U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-04-142
Date: December 2005
Enhanced Night Visibility Series, Volume XI: Phase II—Cost-Benefit Analysis
PDF Version (443 KB)
PDF files can be viewed with the Acrobat® Reader®
CHAPTER 3—COST ESTIMATION
One facet of the cost-benefit analysis encompasses the cost of the VESs and the fluorescent materials, as well as the cost of any changes to required miscellaneous equipment (e.g., headlamp ballast, paint trucks) that accompany these technologies.
The pavement marking materials were monitored for changes over time in retroreflectivity, apparent color, and (in the case of the fluorescent materials) fluorescent efficiency. Table 2 lists the estimated service life of each experimental material and the historical average service life of the conventional materials used in Virginia.
The service lives evaluated in the cost-benefit study are estimates supplied by the manufacturers. The contractor paid its supplier $1,300 each for the UV–A headlamps and ballasts that were installed on the experimental vehicles used for the Smart Road testing. The unit cost of the hybrid UV–A headlamps (provided by Visteon®) was not available. According to Consumer Guide®, the infrared (IR) night vision system in the 2002 Cadillac® DeVille® DHS had an invoice price of $1,913 and a suggested retail price of $2,250.(9) A Cadillac dealer in Charlottesville, VA, quoted a retail price of $2,895 for the infrared night vision system in the 2003 DeVille DHS.(10)
In fall 1999, three different types of pavement markings were installed on the Smart Road. Two of the three types of markings, a hydrocarbon-resin-based thermoplastic (Cleanosol®) and a latex-based paint (Mercalin®), had fluorescent properties. The third, a polyurea binder system (3M® Liquid System 1200TM), served as a nonfluorescent control material. Pavement marking costs were from $0.0645 to $2.3476 per linear foot (lf) for the fluorescent paints and thermoplastics that were used in the field tests. A representative of the supplier provided an estimate of the cost of the polyurea binder system. Table 3 through table 11 itemize the cost per linear foot (lf) of each pavement marking material that was used in the Smart Road tests, plus the average cost per linear foot of the conventional thermoplastics and paints that the Virginia Department of Transportation (VDOT) uses in the field. Materials applied and tested on the Smart Road are marked with an asterisk. Cost figures used in the cost-benefit analysis are marked with two asterisks. Other materials and cost figures are unmarked. Delineator post unit costs are expressed in dollars per post. All other unit costs are expressed in dollars per linear foot; 1 lf = 0.305 linear meter (lm).
It was found that the labor and equipment cost per linear foot of installing a fluorescent marking material is identical to the cost of installing its nonfluorescent counterpart.
In addition to pavement markings—fluorescent and conventional nonfluorescent—delineator posts were installed along the side of the roadway. Their costs are tabulated in table 6 and table 11.(11) Because the delineator posts played no part in the visibility tests, their costs are excluded from the cost-benefit analysis.
The Federal Highway Administration’s (FHWA) annual Highway Statistics supplies historical tallies of motor vehicle registrations, centerline miles of road, and lane miles of road.(12,13) This information was retrieved from 1990–1998. These data help permit a forecast of the total cost of implementing any of the VESs or pavement markings.
The incremental cost of the UV–A technology takes into account the cost of the UV–A headlamps themselves, the cost of their installation, and the cost of modifications to the vehicle’s body and electrical system to fit the headlamps and power them. The differential cost of producing new vehicles that are designed to operate UV–A headlamps would possibly be less than the cost of retrofitting existing vehicle models.
It is quite possible that the unit cost of UV–A headlamps in mass production will differ from the concessionary prices paid for the experimental equipment; however, spokespersons for the auto industry were hesitant to forecast the unit cost of the headlamps in mass production. The cost analysis computes the incremental cost of each headlamp technology on the assumption that the steady-state cost of equipping a new automobile with such headlamps would be equal to the prices the contractor has on record.
The average replacement age of the headlamps is assumed to be 8 years, matching the assumption made in the previous FHWA evaluation of UV–A headlamps.(1) The average replacement age of the thermal imaging system is assumed to be the same.
A computation of the incremental cost of a fluorescent technology takes in the up-front cost of the fluorescent materials themselves, the differential cost of their installation, and the difference in the length of the replacement cycle (i.e., the service life). Each of the fluorescent materials tested was a fluorescent variant of a marking material already in use, namely thermoplastic or paint. The nonfluorescent control was a polyurea binder system. The standard nonfluorescent paint that VDOT uses is less expensive than any of the three alternatives that underwent sight-distance tests on the Smart Road. Because the technology for applying a given type of pavement marking is largely independent of the material’s fluorescent properties, the cost-benefit analysis assumes no differential installation cost (i.e., that the installation cost of fluorescent paint is no different from the installation cost of conventional paint, and that the installation cost of fluorescent thermoplastic is no different from the installation cost of conventional thermoplastic).
As in the case of the headlamps, authoritative estimates of the unit cost of each fluorescent material in mass production were not available. The cost-benefit analysis computes the annualized costs of each pavement marking technology on the assumption that the steady-state cost of procuring fluorescent pavement markings would equal the prices that were actually paid. The average replacement age of the thermoplastic, either fluorescent or nonfluorescent, is taken to be 3 years. The average replacement age of the paint, either fluorescent or nonfluorescent, is taken to be 1 year. Both of these assumptions reflect VDOT experience with the conventional nonfluorescent products.
A simple log linear equation as shown in figure 3—where y is the forecast quantity, x is the year, and m and b are constants—was fitted to 9 years of annual FHWA statistics on the number of motor vehicle registrations in the United States 1990 through 1998.(12,13) The equation is used to create a 20-year forward forecast of motor vehicle registrations. Figure 4 compares the actual number of motor vehicle registrations in each year with the number implied by the fitted equation:
The forecast of registered motor vehicles measures the size of the market that a new VES would have to penetrate. The reported results are given on the assumption that this penetration would occur over a 20-year period in equal 5 percent increments from the first year until 100 percent implementation was achieved. The annualized cost of a given VES is applied to the number of equipped vehicles forecast in each future year to yield a total cost estimate for that year.
The number of stripes needed to mark a given segment of highway is assumed to equal the number of lanes plus one. Under this assumption, the sum of the number of highway centerline miles plus the number of highway lane miles equals the number of miles of striping that would need to be placed on the Nation’s roads.
The simple log linear equation in figure 3 was fitted to 9 years of annual FHWA statistics on the number of highway (centerline) miles of road in the United States from 1990 through 1998.(12,13) The same equation was fitted to 9 years of data on the number of lane-miles of road.(12,13)
Figure 5 compares the actual number of centerline miles in each year with the number implied by the fitted equation. Figure 6 does the same for lane miles on rural roads. Each equation is used to create a 20-year forward forecast of the time series to which it was fitted. The forecast of each quantity is independent of the forecasts of the other.
It is assumed that implementation would occur only on unlighted highway segments. Because a tally of the number of unlighted highway miles was not readily available, the number of miles of rural highway is used as a proxy. It is assumed further that installation of a new marking system would occur over 20 years, in equal 5 percent increments from the first year until 100 percent implementation is achieved. The annualized cost of a given marking material is applied to the number of retrofitted miles of stripe forecast for a future year to yield a total cost for that year.
A case could be made that the graph of market penetration over time should be S-shaped, reflecting hesitant initial adoption, followed by a boom of installation that tails off as the number of unequipped vehicles and highway miles asymptotically approaches 0 percent. Given the results from the Smart Road field tests, it is not conceivable that such a modification would alter the cost-benefit findings.
A case could also be made that a cost computation based on installation cost rather than annualized cost would reflect better the time path of the costs, especially during the early years of implementation. Again, the cost-benefit findings are not sensitive to such a modification. The reported results are based on a present-value calculation using annualized costs.