Review of U.S. Department of Transportation Truck Size and Weight Study - Second Report: Review of USDOT Technical Reports (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

23 3. BRIDGES The MAP-21 Section 32801 study charge requires the USDOT study to evaluate the impacts on the infrastructure (including bridges) of vehicles that operate under federal exemptions or grandfather rights (i.e., vehicles that federal law exempts from the normal federal weight or dimensional limits that apply on the Interstate or National Network systems) and the infrastructure impacts of allowing operation of alternative configurations, compared with configurations now allowed nationally under federal law. The USDOT report considers three kinds of effects of changes in size and weight limits: the change in numbers of bridges that would require posting for restricted truck traffic, strengthening, or replacement because their structural load-bearing capacity would be exceeded after the change in limits; change in the cost of load-induced fatigue in steel bridges; and change in the costs of bridge deck deterioration. Deck deterioration costs were not estimated because the study could not identify a suitable method. The fatigue analysis compared effects of alternative configurations and control vehicles for selected details on four actual steel bridges but did not scale the estimates to nationwide impacts or relate the effects to bridge costs. The cost of the change in the number of bridges with insufficient structural capacity was estimated as the cost of strengthening or replacing all such bridges. This strengthening or replacement cost is described as “the extreme upper bound of possible costs” (Bridge, 58) because states would choose to restrict truck traffic on some deficient structures rather than to retrofit or replace them. The structural analysis simulated the response of 490 actual bridges to changes in loads in each of the alternative configuration scenarios according to the AASHTOWare Bridge Rating software package, which is used by 36 states for rating the structural capacity of bridges. Sample bridges were selected in each of 22 categories: 11 structural types on the Interstates and the non-Interstate National Highway System (Bridge, 13). The number of sample bridges found structurally substandard (according to a specified criterion that is a function of bridge structure and loads imposed) was scaled to a national estimate on the basis of the systemwide total number of bridges in each of the 22 categories.

24 Responsiveness to the Questions Identified by Congress The response of the bridge analysis to the legislative charge is incomplete on account of the following omissions:  Missing costs (as identified in the following section);  Lack of a comprehensive framework that would allow the reader to put the USDOT study’s impact estimates in perspective with respect to total bridge-related costs and the factors that determine costs; and  Lack of estimates of bridge impacts of grandfather and other exemptions from federal limits in the states where the exemptions apply, and lack of evaluation of the ability of each state in which exempt trucks operate to recover the costs of the trucks’ bridge impacts. Methods and Data The conclusions below concern aspects of the bridge analysis that affect the credibility and usefulness of the estimates. Missing Costs The USDOT study does not include estimates of potentially important bridge-related costs:  Costs of changes in truck traffic on bridges on local roads: Local bridges tend to have shorter span lengths than the bridges on the major roads included in the study analysis. The alternative configurations are more likely to cause distress in short-span bridges than longer-span bridges.

25  Bridge deck costs: The USDOT report states: “It was not possible to draw national conclusions or present findings concerning the effect on overall bridge service life. While it is highly likely that bridge deck deterioration will accelerate with additional or heavier axle loads, the complex relationship of parameters that determine that performance is not well-defined.”5  User costs of delay at bridge construction projects or for detours around posted bridges: Even bridges that states selected for eventual strengthening or replacement would be posted for some time, until funds became available for construction.  Cost of fatigue: The USDOT study considers one fatigue-prone detail, cover-plate end welds in steel girder bridges. Other fatigue-prone details and distortion-induced fatigue, especially in gusset plates, are omitted. Fatigue cost in future years will depend on the rate of growth of truck traffic, an effect that the USDOT study does not take into account.  Expected or likely bridge structural costs: The cost estimates reported are described as an “extreme upper bound,” and the USDOT study concludes that “neither the actual costs nor the lower bound costs are determinate due to the range of program and policy decisions available to the States” (Bridge, 58). That is, actual highway agency and user costs will depend on decisions and management practices of state highway agencies, which the USDOT study does not consider. It is not clear that the cost estimated in the study is an upper bound, since the report does not explain the assumption in the calculation about whether bridges with insufficient capacity are strengthened rather than replaced.  Cost of building stronger bridges in the future: If changes in truck size and weight limits increase bridge capacity requirements, the cost of building bridges may be increased. Some states today build bridges to carry permit loads that are heavier than normal legal maximums; this practice may reduce the need to upgrade designs of future bridges if the federal limits are changed.  Costs of structurally deficient bent caps, columns, or foundations. 5 For comparison, the 2000 USDOT report concludes: “If total truck VMT decreases and axle loads do not increase as the result of TS&W [truck size and weight] limit changes, bridge deck deterioration may be reduced somewhat. No direct relationships currently exist between truck traffic, axle loads, and bridge deck deterioration, but research currently is underway to develop such relationships” (USDOT 2000, Vol. III, VI-4).

26 Because of these omissions, the USDOT study provides little information to policy makers about bridge costs, which the past USDOT and TRB truck size and weight studies have concluded would be the major infrastructure impact of changes in truck size and weight limits. Credible estimates of the order of magnitude of the missing costs would have been possible, based on methods used in past studies and reasonable assumptions, and sensitivity analyses could have been used to illustrate the degree of uncertainty in the costs. The following are examples:  An order-of-magnitude estimate of bridge deck costs could have been made on the assumption that the percent change in deck costs would be similar in magnitude to the percent change in pavement cost in a scenario, because both costs increase with frequency of axle loadings and increase nonlinearly with axle weights. Total annual baseline bridge deck costs could be estimated on the basis of deck sufficiency ratings, deck area, and current average deck repair costs in states that have these data. The baseline cost multiplied by the expected percent change in the cost in each scenario would be an estimate of the change in nationwide deck costs in the scenario. Such an estimate would be highly approximate; sensitivity analysis could provide an indication of whether the cost is likely to be important in comparison with other infrastructure costs.  States concerned about bridge cost impacts of accommodating specialized hauling vehicles (short, heavy trucks such as dump trucks and refuse trucks, typically with four to seven axles) are carrying out studies that could have provided a lower bound for bridge costs in some of the USDOT study scenarios. Cost impacts include needs for posting additional bridges, thereby limiting goods movement, or strengthening or replacing bridges. The concern is motivated in part by a new federal requirement that states rate bridges for these vehicles.  The 2000 USDOT study estimated the cost to road users of delay and rerouting caused by bridge construction. It concluded that this cost would be greater than highway agency costs for the construction (USDOT 2000, VI-11). Posting bridges that are not replaced also imposes costs on users.

27  The TRB Truck Weight Limits (TRB 1990a, 102–104) and Turner proposal (TRB 1990b, 152) studies estimated the average annual cost of building future bridges to the higher design standards that would be adopted to accommodate heavier trucks. This cost was estimated to be 10 to 25 percent of total bridge-related costs of various regulatory scenarios.  Two alternative kinds of estimates of structural costs would be relevant for policy decisions: an estimate based on assumptions about likely state responses to changes in truck traffic, determined by observing state practices; and an estimate based on the assumption that states make optimum decisions with regard to bridge maintenance, inspection, posting, weight limit enforcement, and retrofit and replacement, with all agency and user costs of their decision taken into account (TRB 2002, 73). The Turner proposal study described the effect on estimated cost of alternative assumptions about state decisions concerning posting or replacing bridges and presented a sensitivity analysis showing how the timing of bridge expenditures would affect highway agency costs (TRB 1990b, 153–156, 191–194). The TRB Commercial Motor Vehicles study compared the results of the bridge cost estimates in the 2000 USDOT study with actual state practices for selected bridges in one state (TRB 2002, 64–66) and outlined a method for estimating bridge costs if states followed optimal bridge management practices (TRB 2002, 69–75).  Costs of changes in truck traffic on bridges on local roads were estimated in all the past USDOT and TRB size and weight studies (e.g., USDOT 2000, VI-9; TRB 1990a, 99–100). The bridge structure cost estimated in the USDOT study is the cost of strengthening or replacing the increment, caused by introduction of each alternative configuration, in the backlog of structurally substandard bridges on the Interstates and the National Network. The USDOT report could have provided perspective on this incremental cost by noting the cost of eliminating the existing backlog. The 2000 USDOT study presented bridge costs in this way (USDOT 2000, VI-12, Table VI-2).

28 Bridge Sample The report contains only a general description of how the sample of 490 bridges was selected. USDOT explained to the committee that the bridges were sought and selected to make the sample appear representative with respect to age, span length, and other features. Moreover, the sample was limited to bridges from states that use the AASHTOWare Bridge Rating software applied in the USDOT study and to bridges for which these states had coded the data that the software package requires. (Two bridge types, truss and girder-floor-beam bridges, which the software cannot analyze, were assessed by a different method.) The procedure for selecting the 490-bridge sample creates the possibility that it is biased. The states that use the AASHTOWare package may not be representative of all states in their bridge management and maintenance procedures. Also, the bridges for which a state has data available in the model format may not be representative of all bridges of their structural type in the state (e.g., they might be bridges for which the required data were most easily obtained). The report explains that the sample size of 490 was selected according to a calculation of the sample size required to achieve a desired confidence level in estimates derived from the sample (Bridge, 16). However, because the sample was not randomly selected, this calculation is inapplicable and gives no information about the confidence level of the bridge cost estimates. An approach to checking the validity of the bridge sample would have been to repeat the 2000 USDOT bridge analysis method (which used a simpler criterion requiring fewer data items for determining which bridges would be overstressed by alternative configurations) for a random sample of bridges from the National Bridge Inventory and for the 490-bridge sample of the 2015 study. Similar failure rates for bridges of the same type in the two samples would be evidence in support of the validity of the 490-bridge sample. Another check on the validity of the sample would have been to compare the distribution of square feet of bridge deck by bridge type in the sample with the distribution for the U.S. population of bridges.

29 Unit Costs The bridge replacement and strengthening cost estimates in the USDOT study assume an average cost of $235 per square foot, derived from records of federal-aid project costs. The report does not describe the derivation of this cost, the variability in costs from project to project, or the mix of strengthening projects and replacement projects in the data from which it was derived. Although the committee did not attempt any estimate of its own, the reaction of some committee members was that the value appears very low. Truck Weight Assumptions The structural analysis compares the number of bridges that fail the rating criterion when exposed to the control vehicle loaded to 80,000 pounds (for the tractor-semitrailer control vehicle) or 71,700 pounds (for the double-trailer control vehicle) with the number that would fail when exposed to the alternative configuration loaded to the maximum weight defined in each scenario. The increase in the number of bridges failing is attributed to the change in limits, and the cost of replacing or retrofitting them is considered a cost of the change in limits. However, virtually all of the sample bridges are exposed today, with some regularity, to loads heavier than the assumed loads of the control vehicles, from trucks operating under special permits or from grandfathered trucks (not considering illegal overloads). The bridge structural cost of the alternative configuration scenarios, estimated according to the procedure of the USDOT study, will vary greatly depending on whether the base case is defined as existing legal nonpermit loads, existing legal permit loads, or existing illegal overloads. In other words, the bridge cost estimates in the USDOT study are highly sensitive to arbitrary assumptions. The assumption in the bridge impact estimates (and in the vehicle handling and stability simulations) that all vehicles are loaded to their maximum legal weight except the control double vehicle, for which the assumed weight is 71,700 pounds rather than the legal maximum 80,000 pounds, may be

30 appropriate for the purpose of the study. However, the report does not adequately explain the inconsistency in the weight assumptions. A more transparent procedure would have been to conduct the analyses at two assumed weights of the control double (71,700 and 80,000 pounds) to show the consequences of the assumption. Recommendations The final report of the USDOT study should provide a framework to explain the significance of its bridge cost estimates to the nonspecialist reader. The framework should identify all relevant categories of costs and present estimates of each category from past studies as well as the present study in consistent units. Any future USDOT truck size and weight study should define such a framework. Bridge cost estimates should include the effect of changes in truck traffic on all bridge components and all forms of impact that affect bridge performance. The addendum to this chapter outlines practical approaches for filling the major gaps in the present study to allow comprehensive estimates. FHWA is ideally situated to champion research on the infrastructure investment needed to increase highway transport efficiency. Bridge cost estimates in any future study should be based on explicit assumptions about state highway agency responses to changes in truck traffic. These assumptions could be projections of likely state responses based on interviews with states and review of historical state bridge policies and actions. Analysis of the optimal highway agency response according to engineering economic principles also would be valuable in a future study. The bridge analysis in any future study should include cost estimates derived from evaluation of a scientifically designed sample of all U.S. bridges.

31 Addendum to Chapter 3 Ideas are offered below on an approach for a practical but comprehensive study of bridge costs of changes in size and weight limits. The largest costs will arise from three impacts:  Deck degradation leading to deck repair or replacement,  Loads exceeding the strength of main load-carrying members, and  Increased cost of future bridges. Deck cost projections are essential for alternative configuration scenarios that involve higher maximum or average axle weights or that include a triple axle configuration. It is reasonable to expect that the need for deck replacement will be accelerated by heavier axle loads (i.e., that increased loads will bring some bridge decks to the threshold of needing immediate action). The National Bridge Inventory database provides deck sufficiency ratings and contains all bridge lengths and average widths. States have data on their cost per square foot for deck replacement. With this information and an assumption on a cutoff sufficiency rating, an order-of-magnitude estimate of deck costs can be made and added to that for strengthening bridges. Cost impacts of weak girders can be determined without a model of each bridge. The shear and bending moment demands due to each of the scenario vehicles need to be calculated for all practical span lengths. These values can be compared with those for current legal loads. Each state has cost data on bridge strengthenings and replacements. Replacement costs should be used for bridge types such as precast prestressed girder bridges, trusses, and slab bridges that cannot be strengthened. Steel girder bridges can be strengthened in most cases. Finally, projects are not completed instantaneously. Costs must be projected over 5 to 10 years and delay costs added for the time a bridge is posted.

32 Future bridges may need deeper cross sections to carry an alternative configuration. A modified table of depth-to-span ratios for each structure type would need to be developed for each of the scenario vehicles. Construction cost averages in each state for each bridge type could then be used to project a percent increase in new bridge cost. References Abbreviations TRB Transportation Research Board USDOT U.S. Department of Transportation TRB. 1990a. Special Report 225: Truck Weight Limits: Issues and Options. National Research Council, Washington, D.C. TRB. 1990b. Special Report 227: New Trucks for Greater Productivity and Less Road Wear: An Evaluation of the Turner Proposal. National Research Council, Washington, D.C. TRB. 2002. Special Report 267: Regulation of Weights, Lengths, and Widths of Commercial Motor Vehicles. National Academies, Washington, D.C. USDOT. 2000. The U.S. Department of Transportation’s Comprehensive Truck Size and Weight Study: Volume III: Scenario Analysis. Aug.