Comprehensive Truck Size and Weight Limits Study - Volume 1: Technical Reports Summary
Chapter 3: Results of Focus Area Analysis
This chapter summarizes the Technical Reports conducted for the five study focus areas and briefly describes the primary results. The summaries are discussed in the following order:
- Modal Shift Analysis
- Safety Analysis
- Pavement Analysis
- Bridge Analysis
- Compliance Analysis
Each summary discusses the purpose, methodology (including data and models used), assumptions/limitations, and the technical results of each of the focus areas. In keeping with the MAP-21 legislation, the study presents results for each of the assessments outlined by Congress in §32801 of MAP-21. No conclusions or recommendations on national truck size and weight policy were developed as part of the study.
The study examines vehicles that are much closer to current Federal weight and size limits than those that were assessed in previous studies. Specifically, the study examines six alternative truck configurations in six scenarios. The first three scenarios would allow heavier tractor semitrailers than are, in some cases, allowed under current Federal law. Scenario 1 would allow five-axle (3-S2) tractor-semitrailers to operate at a GVW of 88,000 lbs., while Scenarios 2 and 3 would allow six-axle (3-S3) semitrailers to operate at GVWs of 91,000 lbs. and 97,000 lbs., respectively. Scenarios 4, 5, and 6 examine vehicles that would serve primarily less-than-truckload (LTL) traffic. Scenario 4 examines twin trailer combination trucks with 33-ft. trailers (2-S1-2) with a GVW of 80,000 lbs. Scenarios 5 and 6 examine triple-trailer combination vehicles with 28- or 28.5-ft. trailers with a GVW of 105,500 lbs. (2-S1-2-2) and 129,000 lbs. (3-S2-2-2), respectively. Table 4 shows the six alternative truck configurations and summarizes the scenarios examined in this study.
Modal Shift Analysis
The modal shift analysis provides the foundation for assessing a range of potential impacts associated with the truck size and weight scenarios analyzed in this study. “Modal shift” refers to shifts in freight usage between truck and rail modes as well as across vehicle types and operating weights within the truck mode.
The purpose of the modal shift analysis is to quantify the potential nationwide impacts of changes in trucks size and weight limits. Specifically, the work conducted in the modal shift area included:
- Estimating freight shifts between trucks and between truck and other modes due to the introduction of alternative truck size and weight limits under the six scenarios examined in the study;
- Estimating other impacts from shifts in the vehicle or mode carrying freight, including energy, emissions, traffic operations; and
- Providing a framework for assessing the potential impacts if one or more of the six scenarios were implemented in terms of:
- The total number of trips and VMT required to haul a given quantity of freight,
- The transportation mode chosen to haul different types of freight between origin and destination,
- The truck configuration and weights used to haul different types of commodities,
- The costs of enforcing Federal truck size and weight limits,
- Energy requirements to haul the Nation’s freight,
- Emissions harmful to the environment and to public health,
- Traffic operations on different parts of the highway system,
- Total transportation and logistics costs to move freight by surface transportation modes,
- The productivity of different industries, and
- The competitiveness of different segments of the surface transportation industry; and
- Providing data to aid in comparing the studied effects of the alternative configurations under possible modal shifts.
Modal Shift Analysis Methodology
This section summarizes the data and methods used in the modal shift analysis. The analysis begins with an estimation of current (base case) truck traffic (vehicle miles of travel) by vehicle configuration (number of trailers, number and types of axles, etc.), operating weight, and highway functional class. Data sources for base-case traffic estimates included:
- The volumes of truck traffic by highway functional class from the FHWA’s Highway Performance Monitoring System (HPMS),
- The distribution of trucks by vehicle configuration from vehicle classification data collected by the States and reported to FHWA, and,
- The distribution of vehicle operating weights from weigh-in-motion (WIM) data reported to FHWA by the States.
The USDOT study team conducted a desk scan (literature search) to identify and evaluate potential analytical tools and data sources for the modal shift analysis. It revealed that data and analytical tools have improved over the past 20 years, allowing for a much more robust picture of current commodity flows across the country. Data used in this analysis are primarily from 2011, the analysis year for the study, although in some cases WIM data were supplemented by 2010 and 2012 data to provide a more robust distribution of operating weights on different highway functional classes. A summary of base case traffic is presented in Modal Shift Comparative Analysis in Volume II of this 2014 Comprehensive Truck Size and Weight Limits Study.
Based on research and desk scan results, the FHWA’s Freight Analysis Framework (FAF) was selected as the commodity flow database for this study. The FAF integrates data from several sources to provide detailed estimates of freight movement among States and major metropolitan areas by all modes of transportation. One limitation of the FAF for the modal shift analysis is that origins and destinations in the database are reported for only 123 regions representing, overall, the largest markets in the country. This level of detail was too coarse for purposes of the modal shift analysis since it would not allow a detailed assessment of potential impacts of limiting the highway networks available for certain scenario vehicles. The Oak Ridge National Laboratory (ORNL) disaggregated the FAF and provided commodity flows for origins and destinations at the county level. Disaggregation of the FAF data did not produce a data set presenting higher accuracy, but this activity was a necessary step in developing data suitable for use in estimating truck-to-truck and intermodal shifts that may occur as produced in the assessment of each of the scenarios.
The analytical tool used for the modal shift analysis itself was the Intermodal Transportation and Inventory Cost Model (ITIC). The USDOT developed this model during the course of and immediately following the issuance of its 2000 Comprehensive Truck Size and Weight Study (2000 CTSW Study), and it was used for subsequent studies by both FHWA and the Federal Railroad Administration (FRA). The ITIC model is described in detail in the Modal Shift Comparative Analysis report.
In general the model estimates transportation and non-transportation logistics costs for shipments of different commodities by different vehicle configurations and transportation modes between various origins and destinations. Specific costs considered in the ITIC model include vehicle operating costs, shipping rates that vary by market, and inventory carrying costs such as safety stock, cycle costs, and in-transit costs. If the costs for moves by scenario vehicles are lower than the costs for the same move in existing vehicle configurations at current size and weight limits, the move would be assumed to shift to the heavier scenario vehicle. Likewise if shipments by scenario vehicles cost less than shipments by rail, traffic would be assumed to shift from rail to truck.
In the ITIC model, railroads are assumed to respond to increased competition from more productive trucks by lowering their rates to the point where rates equal variable cost. If lowering the rates reduces total transportation and logistics costs for rail below rates for the scenario vehicles, traffic will remain on the railroads, but the contribution of those shipments to covering railroad fixed costs will be reduced.
Modal Shift Analysis Assumptions and Limitations
The USDOT study team made several assumptions when conducting the modal shift analysis:
- Cargo less than 75,000 lbs. GVW will not divert to (3-S3) six-axle tractor-semitrailers.
- Traffic currently moving as (3-S2) five-axle tractor-semitrailers that cannot benefit from the added weight allowed on a six-axle tractor-semitrailer will not shift to the six-axle vehicle.
- Carriers would not shift their entire fleets over to (3-S3) six-axle vehicles simply to increase the flexibility of their fleets.
- All scenario vehicles except triples have the same access to cargo origins and destinations as base case vehicles. In the short run, bridge or other highway improvements may have to be made before scenario vehicles could use the same routes as base case vehicles, but in the long run it is assumed that such improvements would be made. The modal shift analysis is based on this long-run assumption.
- Triple configurations operate in less-than-truckload (LTL) line haul (terminal to terminal) operations. In actuality there may be a few markets where heavy triples could be used for truckload shipments under the network and access restrictions placed on triples operations, but based on discussions with industry experts, those are believed to be localized and would have very little impact nationally.
- Equipment currently being hauled in specialized configurations such as truck-trailer combinations will not shift to scenario vehicles. Specialized configurations are used because of unique commodity characteristics that would not be met by the scenario vehicles.
- The Surface Transportation Board’s (STB’s) Carload Waybill Sample data were used to analyze the potential shifts from rail to truck because it includes more detailed origin, destination, and other shipment characteristics than FAF. The Waybill Sample data also includes information on rates paid for each series of moves. 90 percent of short-line carloads interline with Class 1 railroads and thus are reflected in the Surface Transportation Board’s Carload Waybill Sample.
- The analysis year for the study is 2011. To the maximum extent possible all data used for the study are from 2011 or have been adjusted to reflect 2011 values.
- The analysis assumes Federal and State highway user fees on the scenario vehicles are unchanged.
- The base year for vehicle-miles-of-travel data was set at 2011. No projections of future travel levels were made since results projected for future years may impact the quality of the comparative assessments Congress outlined in MAP-21. Questions and concerns about future projections would have negative effects on the quality of the assessments completed in the study.
Several data limitations affected the analysis, including:
- The precise origins and destinations of shipments are unknown from the FAF. Origins and destinations are assumed to be county centroids 12 for inter-county shipments.
- The precise routes used to ship commodities between origins and destinations are unknown. Shortest path routes between each origin and destination pair are calculated for purposes of estimating transportation costs.
- Characteristics of specific commodities within broad commodity groups may vary significantly.
- Shipment sizes and annual usage rates for freight flows between individual origins and destinations cannot be discerned from the FAF and must be estimated from VIUS and other sources. This affects non-transportation logistics costs.
- Truck/rail intermodal origins and destinations are not reported in the Carload Waybill Sample and have been estimated using the same assumptions that were used in the 2000 CTSW Study.
- Multi-stop truck moves to accumulate and/or distribute freight from/to multiple establishments are not captured in the FAF.
Such limitations are unavoidable in a nationwide study such as this. They also were encountered in USDOT’s 2000 CTSW Study and in other national studies. It is not believed that these limitations affect overall study conclusions, but the limitations must be kept in mind when considering study implications.
Cost Responsibility Issue
The issue of cost responsibility often arises in connection with truck size and weight policy studies. Many truck size and weight policy options, including those examined in the current study, have highway investment implications, both in the near term and over time. These costs can be linked to changes in highway travel by different vehicle configurations at different weights as the result of the truck size and weight policy changes. Many costs including pavement and some bridge costs estimated in this study are related not just to operating weight, but also to specific axle loadings for the various vehicle classes. To estimate the responsibility of different vehicle classes for changes in highway investment requirements, the distribution of axle loadings by vehicle classes affected by introduction of the alternative configurations would have to be known. Estimates on the allocation of highway costs are conducted in the Highway Cost Allocation studies periodically prepared by FHWA and follow a methodology that identifies the cost implications of operating a wide variety of vehicle types and identifies user charges and fees paid by the vehicles studied. This study did not extend beyond the boundary of identifying the impacts that each alternative configuration was estimated to have or identifying impacts that trucks operating above current Federal size and weight limits have on highway infrastructure and safe roadway operations. This is consistent with the comparative assessments outlined in MAP-21.
Summary of Modal Shift Analysis Results
A common set of 2011 vehicle miles of travel (VMT) data was constructed to analyze the base case and the six scenarios; this dataset also was used for assessing safety, pavement, bridge, and compliance impacts (See Table 3 in Chapter 2). The modal shift analysis assessed the shifts between the truck and rail modes, as well as the shifts freight among base case and scenario vehicle types and operating weights within the truck mode. Estimating the impacts on railroads is particularly important as the truck and rail modes are partners in some transportation markets, but are competitors in other markets. Increasing truck productivity could have serious economic consequences not only on railroads but also on the communities they serve. Finally, the analysis also estimated how modal shifts affect energy consumption, emissions, and traffic operations.
The modal shift analysis provides the basis for assessing the range of potential impacts associated with the truck size and weight scenarios analyzed in this study. These various impacts are discussed in each of the five technical reports in Volume II. Impacts are quantified to the greatest extent possible.
The modal shift analysis comprised the following elements:
- Developing a detailed project plan describing how the modal shift analysis was conducted using analytical tools and data identified during research and through the desk scan.
- Estimating truck traffic currently operating within and above existing Federal truck size and weight regulations.
- Specifying truck size and weight scenarios for analysis in the study. The basic vehicle configurations to be analyzed in the study were identified by USDOT, but specifications for those vehicles and how they would operate were developed for use in the various study tasks.
- Developing assumptions necessary for the modal shift analysis and identify limitations in the data and analytical methods that will affect the analysis.
- Estimating modal shifts associated with each scenario using the analytical tools and data chosen for the analysis.
In Table 6, a summary of the impacts is presented for each scenario on total truck VMT required to haul freight included in the 2011 FAF, the cost of moving that freight, and the impact of shifts from rail to truck on railroad profitability. As would be expected, impacts on VMT generally vary with the allowable GVW assumed in each scenario. Percentage changes in VMT reflect changes in VMT from the base case to the scenario size and weight limits for those vehicle configurations affected by each scenario. They do not reflect the percentage change in total VMT or total truck VMT, both of which would be much smaller than the percentage changes in VMT for just those truck configurations affected by the size and weight limits used in each of the scenarios. In terms of cargo tonnage that shifts from base case configurations to scenario configurations, the vast majority of the shifts occur among truck types rather than from rail to truck. Scenarios 1, 2, and 3 affect more tonnage because they primarily affect the movement of bulk commodities while Scenarios 4, 5, and 6 only affect LTL shipments.
Changes in total logistics costs and railroad contribution were much higher for Scenarios 1, 2, and 3 than for Scenarios 4, 5, and 6. Transportation costs are relatively higher for the bulk commodities most affected by Scenarios 1, 2, and 3, and there are few, if any, savings in non-transportation logistics costs associated with changes in the sizes of vehicles used to haul less-than-truckload freight. The greatest reduction in total logistics costs was associated with Scenario 3, where costs decreased by over $13 billion.
The percentage change in total logistics costs (transportation and non-transport logistics costs) for Scenarios 1, 2, and 3 is based on a comparison of total logistics costs associated with moving all traffic in the configurations affected by each scenario to total transportation and non-transport logistics costs associated with hauling the same traffic at the size and weight limits for each scenario. Changes in total logistics costs for Scenarios 4, 5, and 6 are calculated differently because those scenarios are assumed to apply only to LTL traffic. Total logistics costs associated with moving all LTL traffic both by truck and by rail in the base case are compared with total logistics costs associated with moving the same traffic under the size and weight limits assumed for each scenario. For all scenarios, the percentage change in railroad contribution reflects the difference between total revenues and total freight service expense. This contribution represents the amount available to cover fixed cost, income taxes, shareholder profits and capital investment to improve and maintain the system. The negative values indicate that net revenues fell more than freight service expense.
Stakeholders have expressed concerns about the potential impacts of changes in truck size and weight limits on short line railroads. Short lines provide regional/intrastate rail service, 90 percent of which connects to the larger Class 1 railroads. Data on short line operations in the Carload Waybill Sample are limited, but most commodities hauled by short lines are moved in carload quantities that would only be affected by the truck size and weight changes analyzed in Scenarios 1, 2, and 3. Using the same general methods that were used to analyze impacts to Class 1 railroads, estimates produced through the analysis indicated that short line railroads would lose between 1 and 4 percent of total revenue under each of Scenarios 1, 2, and 3. Revenue losses under Scenario 3 would be somewhat greater than those under Scenarios 1 and 2. Losses for some individual short line railroads could be greater. Although the analysis identified waybills that diverted under Scenarios 4, 5 and 6, due to data constraints with the reported revenue, the results were not included in the analysis.
As shown in Table 7, scenario impacts on energy consumption, emissions, and traffic operations reflect the reduced VMT presented in Table 6. Percentage changes in fuel consumption, carbon dioxide (CO2), and nitrogen oxides (NOx) were calculated in the same way that changes in VMT were calculated: changes in base-case fuel consumption and emissions for the vehicle configurations affected by each scenario were compared to fuel consumption and emissions for those same vehicles under the assumed size and weight limits for each scenario. Congestion costs decreased under all scenarios, reflecting changes in the relative VMT for each scenario. Congestion cost savings ranged from $256 million in Scenario 1 to $875 million in Scenario 4. The percentage change in congestion cost is estimated by comparing congestion costs for all vehicles operating on the highway under base case size and weight limits to congestion costs for all vehicles assuming the scenario size and weight limits. Impacts on congestion are not limited just to the vehicles whose VMT is affected by each scenario, but they accrue to all vehicles in the traffic stream. It should be noted that reductions in VMT calculated by the model due to the introduction of alternative configurations are very short in duration. It is estimated that these reductions will be offset by reasonably expected VMT increases in about 1 year.
Key: CO2 = carbon dioxide; NOx = nitrogen oxides.
Safety Comparative Analysis
This section summarizes the approaches and methods used and the results of the Highway Safety and Truck Crash Comparative Analysis in Volume II of this study. The comparative analysis explores the differences in safety risk and truck crash frequency between truck configurations currently operating on the Nation’s roadways at and below current Federal limits to those operating above such limits. The safety analysis also compares crash frequency and severity associated with base-line control vehicles with the six alternative truck configurations shown in Table 4 (page 22).
To determine these safety impacts, three different analytical approaches were pursued: 1) crash-based analyses; 2) vehicle stability and control analyses; and, 3) safety inspection and violations data analyses. The use of multiple approaches provides a richer understanding of the safety performance of the current and alternative truck configurations examined in this study, particularly in light of the crash data uncertainties discussed below. Each of the three approaches has its own advantages and limitations, but the results of the safety task analysis provide a broad picture of the potential safety implications of changes to the current Federal truck size and weight limitations.
Central to the approach was the recognition that Federal size and weight limits (e.g., 80,000 lbs. GVW) 13 apply to trucks operating on the Interstate Highway System and are frequently supplanted by grandfathering clauses and other statutory provisions that allow the legal operation of vehicles exceeding the Federal limits. In addition, State weight limits that apply to trucks traveling off the Interstate System differ from the Federal limits in several cases. These exceptions to national weight limits were considered in designing an approach and methodology that sought to analyze and compare 80,000-lb. control vehicles that operate on most U.S. roads with vehicles weighing more than 80,000 lbs. that operate on a more limited set of U.S. roads.
To ensure a consistent comparison, the study teams used the same data years for analysis, where possible (i.e., 2008 to 2012) and multiple sources of information were sought to reflect accurately the safety performance of the control and alternative configurations on the highway systems noted in Table 4. The main focus of the safety analyses was to estimate the changes in safety for each of the scenarios for multiple functional roadway classifications.
There were several challenges to producing nationally representative estimates of changes in truck safety associated with the scenarios in Table 4:
- Due to a lack of truck weight data for individual trucks in crash databases, the project team found it necessary to compare groups of control and alternative trucks based on the number of axles on the vehicle rather than comparing vehicles at specific weights (e.g., crash rates for an five-axle, 80,000-lb. 3-S2 control vehicle compared to crash rates for a six-axle 97,000-lb. 3-S3 configuration).
- Data limitations in annual average daily traffic (AADT) and WIM data restricted the crash analysis to rural and urban Interstates. These data limitations did not affect the vehicle stability and control and inspection and violation analyses.
- Most State crash databases lacked the data elements needed to identify the configuration of the truck (e.g., 3-S2). As a result, the State crash analysis and the development of crash estimates for Scenarios 2, 5, and 6 were based on configuration data from only one State, while Scenario 3 was based on data from only two States. Scenario 1 could not be analyzed due to the lack of truck weight in the crash data and Scenario 4 could not be analyzed since that alternative truck configuration did not currently begin its limited operations in the United States until very recently and thus does not have sufficient data for analysis.
- Due to the limited number of States with suitable data, the analysis of crash rates cannot be extended to other States or be used to draw meaningful conclusions on a national basis.
- In light of the lack of truck weight data on State crash reports, it was not possible to complete a comparative assessment between trucks operating at and below current Federal limits and trucks that operate above those limits.
Each of these challenges and their implications are discussed further in the Highway Safety and Truck Crash Comparative Analysis technical report.
Crash Analysis
The analysis focused on estimating the changes in crash rates for the control and alternative configuration vehicles on the roadway networks described in Table 4. The crash rate analysis was conducted using crash data from actual operations on U.S. roads, to the extent possible. The data included police-reported crash data in State files, crash information collected by trucking companies, and truck exposure data developed from different sources. The road safety profession has stated that analysis of crash, injury, and fatality data are, in fact, the definition of “safety analysis” (AASHTO 2010 and TRB 2011).
Crash Analysis Methodology
The crash analysis relied on State-based data (e.g., number of axles and trailers) that were combined to infer vehicle configuration in crash and exposure data.
A detailed review of crash databases from 15 States that allow the operation of six-axle heavy semitrailers and 17 States that allow the operation of triple-trailer combination vehicles indicated the absence of one or more of the needed data variables for the analysis. The lack of data describing the weight of the truck involved on State crash reports was the most persistent problem found. State crash data included no information on truck weight; fleet data from carriers were only slightly better. This meant that the only weight data available were the allowable GVW limits for different vehicle configurations within a given State (e.g., the maximum allowable GVW for a triple-trailer configuration). These weight limits were used to define groups of vehicles that could be compared within a State, effectively representing the comparisons shown in Table 8. When the potential States were limited to those where the maximum GVW limits for both the control and alternative vehicle configurations closely matched one of the above scenarios and were then further limited to those that included trailer and axle counts in the crash data, only four States could provide the needed data—Kansas, Idaho, Michigan, and Washington. The Highway Safety and Truck Crash Comparative Analysis technical report describes in detail the process undertaken to review State crash data for suitability in the study. Because of the State database limitations, Scenario 1 could not be conducted with crash data (i.e., the control and alternative vehicle can’t be differentiated in any State-based crash data set). Additionally, no crash data were available for the alternative configuration in scenario 4 as that vehicle type is not currently in wide use on U.S. roads.
All comparisons were conducted of configurations operating within the same State because State reporting practices and data records vary widely, making comparisons combining States inappropriate. Crash data also were obtained and used from three fleets operating triple-trailer combinations and from fleets operating six-axle semitrailer configurations above 80,000 lbs. but were insufficient for full analysis due to the small sample size. The carriers had difficulty in providing the exposure data for fleet-owned trucks in the selected States for the requested years. The primary difficulty was that carriers were not accustomed to analyzing safety based on road segment of travel, so their information systems could not readily supply the data requested. Crash report data and aggregate exposure data from some carriers was received. These data enabled the calculation of aggregate crash rates for triple trailer and double-trailer configurations, but the data were insufficient to allow for a more detailed comparison of configuration crash experience.
In addition to data describing crashes, VMT information was obtained from States and fleets as an exposure measure of the alternative and control configurations. The exposure data from the States was supplemented with WIM data. The limitations in WIM data (i.e., coverage) also limited the analyses done at the State level to the use of crash rates rather than to extensive regression modeling. Likewise, limitations in exposure data obtained from fleets provided a major challenge in the analysis of the fleet data. Crash records were generally available, but carriers did not consistently provide detailed route-level exposure data. As a result, simplified analyses were undertaken with fleet data.
Crash Analysis Assumptions and Limitations
Several key assumptions and limitations apply to the crash analysis. It was assumed that driver skills and management practices of firms in future operations will be similar to those in use today. This is an implicit assumption of the comparisons conducted in each scenario. While other studies have presented evidence from Canada and other nations that long combination vehicles (LCV) in general may experience very low crash rates if stringent restrictions are placed on drivers, routes, bad-weather operation, truck configuration equipment (e.g., dollies), truck components (e.g., brakes) and other safety-related factors (Woodrooffe, Anderson, et al. 2004), the crash analysis methodology used in this study did not take into account the degree to which, if any, such stringent restrictions would apply to actual crash data from the United States.
As previously discussed, the limitations encountered during the analysis included the limited number of triples in the current truck fleet from which to gather data. The lack of vehicle weight and configuration information in State crash data severely limited the analysis of on-road safety. The WIM data and vehicle classification data reported to FHWA by the States was relied on. Data limitations are more fully identified above and are addressed in the Highway Safety and Truck Crash Comparative Analysis technical report.
Summary of Crash Analysis Results
Table 8 summarizes the results of the crash analyses. It includes the results of successfully conducted analyses as well as information on analyses that could not be successfully completed due to data-related issues. This information is included to support the study's conclusion regarding needed changes in truck safety data. The Highway Safety and Truck Crash Comparative Analysis technical report provides more details concerning each of the crash analysis results.
Because of the small sample sizes available for some of the crash analyses, particularly for the triple-trailer configurations, the results of the test for statistical significance, reported in Table 8, have a p-value that is higher (i.e., p ≤ 0.15) than what is typically reported (i.e., p ≤ 0.05) in road safety research. The use of this broader range of significance levels has been suggested by others (e.g., Hauer 2004). In the table below, the term “significant” is used to refer to findings at the p ≤ 0.05 level, and the term “marginally significant” is used for findings with p-values between 0.05 and 0.15.
| Scenario | Data and Analysis Type | Results |
|---|---|---|
| Scenario 2 Target – 3-S3, 91,000 lb. semitrailer vs. 3-S2, 80,000 lb. semitrailer Limited State Crash Analysis – six-axle semitrailer with maximum allowable GVW of 91,000 lb. vs. five-axle semitrailer with maximum allowable GVW of 80,000 lb. (Washington data) Fleet Analysis – No fleet analysis conducted for this Scenario |
State Involvement Rates |
|
| State Regression Modeling |
|
|
| State Injury Severity Distributions |
|
|
| State Longitudinal Barrier Analysis |
|
|
| Fleet Crash Rates |
|
|
| Fleet Severity Distributions |
|
|
| Scenario 3 Target – 3-S3, 97,000 lb. semitrailer vs. 3-S2, 80,000 lb. semitrailer Limited State Crash Analysis – six-axle semitrailer with maximum allowable GVW of 105,500 lb. vs. five-axle semitrailer with maximum allowable GVW of 80,000 lb. (Idaho data) and 86,000 lb. (Michigan data) Fleet Analysis – No fleet analysis conducted for this scenario |
State Crash Involvement Rates |
|
| State Regression Modeling |
|
|
| State Injury Severity Distributions |
|
|
| State Longitudinal Barrier Analysis |
|
|
| Fleet Crash Rates |
|
|
| Fleet Severity Distributions |
|
|
| Scenario 5 Target – 2-S1-2-2, 105,500 lb. triple vs. 2-S1-2, 80,000 lb. twin Limited State Crash Analysis – Triple trailer configurations with maximum allowable GVW of 105,500 lb. vs. five- and six-axle double trailer configurations with maximum allowable GVW of 80,000 lb. (Idaho data) Fleet Analysis – Triple Trailer Configurations with unknown GVW vs. Twins with unknown GVW |
State Crash Involvement Rates |
|
| State Regression Modeling |
|
|
| State Injury Severity Distributions |
|
|
| State Longitudinal Barrier Analysis |
|
|
| Fleet Crash Rates |
|
|
| Fleet Severity Distributions |
|
|
| Scenario 6 Target – 3-S2-2-2, 129,000 lb. triple vs. 2-S1-2, 80,000 lb. twin Limited State Crash Analysis – Triple trailer configurations with maximum allowable GVW of 120,000 lb. vs. five- and six-axle double trailer configurations with maximum allowable GVW of 80,000 lb. (Kansas Turnpike data) Fleet Analysis – Triple trailer configurations with unknown GVW vs. Twins with unknown GVW |
State Crash Involvement Rates |
|
| State Regression Modeling |
|
|
| State Injury Severity Distributions |
|
|
| State Longitudinal Barrier Analysis |
|
|
| Fleet Crash Rates |
|
|
| Fleet Severity Distributions |
|
Based on the analyses conducted to quantify the safety of trucks on Interstate roads, several conclusions may be made.
- The lack of truck weight information recorded on State crash reports led to a comparative analysis of axle-based crash data. The cases analyzed, described previously, resulted in an investigation in truck crash information in three States and one roadway in a single State. The analysis is not robust enough to draw meaningful conclusions of crash relationships among the six alternative configuration vehicles and control vehicles at the regional or national level. Further research and sets of more robust truck crash data are required to present results better tailored to draw conclusions at the national level.
- In Michigan, Washington, and Idaho (the three States where tractor semitrailer data could be analyzed), the crash involvement rate for the six-axle alternative truck configurations is consistently higher than the rate for the five-axle control truck. The consistent crash involvement rates across these three States lend validity to this finding; however, additional study and research are required to develop an understanding of the causes contributing to the results.
- In Michigan, crash involvements of six-axle alternative truck configuration semitrailers increase much more quickly with an increase in exposure compared to the five-axle control vehicle. In Washington State, crash involvements of six-axle alternative truck configuration semitrailers increase similarly to those of the five-axle control as exposure increases. These contrasting results are explored in more detail in the Highway Safety and Truck Crash Comparative Analysis technical report.
- As has been noted in other research, the use of crash involvement rates based on truck crashes per truck VMT does not capture complete information because truck crash rates can vary based on changes in total AADT. Regression modeling was used to examine this issue. There was some indication in the regression modeling that the crash involvements of six-axle alternative truck configurations increase at a much faster rate with an increase in exposure when compared to five-axle semitrailers. This needs to be further verified in future studies in other States.
- Comparisons of crash injury severity distributions for the six-axle versus five-axle semitrailer configurations showed some indication of reduced severity for six-axle configurations. Analysis of Washington State data did not identify differences for the Scenario 2 distributions. Analysis of Idaho data for the Scenario 3 (97,000-lb. vehicle) indicated that the six-axle alternative truck involvements appear to be less severe than for the five-axle involvements on rural Interstates, urban Interstates, and when urban and rural are combined. Analysis of Michigan data for the same Scenario indicated that the six-axle alternative truck involvements on rural Interstates appear to be less severe than those for five-axle involvements, but no differences were found in the severity distributions for the urban or combined situations.
- Based on the Idaho data analysis, the Scenario 5 (2-S1-2-2) seven-axle vehicle crash involvement rates for triple-trailer combinations were lower than for the STAA twin semitrailer-trailer-trailer control vehicle on both rural Interstates and rural and urban Interstates combined, but the differences were marginally significant. (See Table 9 in the full technical report.) No differences were found in the Scenario 6 (3-S2-2-2) nine-axle semitrailer configurations vs. STAA twins semitrailer-trailer control configuration rates based on Kansas Turnpike data, even at the p=0.15 level of significance. (See Table 10 in the full technical report.) In both cases, the small sample of triple trailer configurations crashes makes drawing conclusions difficult.
- The results of the severity distribution analyses for triple trailer configurations and twin trailer configurations were mixed. The Idaho Scenario 5 (2-S1-2-2) seven-axle 105,500-lb. triple semitrailer study configuration appeared to be in somewhat less severe crashes than the STAA control vehicle twin semitrailer-trailer group. No differences were found in severity distributions for the study triple trailer configurations vs. control vehicle twin trailer configurations in the analysis of Scenario 6 (3-S2-2-2) nine-axle 129,000-lb. triple semitrailer configuration operating on the Kansas Turnpike. While the fleet data indicted no differences in severity distributions for twin trailer and triple trailer configurations for both all occupants and for truck occupants, there was a significant difference in the severity distributions of non-truck occupants who experienced less severe injuries in crashes with STAA twin trailer configurations.
- Due to data issues primarily related to either missing data or small sample sizes of the alternative truck configurations, planned analyses that could not be completed included the regression modeling for Idaho alternative truck configurations, the regression modeling for both Idaho and Kansas triple trailer configurations, the route-based analysis and the fleet crash rates analyses for the alternative truck configurations.
Vehicle Stability and Control Analysis
This analysis focused on the performance of the control vehicles and alternative configuration vehicles operating at various speeds under a variety of roadway geometric and braking ability conditions. These comparisons were completed in a simulation modeling environment with the exception of some supplemental braking distance testing that was previously completed and shared by the Federal Motor Carrier Safety Administration (FMCSA).
Vehicle Stability and Control Analysis Methodology
A set of vehicle stability and control analyses was defined to compare the simulated performance of the control and alternative vehicle configurations during specific maneuvers. The maneuvers included low speed off-tracking, high-speed off-tracking, straight line stopping distance, brake in a curve, and avoidance maneuver. Performance metrics included stopping distance, maximum path deviation, off-tracking, rearward amplification and lateral load transfer ratio. The analyses were performed using TruckSim®, a widely available numerical modeling package.
Simulated performance under several braking conditions was also analyzed. To supplement the results of the braking assessments, data and results from actual field testing done by the FMCSA and ORNL was added to the analysis to provide a more robust evaluation of stopping performance associated with the control vehicle and heavier single-trailer configurations. The analyses did not include vehicles equipped with electronic stability control since this equipment was not required at the time of the study under the existing Federal Motor Vehicle Safety Standard.
Vehicle Stability and Control Analysis Assumptions and Limitations
The assumptions applied in the vehicle stability and control analysis included the following:
- Dry van trailers with fixed, rigid loads;
- Steer axles with two tires, all others with duals on both ends;
- Multi-trailer combinations modeled with pintle hitch between trailer and converter dolly;
- Air ride suspension, not leaf spring;
- Vehicle characteristics common to U.S. practice;
- Simulations on dry pavement except brake in curve;
- Three braking conditions simulated:
- Anti-lock braking system (ABS) on all axle ends,
- ABS malfunctioning on one axle or both axles in tandem, and
- Brake failure on one axle end or one tandem end.
Vehicle Stability and Control Analysis Results
The results of the vehicle stability and control analyses for each of the scenarios are discussed here. The maneuvers simulated and analyzed included low- and high-speed off-tracking, stopping distance, stopping distance with or without brake failure or ABS malfunction, and avoidance.
The results of the maneuver simulations indicated that the alternative truck configurations in Scenarios 1, 2, and 3 did not differ appreciably from those of the five-axle control vehicle. Specific results include:
- None of the maneuvers identified a condition where the stability of a single-semitrailer combination was severely impaired by the addition of payload weight or a third trailer axle.
- Low- and high-speed off-tracking results were changed by amounts that would be difficult to measure in practice.
- Adding weight to the payload increased the stopping distance on dry road by less than 10 percent; in the proportions selected for the study, the additional brakes on the third trailer axle compensated for the additional payload in Scenario 2.
- Simulating a complete right-side brake failure on both drive axles increased the stopping distance, and the effect of that failure on the scenarios was similar to its effect on the control vehicle.
- The ABS malfunction caused a jackknife on all single-semitrailer combinations as expected; its severity did not appreciably differ between scenarios.
- The differences between the results for the four single-trailer combinations are not significant. Off-tracking is minimal for all scenarios.
The vehicle stability and control analysis for the Scenarios 4, 5, and 6 was compared to the control truck. Note that the payload weights used in the simulations are different from the allowable maximum weights that define the scenario configurations. (See Figure 4 in the full safety report for the payload weights.) The analysis yielded the following findings:
- Multi-trailer combinations were most challenged by the avoidance maneuver, which was formulated for that purpose. The final trailer in all four vehicles (i.e., the three alternative truck configurations and the twin 28.5-ft. control configuration) traced a wider path, experienced greater lateral acceleration, and put more load on the outside tires than did the tractor. The greater length of the 33-ft. trailers in Scenario 4 lowered the response slightly below that of the control vehicle with 28-ft. trailers. The amplification of the third trailer’s response in Scenarios 5 and 6 was greater than that of the second trailer in the control vehicle, as would be expected.
- Differences between the twins and the triples combinations in the off-tracking and braking maneuvers were present but not as significant as in the avoidance maneuver.
- The 33-ft. twin configuration (Scenario 4) had a higher average axle load than the other combinations and had a marginally higher stopping distance.
- When the ABS on the lead dolly malfunctioned during the brake in a curve, all 28-ft. combinations (i.e., twins and triples configurations) experienced a path deviation of 35 inches, which was short of a jackknife but would violate a 12-ft. lane. The 33-ft. combination of Scenario 4 was on the verge of instability, but its path deviation was not affected by the ABS malfunction under the specific conditions of this study.
- The high-speed off-tracking of the triple-trailer combinations was 8 to 9 inches greater than the control vehicle, but was still well within the width of a highway lane for that speed and curvature.
- All three multi-trailer study vehicles had a low-speed off-tracking roughly one-third higher than did the control double.
Inspection and Violation Analysis
The safety inspection and violation analysis compares vehicles currently operating at or below 80,000 lbs. with those operating above 80,000 lbs. The focus was to examine patterns of violation rates, out-of-service rates, and citation rates among the alternative truck configurations and the control configurations for different scenarios.
Inspection and Violation Analysis Methodology
Analysis of inspection and violation patterns for the control vehicle and alternative truck configurations used Level 1 truck inspection 14 data from FMCSA’s Motor Carrier Management Information System (MCMIS) database and GVW reported by roadside inspectors for select States over multiple years. MCMIS data from 2008-2012 were initially screened from 15 States allowing the operation of six-axle heavy semitrailers and from 17 States allowing the operation of triple-trailer combinations. After review of WIM data, 14 States were included in detailed statistical comparisons for vehicles in Scenarios 1, 2, and 3, and 10 States were included for detailed statistical comparisons for scenarios 5 and 6. (Note that the Scenario 4 alternative vehicle does not currently widely operate on U.S. roadways.)
A close inspection of the MCMIS data indicated that GVW data contained variable values, such as actual GVW versus manufacturers’ weight ratings. Because of the variability of the weight values in the MCMIS database, MCMIS data were supplemented with data generated through a cooperative data collection project with the Commercial Vehicle Safety Alliance (CVSA). Violations were further segmented by tractor semitrailers, twin trailers, and triple trailers.
Inspection and Violation Analysis Assumptions and Limitations
The USDOT study team applied several assumptions and limitations to the safety inspection and violations analysis. The study team assumed that the majority of MCMIS inspection data came from roadside inspections at both fixed and roadside facilities. WIM was widely used as a prescreening tool, but there is no indicator in MCMIS to identify whether GVW was captured from WIM or static scales.
In terms of limitations, there is an insufficient number of triple-trailer level 1 inspections to allow a comparison to double-trailers. In addition, MCMIS does not include exposure data.
Safety Inspection and Violations Results
The main results of the inspection and violation analysis are:
- Compared with commercial motor vehicles (CMV) operating at or below 80,000 lbs., CMVs operating over 80,000 lbs. show a higher percentage (18 percent) of brake violations and a higher number (0.76) of brake violations per inspection
- Legally operated trucks (i.e., trucks without overweight violations) weighing over 80,000 lbs. had higher overall violation and out-of-service (OOS) violation rates compared to those at or below 80,000 lbs.
- Twin-trailer configurations had the highest violation and OOS violation rates compared to tractor semitrailer and triple-trailer configurations.
- Triple-trailer configurations had four percent more brake violations (i.e., out of adjustment and all other violations) when compared with twin trailer configurations weighing 80,000 lbs.
Specific comparisons also were made between the 80,000-lb. 3-S2 configurations and the 88,000-lb. 3-S2, the 91,000-lb. 3-S3, and the 97,000-lb. 3-S3 configurations. A comparison of the 80,000-lb. twin trailer to the heavier triple-trailer configurations was considered, but could not be accomplished because of limited sample sizes. The results include the following:
- Alternative tractor semitrailer configurations (88,000 lbs., 91,000 lbs., and 97,000 lbs.) generally have higher violation, citation and OOS violation rates than the control semitrailer configuration group (80,000 lbs.). The exception is that the 88,000-lb. configuration had a lower out-of-service rate.
- Nevertheless, when placed in a regression model that accounts for other predictor variables, the tractor semitrailer configuration was not a significant predictor of the likelihood of a violation. That is, no significant difference was observed between the alternative tractor semitrailer configurations and the 80,000-lb. semitrailers with respect to violations, when controlling for other factors in the regression.
- Driver age, vehicle age, and carrier OOS rates were strong predictors of the likelihood of a violation. Driver age was negatively associated, while vehicle age and company OOS rate were positively associated with likelihood of a violation.
- Percentage of brake violations was roughly two percent higher in alternative semitrailer configurations than the reference configuration. This is mainly because of a higher percentage of “Brakes, out of adjustment” violations in those three alternative truck configurations.
Scenario Analysis Results: Estimating Changes in the National Number of Crashes
The concept underlying the development of the estimates of changes in truck crashes for each scenario required two components: nationally representative crash rates for each truck configuration and estimates of national VMT for both a base case of existing truck configurations and networks and for a scenario case involving alternative truck configurations and networks. Note that results were not generated for Scenario 1 and Scenario 4 since crash rates were not developed for the alternative truck configurations in these scenarios for the reasons previously noted.
- The impact associated with each scenario assessed in the study cannot be completed due to the lack of truck weight and vehicle characteristic information uniformly and reliably reported on State crash reports. This lack of information creates a situation where meaningful analysis leading to an understanding of the implications of each scenario cannot accurately be performed with an adequate degree of confidence.
- The findings for the Scenario 2 (91,000-lb. 3-S3) configuration and the findings for the Scenario 5 and 6 triple configurations were each based on crash rates for one State. The findings for the Scenario 3 (97,000-lb. 3-S3) configuration were based on crash rates from two States. The use of rates from this limited number of States clearly raises questions as to whether these rates can be considered nationally representative and whether using them to predict nationwide estimates is appropriate.
- Because State crash data do not include information on operating GVW for each truck, the definition of truck crashes used in the different scenarios was based on trailer and axle counts and State GVW limits. Is it not known whether actual truck GVWs in the fleet analyzed in this study will be similar to actual GVWs in an expanded future fleet.
- The composition of future fleets of alternative truck configurations may differ in unknown ways from the current fleet that was analyzed in this report. For example, the same alternative truck configuration analyzed here (e.g., 129,000-lb. triple configurations) may carry different commodities in the future. If so, the carriers may differ, which in turn may cause the “safety culture” to differ (e.g., driver training and experience, truck maintenance procedures, equipment age, etc.). The effect of such possible differences could not be analyzed here. For example, while crash data contains information on driver age, there is no driver age-specific truck exposure data, a critical need in any analysis of driver age effects.
These data limitations raise significant questions concerning the accuracy, reliability, and validity of any nationally representative crash rate estimates that could be calculated for each truck configuration. As a result, meaningful national-level crash-rate results could not be developed for this study.
Summary of Inspection and Violation Analysis Results
As noted earlier, the crash rates used in all scenario analyses were based on either one or two States. The use of rates from this limited number of States clearly raises significant questions concerning whether estimates could be considered nationally representative. FHWA does not believe nationally representative estimates can be developed from the data.
The analyses indicate that the safety implications of allowing alternative truck configurations to operate vary by vehicle. In general, for Scenarios 2 and 3, the six-axle configurations have higher crash rates than the five-axle tractor-semitrailer control configurations in Washington, Idaho, and Michigan. This is particularly evident in the two study States where six-axle trucks could run at weights close to the 97,000-lb., six-axle alternative configuration. Similar findings with respect to inspections and violations were observed. The six-axle configuration had higher violations, OOS rates, and brake-related violations per inspection when compared to the control group (i.e., the five-axle tractor semitrailer configurations at 80,000 lbs.).
The vehicle control and simulation analyses showed very marginal differences between the control and alternative truck configurations for the set of maneuvers evaluated. The differences between the crash and vehicle control and simulation results could stem from the fact that crash rates for actual operations versus simulation-based operations do not reflect the same range of operators and/or operating conditions. It was not possible to determine in this study what factors led to these differences. Further exploration is needed.
Scenarios 5 and 6 results for triple-trailer alternative truck configurations also differed between the crash and vehicle stability and control methods. While no differences between triple-trailer and twin-trailer configurations were seen in the Scenario 6 Kansas Turnpike data, the crash rate analyses for Idaho (Scenario 5) indicated that the rates for the triple-trailer configuration were lower than those of the twin trailer configuration. The Level 1 inspection summary data for safety inspections and violations also showed that triple-trailer configurations tend to have lower violation rates than twin-trailer configurations. However, this is based on a very small sample size, and as a consequence, more rigorous analysis could not be conducted to explore this further.
A major result of this overall effort is that crash-based studies focusing on truck size and weight and using U.S. data are very difficult to conduct successfully. This is particularly true if the studies are based on the primary data sources in existence today – State crash files, State roadway inventory data, State AADT data, and additional data on VMT for specific truck configurations. Fleet supplied and MCMIS data were also inadequate to conduct the desired analyses. The issues found in this safety analysis are not new. These include the following:
- Crash data do not include precise information about the configuration (for example, number of axles and number of trailers and semitrailers) and the weight of trucks involved in crashes.
- The single source of State and national truck VMT information for the specific configurations of interest in this safety study is FHWA’s Traffic Monitoring Program traffic volume, vehicle classification and WIM data described earlier. WIM data is especially important since vehicle weight is a key factor used in in performing the several comparative analyses required. The number of current WIM data collection points is so limited that the estimate of truck travel by weight category was extremely constrained and limited to Interstate System roadways in a number of cases. Truck VMT using traffic volume and classification count data could only be provided at the State functional class level and not at finer levels, such as corridor-based comparative assessments as was proposed and attempted at the outset of this study.
- The data used to support the inspection and violations analysis included the selection of the GVW variable from the MCMIS database. Discussions with FMCSA indicated that this variable is not always available in the database as a measured weight, and that no better variable exists in MCMIS for a description of combination vehicle weight.
References
E. Hauer (Hauer). 2004. “The harm done by tests of significance.” Accident Analysis and Prevention, Vol. 36 (5), pp. 495-500.
Woodrooffe, J., D. Anderson, et al. 2004. The Influence of Policy on Crash Rates of Long Combination Vehicles. 8th International Symposium on Heavy Vehicle Weights & Dimensions, Johannesburg, South Africa.
Pavement Analysis
This section summarizes the pavement analysis conducted as part of the study. The pavement analysis assessed the impacts that trucks operating at or below current Federal weight limits have on pavement infrastructure compared to trucks operating above those limits. The analysis also assessed the impacts on pavement infrastructure associated with potentially allowing alternative the truck configurations identified in the six scenarios described in Table 4.
The purpose of the pavement analysis is to address two major questions:
- How will changes in axle weights and types resulting from each scenario affect pavement performance and expected pavement costs?
- How much pavement damage is currently caused by trucks operating above the current Federal weight limit versus trucks operating at or below those limits?
Pavement Analysis Methodology
A multi-step approach was used to assess the impacts of various truck types and traffic scenarios on pavement performance and life-cycle costs. Key to the process was the selection of representative pavement sections (flexible and rigid along with their local materials and design inputs) within each of the four primary climate zones in the United States—wet freeze, dry freeze, wet no-freeze, and dry no-freeze—and a single location within each climate. Three truck traffic levels—high-, moderate-, and low-volume—were also identified.
Through the desk scan, a thorough understanding of the current state of research and practice regarding pavement cost analysis related to heavy-vehicle use was gained. The information gleaned from the scan assisted researchers in the selection and application of analytical tools and in the compilation of data required for those tools. Some of the sources evaluated are listed below.
The approach used differs from past truck size and weight studies in that the current AASHTOWare™ Pavement ME Design® software was used to assess the structural impacts of evolving vehicle types and traffic scenarios on pavements. The Pavement ME Design® software is the tool based on the AASHTO Mechanistic Empirical Pavement Design Guide (M-EPDG) procedure that was adopted by AASHTO in 2007. The MEPDG directly applies an axle-load spectrum to calculate the amount of damage produced by the estimated range of traffic loads. The axle load spectra data are obtained from processing WIM data and include axle-load distributions (e.g., single, tandem, tridem, quads) and axle-load configurations (e.g., axle spacing and wheelbase). Neither the AASHTO software nor the M-EPDG was available at the time of the 2000 CTSW Study.
Several FHWA data sources were used to estimate the pavement impacts, including the Highway Performance Monitoring System (HPMS); vehicle classification and weight data reported by the States, the Long-Term Pavement Performance (LTPP) database, and calibration data provided by four State departments of transportation for use in the AASHTOWare™ Pavement ME Design® model.
Pavement Analysis Assumptions and Limitations
Several key assumptions and limitations apply to the pavement analysis. Analysis of the relative impacts of one group of vehicles compared with another at the national-system level requires some simplification of assumptions about the vehicles themselves. For each scenario in this study, freight was shifted either from one vehicle to another, or to a vehicle of the same type but with a different weight. The approach used in this study assumed that both the before and after vehicles in each scenario had the same temporal use patterns, the same tire and suspension characteristics, were traveling at the same speeds, and behaved in similar ways. The only variables considered for pavement analysis were the change in axle weights and vehicle types.
The main limitation of this study is that it considers only the initial service lives predicted by the AASHTOWare™ Pavement ME Design® model,version 2.0, and only for the distresses and pavement types that the software could suitably model. By implication, this means the study concentrates only on a subset of the impacts of the proposed changes in truck size and weight on pavement life and consequent costs. Deterioration caused by the interaction of loads, construction deficiencies, or materials durability (e.g., deterioration of HMA 15 transverse cracks caused by low temperatures, deterioration of PCC “D” cracking 16) are outside the scope of this study, although it should be recognized that they can significantly impact the performance of pavements. In addition, the impacts of truck tire types (e.g., wide-based radial) and tire-pavement interaction (e.g., braking, torqueing, and other physical responses) are not considered. And again, a lack of data on local roads prevented the modeling of these facilities for the pavement analysis.
Pavement Analysis Summary of Results
The study analyzed the effect of overweight axles in current operations, defining overweight as single axles weighing more than 20,500 lbs. and tandem axles weighing more than 35,000 lbs. to be consistent with the axle weight group boundaries used in the vehicle weight analysis. Initial service intervals were found to increase significantly for both flexible and rigid pavement sections, except in the case of one rigid pavement section that did not reach the end of its initial service interval during the analysis period. Flexible pavement initial service intervals increased by between 19 percent and 34 percent and rigid initial service intervals increased by between zero percent and 10 percent when overweight axles were removed from the traffic mix.
The estimated impacts of the truck size and weight scenarios vary by scenario and by the pavement type and service conditions considered in the analysis. The use of the alternative truck configurations resulted in the following findings in comparison with the base case of current vehicle usage patterns:
- Scenario 1, which allows five-axle semitrailer configurations to operate at an 88,000-lb. GVW, resulted in a heavier array of tandem-axle loads;
- Scenarios 2 and 3, which allow six-axle, tractor-semitrailer configurations to operate at a 91,000-lb. GVW and a 97,000-lb. GVW, respectively, resulted in a transfer of some heavier tandem axles to tridem axles;
- Scenario 4, which allows five-axle, twin-trailer configurations with 33-ft. trailers to operate, showed an increase in the weight distributions of single-axle loads;
- Scenario 5, which allows seven-axle, triple-trailer configurations to operate at a 105,500-lb. GVW, resulted in the transfer of some tandem-axle loads to lighter single axles; and
- Scenario 6, which allows nine-axle, triple-trailer configurations to operate at 129,000 GVW, resulted in lower tandem axle weights as well as a similar shift from tandems to lighter single axles.
Table 9 summarizes the average impacts of each scenario, both in terms of time to first rehabilitation and in life-cycle cost. Flexible pavements exhibited more accelerated deterioration in Scenarios 1 and 4, while rigid pavements were more negatively impacted by Scenarios 4, 5, and 6.
The more significant modal shift impacts are predicted to occur on lower volume facilities, specifically the low volume Interstate highways and the low volume (other NHS) arterials, typically constructed with thinner cross-sections. The estimated impacts of the scenarios are relatively minor for the thicker pavement sections built to handle higher truck volumes. The range of impacts for each scenario results from varying pavement conditions, climatic conditions, and highway types
The life cycle cost (LCC) implications of the scenarios also varied. Table 9 also summarizes the differences averaged over all pavement types, climate zones, and types of facilities. Two discount rates were employed in estimating the present value of the repair and restoration costs modeled for each pavement section sample. A conservative discount rate of 1.9 percent was applied and a more widely used discount rate of 7 percent was applied so as to frame the range that the results of the analysis completed. On average, Scenario 4 resulted in the largest LCC overall increase of 1.8 to 2.7 percent from the base case, whereas Scenarios 2 and 3 resulted in 2.4 to 4.2 percent and 2.6 to 4.1 percent decreases, respectively, in predicted LCC from the base case. Scenarios 1, 5, and 6 showed lower increases in LCC, which is defined herein as agency cost for pavement rehabilitation (e.g., overlays, retexturing) over a 50-year analysis period.
Note: Individual pavement sections were weighted based on the number of lane-miles of pavement of each type, thickness range, and highway type.
References
Chatti, K. “Effect of Michigan Multi-Axle Trucks on Pavement Distress.” Michigan DOT and Michigan State University, Final Report, Executive Summary, Project RC-1504. February 2009. http://www.michigan.gov/documents/mdot/MDOT_Research_Report_RC-1504__ExecSum_272183_7.pdf
Chatti, K., Salama, H., and C. Mohtar. “Effect of Heavy Trucks with Large Axle Groups on Asphalt Pavement Damage.” Presented at 8th International Symposium on Heavy Vehicle Weights and Dimensions, Johannesburg, South Africa, March 2004. http://road-transport-technology.org/Proceedings/8%20-%20ISHVWD/EFFECT%20OF%20HEAVY%20TRUCKS%20WITH%20LARGE%20AXLE%20GROUPS%20ON%20ASPHALT%20PAVEMENT%20DAMAGE%20-%20Chatti.pdf
Timm, D., Turochy, R., and K. Peters. Correlation between Truck Weight, Highway Infrastructure Damage and Cost. Auburn College of Engineering for FHWA, DTFH61-05-Q-00317, Subject No 70-71-5048. October 2007. http://www.eng.auburn.edu/files/centers/hrc/DTFH61-05-P-00301.pdf
Tirado, C., Carrasco, C., Mares, J., Gharaibeh, N., Nazarian, S., and J. Bendaña. “Process to Estimate Permit Costs for Movement of Heavy Trucks on Flexible Pavements.” Transportation Research Record: Journal of the Transportation Research Board, 2154, National Research Council. Washington, D.C., pp. 187-196, 2010. http://pustaka.pu.go.id/files/pdf/BALITBANG-03-C000066-610032011103843-process_to_estimate_permit_cost.pdf
Bridge Structure Comparative Analysis
This section summarizes the results of the Bridge Structure Comparative Analysis technical report and the methods used to assess the impacts that certain alternative truck configurations may have on bridge infrastructure. It provides estimates of the impacts to bridge infrastructure from trucks operating at or below the current Federal weight limits compared to trucks operating above those limits. Bridges located on the Interstate System (IS) and all other highways comprising the National Highway System (NHS) were assessed. The scope of the analysis was limited to the immediate structural effects on the existing bridge inventory and the impact on bridge load-induced fatigue that would result due to that change.
The bridge technical analysis work focused on two main analytical objectives:
- Structural Analysis: Determine and assess the implications of the structural demand on U.S. bridges due to the introduction of the proposed alternative truck configurations that have a GVW of more than 80,000 lbs. versus trucks in the current fleet that are subject to a maximum weight limit of 80,000 lbs. This task included an assessment of one-time bridge costs that might be incurred as a result of resolving posting issues (bridges that are not capable of handling the weight for which they were constructed legal loads are “posted” at a lower, safe weight) leading to the strengthening or replacement of those bridges).
- Bridge Damage Cost Allocation: Determine the increase or decrease in bridge damage-related costs expected to accrue over time due to the introduction of the proposed alternative truck configurations vs. the costs attributable to the current truck fleet. While it is strongly believed that an increase in axle load or number of axles accelerates bridge deck deterioration, because a suitable model based on generally accepted procedures was not available, this aspect of the analysis and the associated long-term costs were not included in the study results.
Bridge Structure Comparative Analysis Methodology
Both structural demand and bridge damage cost allocation analyses were conducted on bridges located on the three highway networks noted above. The load-induced, fatigue-related effects of trucks were evaluated regarding the impact on service life of bridges with respect to the degree to which structural fatigue may be affected by the introduction of the proposed alternative truck configurations on a national basis.
The results of the extensive desk scan and previous research affirmed the approach to the structural analysis of a representative sample of bridges screened from the National Bridge Inventory (NBI) database, and for determining bridge posting issues and one-time structural costs. Investigations into previously completed studies also assisted in the development of the framework for assessing the impacts that the alternative configurations would have on bridge load-induced fatigue.
Structural Analysis Methodology
The NBI database was first screened to determine both the total bridge count and the relative number of bridges on the NHS and NN by bridge type that are on the Interstate System and on the non-Interstate System within the two subject highway networks. The 12 most common bridge types were chosen for inclusion in the structural analysis, representing 96 percent of all bridges. More than 500 representative bridges were analyzed using AASHTO’s AASHTOWare™ Bridge Rating®Program (ABrR) using the load resistance factor rating (LRFR) method. Of these, 490 bridges were selected to best represent the mix of bridges in terms of bridge types, span length, and age, on the two highway networks referred to above. The only exceptions were for thru-trusses and girder-floor beam bridges for which there was not yet any LRFR capability in ABrR. The load factor rating (LFR) method was employed for those bridges.
The bridge models selected for analysis were in proportion to the number of bridges in the NBI by bridge type on the subject highway networks. The bridges were further screened to ensure that they were representative in terms of age, condition, and span length. The results of the analysis were recorded for maximum moment and shear, and the Rating Factors (RF) for the alternative truck configurations were compared to (normalized relative to) the 80,000-lb. control vehicles.
This analytical process is the basis for assessing the increase in the gross number of bridges that would have structural/posting issues potentially requiring strengthening or replacement as a result of the introduction of the alternative truck configurations. From this assessment, the one-time costs resulting from structural and posting related issues were derived.
Bridge Damage Cost Allocation Methodology
Prior work completed in the United States and around the world was exhaustively investigated, confirming that there is no generally accepted and applied approach for measuring the cost effects of heavy vehicles on bridges on a national scale. Consequently, a methodology for bridge damage cost allocation was developed as an axle-load based method, aggregated by truck class. Requests for alternative bridge deterioration models were made to stakeholders at publicly held meetings conducted by USDOT, and to the National Academy of Science, Peer Review Panel. Unfortunately, no generally accepted or state-of the-practice methodology was identified. Bridge Program subject matter experts recommended that this area of analysis be eliminated from the study due to a lack of a tool capable of estimating impacts at the national level.
However, the FHWA’s Long-Term Bridge Performance Program (LTBP) is in the process of collecting useful data to better understand bridge element performance and heavy-vehicle interactions. This effort is intended to lead to the development of the tool needed to assess heavy truck impacts on bridge decks.
Bridge Comparative Analysis Assumptions and Limitations
Two of the key assumptions applied in conducting the bridge analysis were that:
- Maximum legal axle weights would be used for both the structural and load-induced fatigue analysis, and
- Bridge capital costs would be based on the 2011 Financial Management Information System (FMIS) cost summaries, including both State and Federal shares.
Limitations affecting the analysis include:
- Lack of a generally accepted methodology prevented the estimation of costs associated with accelerated bridge deck deterioration due to increased truck weights or number of axles at this time. This limitation resulted in the inability to provide a complete bridge impacts analysis in accordance with the statutory direction.
- Little segregated cost data was available for deck preservation and preventative maintenance;
- The limited load-induced fatigue analysis performed supported only a qualitative assessment;
- LRFR capability was not available in ABrR for structural analysis of trusses and girder-floor-beam bridges.
Summary of Bridge Comparative Analysis Results
The Bridge Analysis examined a multiplicity of contributing factors and issues, including two 80,000-lb. control vehicles, six scenario alternative truck configurations, two regions, and two primary highway networks. The following discussion presents the results for the two areas of assessment completed: 1) Bridge Structural Analysis and 2) Bridge Load-Induced Fatigue.
Bridge Structures
Based on the derived rating factors for each of the alternative truck configurations in each scenario, an assessment was made of how many bridges had posting issues and would potentially require either strengthening or replacement. A threshold Rating Factor (RF) value of 1.0 establishes a potential need for bridge strengthening or replacement. Table 10 shows the projected number of posted bridges.
| Number of Bridges in the NBI | Load Rating Results | Projected Number of Bridges W/ Posting Issues For Entire Inventory | ||||||
|---|---|---|---|---|---|---|---|---|
| # of IS Bridges in the NBI | # of Other NHS Bridges in the NBI | # of IS Bridges Rated | # of Other NHS Bridges Rated | Vehicle Configuration | IS Bridges Rated w/ RF < 1.0 (percent) |
Other NHS Bridges Rated w/ RF < 1.0 (percent) |
# of IS Bridges w/ Posting Issues | # of Other NHS Bridges w/ Posting Issues |
| 45417 | 43528 | 153 | 337 | Scenario 1 | 3.3 | 5.0 | 1485 | 2194 |
| Scenario 2 | 3.3 | 7.7 | 1485 | 3360 | ||||
| Scenario 3 | 4.6 | 9.5 | 2080 | 4135 | ||||
| Scenario 4 | 2.6 | 3.0 | 1185 | 1293 | ||||
| Scenario 5 | 2.0 | 0.9 | 890 | 387 | ||||
| Scenario 6 | 6.5 | 5.6 | 2970 | 2455 | ||||
Comparing the number of bridges to be posted for each alternative truck configuration to the posting required for control vehicles (3-S2 and 2-S1-2) provided a reliable indication of how many additional bridges would need to be posted (or strengthened) if these alternative truck configurations were to be introduced as legal trucks on the NHS. Table 11 shows both the percentages and the actual number of bridges that have posting issues.
In order to estimate the probable one-time cost effect of employing alternative truck configurations, the increase in the potential strengthening or replacement costs relative to the control vehicles was developed. The calculated one-time cost of bridge improvements addressed herein could pertain to either superstructure strengthening or superstructure replacement triggered by the need to increase live load capacity. Costs were estimated for bridge strengthening and replacement using project cost information from FHWA’s Financial Management Information System (FMIS). A unit cost for this type of work was calculated ($235.00 per square foot of deck space), applied to bridges requiring strengthening or replacement and summarized for each scenario modelled. Bridges requiring improvement action on the Interstate System (IS) and National Highway System (NHS) were flagged for improvement when a rating factor equal to or less than 1.0 was observed. Costs by span length for IS and NHS bridges are found in Table 23 of the Bridge Structure Comparative Analysis Report. A full description of the cost analysis is found in Chapter 3 of the Bridge Structure Comparative Analysis Report.
The choice of strengthening vs. replacement would depend on superstructure type and whichever is the more economical alternative. The summary of what is considered the upper bound of these projected costs for each scenario’s alternative truck configuration is presented in Table 11.
Bridge Fatigue
The USDOT study team also investigated load-induced steel fatigue resulting from truck loadings. Four steel bridges of various span lengths, configurations (simply supported and continuous), and fatigue category details were investigated using a comparative analysis approach.
Results from the analysis showed that relatively heavier axle loads and axle groupings tend to affect fatigue life negatively when compared to the control vehicles. However, any overall reduction in bridge fatigue life depends on the number of relatively heavier trucks that are in the traffic stream. In general, fatigue-related costs of steel bridges are small compared to total bridge program costs.
Bridge Deck Deterioration, Service Life, and Preventative Maintenance
Initially, bridge deck repair and replacement costs and bridge deck preservation and preventative maintenance were investigated together because the topics are innately linked. Bridge deck limit states include the ultimate deck strength limit and the deck durability service limit. AASHTO design criteria (AASHTO 2002, 2011) provide bridge decks with adequate strength to carry the potentially heavier alternative truck configuration axle loads. However, cyclic axle loadings diminish deck service life or durability.
As noted above, the lack of a bridge deck impact model suitable for estimating bridge deck wear caused by commercial motor vehicles of various weights limited USDOT’s ability to evaluate the consumption of bridge deck service life and provide an estimate for related cost responsibility attributable to specific configurations and alternative gross vehicle weights. Because a suitable model based on generally accepted procedures and sound engineering principles was not available, this likely significant aspect of the analysis is not included in the study results.
References
American Association of Highway and Transportation Officials (AASHTO):
2011. Load Resistance Factor Design Specifications, 6th Edition, Washington, D.C.
2002. Standard Specifications for Highway Bridges, 17th Edition, Washington, D.C.
Compliance Comparative Analysis
The goal of this area of the study is to assess the cost and effectiveness of enforcing truck size and weight (TSW) limits for trucks currently operating at or below current Federal truck weight limits as compared with a set of alternative truck configurations in six scenarios.
At this point it is important to note that while the control double has an approved GVW of 80,000 lbs., the GVW used for the control double in the study is 71,700 lbs. based on actual data collected from WIM-equipped weight and inspection facilities and is a more accurate representation of actual vehicle weights than the STAA authorized GVW. Using the WIM-derived GVW also allows for a more accurate representation of the impacts generated through the six scenarios.
The cost analysis portion of this study includes a description of the principal TSW enforcement methods used in the U.S., including the application of enforcement technologies, meaning that the enforcement costs assessed reflect the resources required to undertake the truck size and weight enforcement task. The analysis examines national-level trends in enforcement program costs and conducts enforcement cost comparisons between States and for different truck configurations. Finally, the analysis estimates the enforcement cost impacts of introducing the alternative truck configurations into the traffic stream.
Enforcement program effectiveness reflects how the resources provided to the enforcement program translate into TSW enforcement actions and ultimately contribute to achieving regulatory compliance. The effectiveness analysis examines trends and relationships pertaining to enforcement program activities (such as weighing trucks) and compares the effectiveness among States and for different truck configurations. WIM data gathered at sites where alternative truck configurations currently operate provide the basis for comparing the compliance impacts of introducing these configurations into the traffic stream.
Compliance Comparative Analysis Methodology
Despite the widely held notion of a linkage between truck weight enforcement and compliance, there remains an inability to fully understand this relationship because of differences in how enforcement occurs and a lack of systematic and reliable evidence concerning overweight trucking. Additionally, understanding this relationship for specific truck configurations—one of the main issues of interest in this study—has generally been constrained by insufficient data. Increasing investments in proven enforcement technologies, including tools for identifying non-compliant trucks or carriers and the expanded use of WIM devices for monitoring truck weights, provide some opportunity to address these historical data limitations; however, certain data gaps persist which preclude a definitive analysis of the subject.
The analysis of costs and effectiveness undertaken in this study takes a performance-based approach. This approach considers enforcement program performance (or effectiveness) in terms of inputs, outputs, outcomes, and pertinent relationships between these measures. Enforcement program inputs reflect the resources (i.e., personnel, facilities, technologies) available to carry out the TSW enforcement task. State Enforcement Plans (and the subsequent certification of these plans) submitted by each State are the principal data source used to analyze program inputs.
Outputs reflect the way enforcement resources are used, the scale or scope of activities performed, and the efficiency of converting allocated resources into a product (e.g., quantity of trucks weighed, weight citations). These output measures are sourced from the Annual Certifications of Truck Size and Weight Enforcement database. While these outputs on their own provide some indication of program effectiveness, additional outputs and inputs can improve the overall understanding of program effectiveness.
The relationship between citation rate and enforcement intensity (measured as the number of trucks weighed per truck VMT) is one example. Outcomes reflect the degree of success of the TSW enforcement program in achieving its goal which from an operational and programmatic perspective is to achieve compliance with TSW regulations. The outcome measures used in this study are the proportion of axle or truck observations that fall within the Federal weight compliance limits compared to the severity of overweight observations.
Applying the performance-based approach provides the supporting framework for a comparative analysis designed to reveal insights about the costs and effectiveness of TSW enforcement programs. Data limitations, consistency, and availability constrain a comprehensive, representative understanding of these costs and effectiveness, particularly regarding vehicle-specific comparisons. To accommodate these limitations and leverage existing datasets and institutional knowledge, this study applies two types of comparisons:
- At a broad level, readily available State-specific data provides the foundation for comparing costs and effectiveness between States that currently allow trucks above Federal weight limits and those that do not. As the State-level data used in these comparisons do not allow disaggregation by vehicle configuration, these comparisons can be understood as a surrogate way of revealing potential vehicle-specific differences at a State level.
- A more detailed comparative analysis of enforcement program costs and effectiveness involves vehicle-specific comparisons (where possible). These comparisons focus on enforcement cost and effectiveness differences between the control vehicles and the six alternative truck configurations introduced into the traffic stream for the six scenarios in the study. Therefore, the results of the vehicle-specific comparisons directly support the scenario analysis, which estimates system-wide cost and effectiveness impacts that could result from the operation of the alternative truck configurations relative to the 2011 base case.
Summary of Compliance Comparative Analysis Results
Owing mainly to a lack of systematic and consistent data, prior research on TSW enforcement identifies the need for improved understanding of how enforcement resources, methods, and technologies can be effectively deployed to achieve better compliance. A configuration-specific understanding is particularly needed when considering the potential introduction of alternative truck configurations into the traffic stream, as is the case in this study. The State-level and particularly the vehicle-specific comparisons conducted in this analysis leverage existing datasets and, together, reveal insights about potential differences in enforcement costs and effectiveness for trucks operating within current Federal sizes and weight limits versus alternative truck configurations with higher sizes and weights. Additionally, these comparisons support a system-wide estimation of overall cost and effectiveness impacts that could occur under the scenario conditions.
Key findings concerning enforcement costs follow:
- From a national-level programmatic perspective, States spent a total of approximately $635 million (in 2011 U.S. Dollars) on their TSW enforcement programs in 2011. Personnel costs represented about 85 percent of total costs, while facilities expenditures (including investments in technologies) accounted for the remaining costs. Technologies play an important role in TSW enforcement and are increasingly deployed by State enforcement agencies.
- Based on the State-level comparisons, there is no indication of a change in enforcement costs that can be attributed to whether or not a State allows trucks to operate above Federal limits. Rather, differences in how States deliver enforcement programs (e.g., methods of enforcement used, technologies, intensity of enforcement) may have greater influence on total costs.
- The vehicle-specific comparative analysis indicates that, because the alternative truck configurations have more axles or axle groups than the control vehicles (except the Scenario 4 configuration with two 33-ft. trailers); they will require more time to weigh using certain standard weighing equipment and thus result in higher personnel costs.
- When estimating cost impacts on a system-wide basis in the scenario analyses, personnel costs decrease because the reduction in VMT predicted by the scenarios necessitates fewer weighings overall (assuming the rate of weighing vehicles relative to VMT is held constant) and this outweighs the increased costs associated with weighing the alternative truck configurations. Viewed another way, the rate at which weighings occur (per VMT) or the time spent conducting a weighing could be increased under the scenario conditions for the same level of expenditures on enforcement personnel.
Key findings concerning enforcement effectiveness follow:
- Considering national-level trends, both the weighing cost-efficiency (personnel costs per non-WIM weighing) and citation rate (citations per non-WIM weighing) decreased during the period from 2008 to 2012. The relationship between citation rate and enforcement intensity revealed that the citation rate decreases as enforcement intensity increases (i.e., more weighings per million truck VMT), but reaches a point of diminishing return. Moreover, those States that conduct a higher proportion of portable and semi-portable weighings generally have lower overall enforcement intensity and a higher citation rate. Measuring enforcement effectiveness in terms of a citation rate is complex because both relatively low and relatively high citation rates could be interpreted as a reflection of an effective enforcement program.
- Based on the State-level comparisons, as with the cost results, there is no indication of a change in enforcement effectiveness (as measured by the relationship between citation rate and enforcement intensity) that can be attributed to whether or not a State allows trucks to operate above Federal limits.
- For the vehicle-specific comparison of enforcement effectiveness, an analysis of data from selected WIM sites indicates that, except for six-axle tractor semitrailers operating off Interstates, the alternative truck configurations exhibit a higher proportion of compliant GVW observations than the control vehicles—hence our use of the 71,700 pound average GVW for those calculations involving the control double configuration. However, for all the comparisons, the intensity of overweight observations is higher for the alternative truck configurations than the control vehicles.
- In each of the scenarios analyzed, the system-wide impact on the proportion of total weight-compliant VMT for the control vehicle and alternative truck configuration is limited relative to the base case.
Identification of Statutes and Regulations
A final component of this Compliance Task identifies statutes and regulations impacted by the potential allowance of alternative truck configurations on all roads and highways on which Surface Transportation Assistance Act (STAA) vehicles can now operate. The review focuses on relevant language contained in:
- US Code Title 23: Highways,
- US Code Title 49: Transportation, as well as the corresponding regulations in
- Code of Federal Regulations, Title 23, Part 658 and Title 49 Parts 390-399.
The impacts identified by this review principally involve:
- Enactment dates for all applicable sections in 23 USC 127, 49 USC Chapter 311 and 23 CFR Part 658 pertaining to vehicle size and weight limits, as identified in the analysis;
- Length provisions replacing references to the twin 28-ft. and twin 28.5-ft. trailer combination vehicles as STAA vehicles with the twin 33-ft. trailer combination;
- The Federal Bridge Formula to enable operation of non-compliant configurations being assessed in the study; and,
- The listing of States and vehicle and route specific allowances provided in Code of Federal Regulations Title 23, Part 658 Appendix C.
12 A county centroid is the latitudinal and longitudinal (i.e., geographic) center of a county. See http://opengeocode.org/tutorials/USCensus.php for more information. return to Footnote 12
13 Although Title 49 (Transportation) of the Code of Federal Regulations differentiates between the gross weight of single vehicles [Gross Vehicle Weight (GVW)] and combination vehicles [Gross Combination Weight (GCW)], Title 23 (Highways) only refers to vehicle weights. Because this Study addresses truck size and weight assessments, it uses the term GVW to refer to the gross weight of combination vehicles. return to Footnote 13
14 There are six levels of DOT inspections. The comprehensive Level 1 inspection (referred to as the North American Standard Inspection) evaluates both the driver (license, medical certificate, and hours-of-service records, etc.) and the vehicle (brake and exhaust systems, suspension, steering mechanism, and frame, among other items). return to Footnote 14
15 Hot mix asphalt is a combination of aggregate (stone, sand, or gravel) bound together by asphalt. It is used primarily as a surface course to provide structural strength and distribute loads to underlying layers of the pavement. return to Footnote 15
16 Progressive deterioration of Portland cement concrete normally is caused by the winter freeze-thaw cycle. return to Footnote 16
