Guide for Pavement-Type Selection (2011)

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.

13 4.1 Overview LCCA is an integral part of the pavement-type selection process. While pavement-type selection based solely on ini- tial costs allows for more to be accomplished with a specified annual budget, it does not account for long-term costs paid by taxpayers and facility users. Considering the time value of money, the initial and future costs occurring at different points in the life cycle of various alternatives should be compared to determine their cost-effectiveness. In this manner, the LCCA helps to evaluate the overall long-term economic efficiency of competing pavement alternatives and provides an economic basis for optimum strategy selection. The quality of LCCA results is only as good as the quality of the inputs. Therefore, all available, applicable, and reliable data should be used in quantifying the many LCCA inputs, although experience-based estimates can be used in this effort. The LCCA procedure involves establishing the LCCA frame- work, estimating initial and future costs, computing life-cycle costs, and analyzing/interpreting the results. Figure 8 presents a flow chart of this procedure. 4.2 Establish LCCA Framework This section discusses the fundamental economic indica- tors required for establishing the LCCA model. Additional information on economic indicators can be obtained from the Federal Highway Administration’s (FHWA) most current guidance on LCCA. 4.2.1 Analysis Period The life-cycle analysis period must be sufficiently long to dis- tinguish any differences in the cost-effectiveness of pavement alternatives and long enough such that each alternative pave- ment strategy includes at least one future major rehabilitation event. For new/reconstruction projects, an analysis period of at least 40 years is suggested. For rehabilitation projects, an analysis period of at least 30 years is suggested. Longer analy- sis periods may be warranted for long-life pavement designs, and greater effort must be made to make reliable long-term forecasts. 4.2.2 Discount Rate The discount rate represents the real value of money over time and is used to convert future costs to present-day costs. It is approximately the difference of the interest and inflation rates. Historically discount rates are in the 3 to 5 percent range. However, it is proposed that the long-term real dis- count rate values provided in the latest edition of the Office of Management and Budget (OMB) Circular A-94, Appen- dix C, which is updated annually (currently OMB 2010), should be utilized. 4.2.3 Economic Analysis Technique The NPV is suggested for pavement-type selection applica- tions. The suggested practice is to use constant or real dollars and a real discount rate in NPV computations. 4.2.4 LCCA Computation Approach There are two basic approaches for computing life-cycle costs: deterministic and probabilistic. In deterministic LCCA, a single value is selected for each input parameter (usually the value considered most likely to occur based on historical evidence or professional experi- ence), and the selected values are used to compute a single projected life-cycle cost. No “real world” uncertainties or vari- ations are considered in this approach. The probabilistic approach takes into account the variabil- ity associated with the input parameters in LCCA. In this approach, for a given pavement strategy, sample input values C H A P T E R 4 Life-Cycle Cost Analysis

are randomly drawn from the defined frequency distributions, and the selected values are used to compute one forecasted life-cycle cost value. The sampling process is repeated hun- dreds or even thousands of times, thereby generating a “pseudo-population” of forecasted life-cycle cost values for the pavement strategy. The resulting forecasted costs are ana- lyzed and compared with the forecasted results of competing alternatives to identify the most cost-effective strategy. It is proposed that the probabilistic LCCA computation approach be used when reliable historical data exist to model one or more of the input parameters (e.g., standard devia- tions of discount rate, unit costs, pavement service life). These data can be obtained from agency files (variable bid prices, survival analysis of pavement lives to get means and standard deviations and annual discount rates over time). If such data cannot be obtained, a deterministic approach should be used. 4.3 Estimation of Initial and Future Costs The estimated life-cycle costs are used in the economic analysis to identify the most cost-effective pavement alter- native. The life-cycle costs include direct costs incurred by the agency for initial construction and future M&R activ- ities, as well as costs incurred by road users. Only the dif- ferential costs of alternatives are considered to evaluate whether substantive differences can be identified among competing alternatives. Guidance on LCCA is evolving; therefore, FHWA’s current recommendations on LCCA procedures for estimating life-cycle costs for pavement alter- natives should be used. 4.3.1 Direct/Agency Costs Direct/agency costs include the physical costs of pave- ment activities (initial construction/rehabilitation costs and future M&R costs), salvage value, and supplemental costs. Routine reactive-type maintenance costs typically are ignored because they generally are not very high and not substantially different between pavement types (Walls and Smith 1998). Physical Costs of Pavement Activities Physical cost estimates are developed by combining quanti- ties of construction pay items with their unit costs. It is impor- tant to obtain sufficient and reliable unit cost data from available sources, such as historical bid tabulations. Recent highway projects, undertaken within the last 5 to 7 years within the region, are preferred sources for historical cost data. These data often are compiled and summarized on a regular basis for project estimation purposes. Unit cost estimates can be derived from the unit price data of the lowest bid, three lowest bids, or all bids tendered on similar projects. Each average unit price must be adjusted to the present day to account for the effects of inflation, and consideration should be given to filtering out prices biased by projects that included atypically small or large quantities of a particular pay item. Using inflation-adjusted and quantity-filtered unit price data, the mean cost of each pay item, as well as key variability parameters (standard deviation, range), can be computed for use in the economic analysis. The effect of cost adjustment factors on the final cost of the pavement should be evaluated when bid tabula- tions are used. Cost-based estimates can be used in situations where the historical bid-based estimates are not available or defendable. Examples of such situations include projects with unique characteristics, new materials/technologies, geographical influ- ences, market factors, and the volatility of material prices. A combination of cost-based and historical bid-based estimates also can be used. Pavement industry groups can be consulted to help identify appropriate data sources for cost-based esti- mates. Engineering judgment must be applied wherever necessary. The time from completion of estimate to bid advertisement should be as short as possible to reflect market prices at the time of construction, while allowing sufficient time for internal review. If necessary, the cost estimates should be updated before bid advertisement. 14 Figure 8. Evaluation of life-cycle costs. Estimate direct agency costs Estimate road user costs Historic bid tabulations/ actual costs Compute NPV of agency costs LCCA framework Pavement life-cycle strategies (see Figure 4) Next step: see Figure 15 Determine LCCA parameters Compute NPV of road user costs

Salvage Value Salvage value is the estimated value of a pavement asset at the end of the analysis period. It includes two components: • Remaining service life. The structural life remaining in the pavement at the end of the analysis period. • Residual value. The value of the in-place pavement materi- als at the end of their service lives less the cost to remove and process the materials for reuse. Salvage values are used in the economic analysis, as the in- service performance of different alternatives depreciate at different rates. The end of the analysis period very often does not coincide with the end of the alternative’s service life. On the other hand, the residual values are different for compet- ing alternatives but not very large; when discounted over the analysis period, the residual values generally have little effect on the NPV. One method of determining the value of a pavement’s remaining life is to determine the depreciated value (at the end of the analysis period) of the costs of initial construction and subsequent M&R. Depreciation is an accounting term used to attribute costs across the life of the asset. Straight-line depreci- ation is the simplest and most commonly used technique for estimating salvage value at the end of the analysis period. Depreciation can be applied to both the structural and func- tional life components of a pavement (see Figure 9). A func- tional improvement cost relates to those treatments that do not add structural capacity. Typically, this includes preventive and corrective maintenance and improvements to the pavement ride, such as surface treatments, thin overlays, and localized mill-and-fill treatments. In computing the depreciation of functional treatments, the life of the functional treatment can- not exceed the structural life of the pavement. The structural remaining life of the pavement can be deter- mined by several methods. One method is to compare the design traffic loads to the cumulative traffic to date, with the difference between these values representing the remain- ing life of the pavement. Another method is to perform struc- tural capacity testing and compare the measured value with structural design limiting criteria to determine the remaining life of the pavement. Supplemental Costs Supplemental costs, applicable to anticipated future M&R events, can be grouped into three categories: • Administrative costs. Contract management and adminis- trative overhead costs. • Engineering costs. Design and construction engineering costs, construction supervision costs, and materials testing and analysis costs. • Traffic control costs. Traffic control setup and communi- cations costs. If these costs are approximately the same for different alter- natives, then these costs can be ignored. Because estimating these costs can be difficult and time-consuming, an alternative method to consider is to specify them as a percentage of the total project-level pavement costs. 4.3.2 Indirect/User Costs Although borne by highway users, the user costs are given serious consideration by agencies, since an agency acts as the proxy for the public. User costs are an aggregation of time delay costs, vehicle operating costs, crash costs, environmental costs, and discomfort costs associated either with work zones or any time during normal (nonrestricted) operating conditions. In many instances, since the absolute value of user costs of the project far exceeds the direct-agency costs, user costs are eval- uated independently without combining with the direct costs. 15 Figure 9. Example of depreciation curves for structural and functional improvements. Structural addition Analysis period Value of remaining life Remaining life Pavement Age, years Functional treatment Im pr o v em en t c o st $ Annual Depreciation = Cost of Improvement / Life Span

For economic analysis, only the differential user costs incurred from the use of one alternative over another are con- sidered. The suggested practice is to consider only the time delay and vehicle operating cost components associated with work zones. These components can be estimated reasonably well and constitute a large portion of the total user costs. Other user cost components are difficult to collect and quantify accurately. More detailed discussion on user costs can be found in the FHWA’s guidance on the LCCA. 4.3.3 Develop Expenditure-Stream Diagrams Expenditure-stream diagrams are graphical or tabular repre- sentations of direct-agency expenditures over time. These dia- grams help the designer/analyst visualize the magnitudes and timings of all expenditures projected for the analysis period for each alternative. As illustrated in Figure 10, an expenditure- stream diagram shows the costs and benefits associated with various activities of an alternative’s pavement life cycle on a time scale. Costs normally are depicted using upward arrows, and benefits (e.g., salvage value) are depicted using downward arrows. 4.4 Compute Life-Cycle Costs Once the expenditure-stream diagram for each alternative pavement strategy has been developed, the task of computing projected life-cycle costs is undertaken. This step combines ini- tial and future agency costs that are projected to occur at dif- ferent points in time on a comparable scale. Considering the time value of money, all future costs are converted to present values using a specified discount rate. The initial costs and all discounted future costs are then summed together to produce the NPV of agency costs. The user costs are not combined with the agency costs but are evaluated separately using the process discussed in Chapter 5. In deterministic analysis, computing life-cycle costs involves a simple application of NPV. In probabilistic analysis, the NPV is calculated after each iteration to generate an array of fore- casted costs. These costs are then analyzed and compared with the forecasted costs of other alternatives to identify the most cost-effective strategy. Probabilistic simulation requires the use of a computerized spreadsheet program equipped with the necessary probabilistic distribution functions, such as FHWA’s RealCost (Office of Assett Management 2004), or a stand-alone computer program to perform the simulation. When performing a probabilistic simulation, it is important to make sure that each iteration represents a scenario that can actually occur. Two modeling errors with the potential to cre- ate unreal scenarios are as follows (Walls and Smith 1998): • Lack of appropriate predefined relationships between input parameters. Although each randomly selected value for a given iteration may be legitimate on its own, reality may dictate that certain relationships exist between the input parameters. For example, since higher traffic volume generally is linked with shorter pavement life for a given design cross section, it is important to establish an appro- priate sampling correlation between these two inputs. Such a correlation would ensure that, for each iteration, a sam- ple from the high side of the traffic probability distribution is countered with a sample on the low side of the pavement life probability distribution, and vice versa. • Lack of fixed limits on input sampling distributions. For some types of sampling distributions, the limits for sampling are not among the criteria used to define the dis- tribution (e.g., in defining a normal sampling distribu- tion, only the mean and standard deviation are needed). However, it is important to know the minimum and max- imum values for sampling so that reasonable values are used in the simulation. Misleading simulation results can be expected, for instance, if the distribution for a cost or pavement service-life parameter allows negative values to be selected. 16 t1 t2 t3 t40 Time, years Salvage $$ Initial Const $$ Rehab $$ Rehab $$ Rehab $$ Maint $$ Maint $$ Maint $$ Maint $$ Maint $$ Maint $$ Maint $$ Figure 10. Example expenditure-stream diagram.

4.5 Analyze/Interpret Results Regardless of whether deterministic or probabilistic life-cycle costs are computed, the results must be analyzed and inter- preted carefully to evaluate the cost-effectiveness of a pavement strategy. Because the outputs of each computational approach are different, the ways in which they are evaluated and inter- preted also are different. 4.5.1 Analysis of Deterministic Life-Cycle Cost Results In the analysis of deterministic results, the percent differ- ence in life-cycle costs of alternatives is computed. The cost- effectiveness of alternatives is established by comparing the percent difference against some established threshold require- ment. Since the agency has to take other financial considera- tions into account, such as the cost feasibility of alternatives, available funding levels, and the impact on overall system needs, the decision to eliminate or further evaluate an alterna- tive is made after the evaluation of pertinent economic criteria. Chapter 5 discusses the evaluation of alternatives using eco- nomic factors. 4.5.2 Analysis of Probabilistic Life-Cycle Cost Results In the analysis of probabilistic results, the likelihood of an alternative’s cost-effectiveness is evaluated with those of other alternatives. This can be accomplished by risk assessment of forecasted NPV distributions. This approach involves comparing NPV distributions of dif- ferent alternatives at a specified level of probability. A proba- bility level between 75 and 85 percent will provide reliable estimates. Figure 11 shows the cumulative probability distribu- tions of NPV for two alternative strategies at 50 percent (mean value) and 75 percent probabilities. The figure indicates that Alternative B has a lower NPV than Alternative A at both prob- ability levels. Suppose that Alternative A had a slightly lower mean NPV ($1.608 million instead of $1.611 million) and a more dis- persed distribution, as shown in Figure 12. In such a case, the tails of the frequency distribution curves should be evaluated for any potential cost-associated risks. The distribution curves shown in Figure 12 indicate clear differences in the forecasted NPV at the tails. For Alternative A, there is potential for a cost underrun if the true NPV is low (say, less than $1.45 million). This opportunity for cost savings is called upside risk. If, on the other hand, the true NPV is high (say, greater than $1.75 mil- lion), there is a potential for a cost overrun associated with Alternative A. This chance for financial loss is called down- side risk. In the cumulative distributions shown in Figure 13, it can be seen that there is a 10 percent probability that the NPV of Alternative A will be less than that of Alternative B by as much as $26,000. At the other end of the spectrum, there is a 10 percent probability that Alternative A will exceed the cost of Alternative B by up to $41,000. Although many agen- cies may find this information insufficient for identifying the most cost-effective strategy, to some risk-averse agencies it may provide enough assurance that the allocated budget is best served by choosing Alternative B. In other words, there is a greater risk of the true cost of Alternative A exceeding the cost of Alternative B. 4.5.3 Reevaluate Strategies In the final step of the LCCA process, information result- ing from the LCCA is reevaluated to determine if any modi- fications to the alternative strategies are warranted, prior to making a final decision on which alternative to use. Such adjustments may entail changes to the original structure or 17 Figure 11. NPV frequency distributions for alternative strategies A and B. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1200 1300 1400 1500 1600 1700 1800 1900 2000 NPV, $1000 Alternative A Alternative B 50 percent probability 75 percent probability C um ul at iv e P ro ba bi lit y

rehabilitation treatment, revisions to the maintenance of traf- fic plans, reductions in construction periods, or changes in future M&R activities. Probabilistic sensitivity analysis can provide insight on the refinement of strategies. This technique uses correlation analy- sis and tornado plots to show the impacts of key input param- eters on life-cycle costs. Inputs found to be driving the LCCA results can be scrutinized to determine if actions can be taken to improve cost-effectiveness. Figure 14 presents an example showing the correlation coefficients of factors influencing the NPV of a pavement alternative. The correlation coefficient is a statistical measure that indicates the strength of the linear association between two variables. A correlation coefficient of +1 indicates that two vari- ables are perfectly related in a positive linear sense, while a value of −1 indicates perfect negative correlation. Values closer to zero indicate poor or no correlation, and other intermedi- ate values indicate partial correlation. In this example, the NPV of the pavement alternative is pos- itively correlated with cost factors while negatively correlated with the discount factor and pavement service-life estimates. The initial construction cost appears to be the dominating fac- tor influencing NPV, followed by the initial life of the original pavement. In other words, to reduce the NPV of this pave- ment alternative, a strategy to reduce initial construction cost would be more effective than other possible strategies. 18 0 20 40 60 80 100 120 140 160 180 200 1200 1300 1400 1500 1600 1700 1800 1900 2000 Fr eq u en cy NPV, $1000 Alternative A Alternative B Alternative A Mean = $1,611,702 Std Dev = $100,190 Alternative B Mean = $1,600,344 Std Dev = $89,996 Figure 12. Risk assessment—NPV frequency distributions. Figure 13. Risk assessment—NPV cumulative distributions. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1200 1300 1400 1500 1600 1700 1800 1900 2000 C u m u la tiv e Pr o ba bi lit y NPV, $1000 Alternative A Alternative B - $41,000 Alternative A $1,752,000 Alternative B $1,711,000 Alternative B $1,489,000 Alternative A $1,463,000 $26,000 Upside Risk (opportunity for cost savings)

19 Figure 14. Correlation coefficients of factors affecting the NPV of a particular pavement strategy. 0.12 -0.20 -0.22 0.25 -0.35 0.60 -1.00 -0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00 Correlation Coefficient Initial Const. Cost Initial Const. Life First Rehab Cost Discount Second Rehab Cost First Rehab Life