groups are applied to each VC in the group. Linear growth rates specified for one of the VC groups are distributed among the corresponding VCs in proportion to their base-year volumes (AADTi). For Level 1, 2, and 3A sites, if TrafLoad has been requested to estimate AADTi for 13 The second restriction applies only if forecasts are entered into the system via TrafLoad. See the next subsection. 2-43

OCR for page 39
Table 3.4 TrafLoad Input Options for Forecasts User Inputs Input Option Linear Growth Exponential Growth Annual Annual change Annual percentage change Overall change Total change over period* Percentage change over period* Forecast AADT Forecast AADT for VC group* Forecast AADT for VC group* * Also requires specification of the base year and forecast year. the design lane, then linear forecasts of growth are interpreted as being for this lane; if TrafLoad has been requested to estimate AADTi for a given direction, then linear forecasts of growth are interpreted as being growth in traffic for that direction. For Level 1, 2, and 3A sites, TrafLoad also requires that the same input option and the same type of growth (linear or expo- nential) be used for both VC groups. For Level 3B sites, TrafLoad accepts only a single forecast of growth. This forecast is applied to total two-way truck volume, with no distinctions by lane, direction, or VC. 2-44