Value Pricing Pilot Program References

Printable Version (PDF 70 KB)

TRUCE Model - Interactive Spreadsheet (Excel, 72KB)
To view Excel files, you can use the Microsoft Excel Viewer.

TRUCE 2.0 - Users' Guide

Introduction

When freeway traffic flow collapses under high traffic densities, not only do speeds drop, but highway capacity is also lost (1). By preventing congestion from taking hold, congestion pricing of freeways can recover the daily waste of public investment that occurs on congested highways when traffic flow breaks down. Also, as noted by Krusee (2), a financing system based on direct user charges can be fairer to motorists than the current fuel-tax based system, since charges are better aligned with costs imposed on the system by location and time of travel.

The high performance highway concept (3) involves conversion of all lanes on existing freeways during peak periods only into premium-service free-flowing highways that provide fast, frequent and inexpensive express bus service, and charges all private vehicles a variable toll -- except for authorized buses. The toll would vary by level of demand and would be set high enough to guarantee that high demand will not cause a breakdown of traffic flow. Transit, carpool and park-and-ride services, and employer-based flextime and telecommuting programs, would complement congestion pricing to provide additional travel choices for those may not be willing or able to pay the market-based toll rates.

The ability of active freeway traffic management to increase vehicle throughput at a network level was demonstrated in the fall of 2000 in the Twin Cities, MN (4). A ramp meter study was conducted in the fall of 2000. Ramp meters were shut down for a period of 5 weeks beginning in October 2000. After the meters went off, there was an average reduction of daily traffic volume of 9% on freeways. During peak traffic conditions, freeway mainline throughput declined by an average of 14% in the meters-off condition.

A sketch-planning tool has been developed to assist in evaluation of the potential of high performance highway concept in specific metropolitan areas. This sketch-planning tool, called Tool for Rush-hour User Charge Evaluation, i.e., TRUCE, has been developed by Patrick DeCorla-Souza, FHWA, to assist transportation professionals in the "ball park" estimation of the potential impacts of a high performance highway facility or network, in particular the costs, benefits and revenues from operation of a high performance highway network (3).

Key Model Inputs

The spreadsheet model requires a minimal amount of data inputs, and provides ball-park estimates of toll rates, revenues, costs, and benefits from implementing the high performance network concept in a sub-area within a metropolitan region, or the entire metropolitan area. There are five spreadsheet tables that are included in the model. Key information that is needed as input is noted in the last column of the spreadsheet tables. Many of the existing parameters used in the spreadsheet may be used as defaults, if the user does not have access to specific data for his or her region.

The chief inputs required are:

Average peak period freeway speed, which can be derived from the Travel Time Index (TTI) for the area, and the freeway free-flow speed.
Average number of freeway lanes (both directions).
Average number of hours of freeway congestion.
Lane miles of the freeway network to be priced.
Number of gantry units (total) on priced freeway network, or average spacing intervals for toll gantries, needed for tolling technology cost.

Model Structure

The structure of the model is presented in Figure 1. The figure provides references to tables in the spreadsheet that include each model component. The tables illustrate the model calculations for the Los Angeles metropolitan area. The rest of this model documentation discusses the procedures and assumptions used in each model component.

For a quick and simple analysis, the user may use as input the Travel Time Index or average peak period speed best representing the full network to be priced. For a more detailed analysis, the user may break down the priced network into several levels of congestion and perform the analysis separately for each portion of the network. Data on number of lane miles and duration of various levels of congestion would be needed, and the spreadsheet calculations will need to be repeated for each level of congestion, by creating separate columns for each congestion level, and summing the results. The user would, in this case, obtain the summarized results in the last column of the last spreadsheet table (Table 8).

Step 1: Benefits to Toll-paying Motorists

TRUCE begins with estimation of the average travel time that would be saved on a trip that uses a 10-mile segment of the freeway network. Input data for 2003 may be obtained from the Texas Transportation Institute's (TTI) Urban Mobility Study (8). Inputs include the peak-period "travel time index" (8), which, along with a free flow freeway speed of 60 mph, is used to calculate average network speeds during peak periods. Travel time savings are converted into monetary values based on the average value of time per hour per person. Although generally not perceived by motorists, delay reductions also result in significant fuel consumption savings, due to fewer accelerations and braking events. Estimates of fuel consumption savings are based on estimates of fuel saved per minute of delay reduced.

Figure 1. Structure of the TRUCE Model D

Value of time: Travel time savings are converted into monetary values, based on average value of time per hour per person. If local data is not available, possible sources for input are the values used in the TTI study or those recommended by U.S. DOT (15), adjusted for inflation and average vehicle occupancy.

Fuel consumption rates: Delay reductions also result in fuel consumption savings, due to fewer accelerations and braking events. To be conservative in estimating these benefits, default estimates of fuel consumption savings are based on estimates of fuel saved by a small car per minute of delay reduced, as documented in the American Association of State Highway and Transportation Official's (AASHTO) User Benefit Analysis for Highways Manual (9).

User cost savings: Table 2 presents user cost savings per freeway trip for those paying the toll. Net user cost savings per freeway trip are estimated by subtracting the toll cost from the monetary value of time and fuel cost savings. For multi-occupant vehicles that pay tolls, benefits are underestimated because it is conservatively assumed that the total value of time savings of all occupants will be equal to that of a single occupant whose value of time is equal to the average value. On the other hand, for carpoolers who may have previously used HOV lanes on the freeway, the estimated benefits may be more reasonable, reflecting benefits to such former HOV lane users because: (a) speeds on HOV lanes that are not barrier separated are often lower than the speed limit due to "friction" from adjacent slower-moving GP lanes, while high speeds are guaranteed on high performance highways; (b) the HOV lanes themselves may be congested, while high performance highways manage traffic for free flow; and (c) carpools are faced with extra time and inconvenience in merging into and out of HOV lanes, while such merges are unnecessary on high performance highways.

Average peak period toll rate: For the purpose of estimating the average peak period toll rate, it is assumed that:

In deciding whether to pay the toll, motorists would consider how much delay they would incur on an alternative toll-free route and compare the equivalent monetary cost of that delay to the going toll rate.
A current freeway motorist who wants to avoid the toll by taking an alternative toll-free route would face a travel time 25% higher than the "base" congested travel time on the freeway, i.e., travel time prior to introduction of pricing.
Trucks would pay toll rates that reflect their relative passenger car equivalents.
Of those motorists who decide to pay the full toll, the solo driver who values his or her time the least would have a value of time equal to the average value of time for all travelers.
The average monetary value of the delay time on alternative routes would be in equilibrium with the toll rate in effect.

Assumption for truck toll rates: It is assumed that trucks would pay toll rates that reflect their relative passenger car equivalents. Since a heavy truck on average consumes two to three times the lane capacity of a passenger car in free-flowing traffic, toll rates for trucks would average about 2.5 times the toll rates for passenger cars. Since TRUCE analyzes all traffic in terms of passenger car equivalents, separate estimates for trucks are not produced.

Step 2: Benefits on a 10-mile Highway Segment

Table 3 provides calculations of daily estimates of highway benefits and toll revenues for a 10-mile highway segment. Inputs include total daily freeway vehicle miles of travel (VMT), which is used to calculate peak period VMT in the number of rush hours used by TTI study, based on the share of travel in the AM and PM peak periods from the 2001 National Household Travel Survey (7). The number of freeway lane miles is used to calculate the average hourly traffic volume per lane over the 8-hour period, for both directions. Existing peak period demand for freeway use is equal to the total vehicle volume that currently uses the freeway during the congested 8-hour peak period. The existing hourly peak period traffic volume per lane accounts for both lost throughput in the heavy traffic direction, as well as lower vehicle volumes in the reverse direction.

Hours of congestion: In 2003, the average daily congested travel period in major U.S. metropolitan areas amounted to about six and one-half hours, and percent of daily travel that occurred under congested conditions amounted to about 40 percent (8). If TTI study data is used as input, the TTI estimates may be used as input.

Percent of traffic reduction: Tolls would need be just high enough to deter a few motorists from driving on the priced highways during key times during congested periods so that traffic flow does not break down. The Highway Capacity Manual (16) indicates that a freeway remains in stable flow until about 85% of its maximum possible traffic volume is achieved. Due to the fact that hourly throughput at bottleneck locations on freeways is much lower than this maximum volume (1, 4), the default assumption is that currently congested freeways would need a reduction of only 10% of their average peak period traffic volumes for stable flow to be guaranteed.

This level of auto demand reduction may be achieved as follows:

Due to significant reductions in transit travel time resulting from free-flowing traffic conditions, about 2% of total existing peak period travelers may shift to use of express buses. This is consistent with cross-elasticity estimates (i.e., the % change in auto use resulting from a 1% change in transit user costs) ranging from 0.025 to 0.056 (5), assuming a 50% reduction in transit travel time due to the use of free-flowing freeways by express buses.

Use of multi-occupant vehicles may result in a 6% increase in average vehicle occupancy (AVO) for autos, carpools and vanpools, from 1.10 to 1.17. This 6% increase in AVO is less than half the average AVO increase of 14% observed for 10 HOV lane projects implemented in the U.S. (6) - and is consistent with the 50% toll cost advantage that a 2-person carpool would have over a solo-driver.

An additional 2% of previous solo drivers may choose to telecommute or travel at other times of the day. Given the potential of teleworking, and National Household Travel Survey data indicating that 10% to 23% of peak period trips are made solely to shop (7), this is a plausible assumption.

Ratio of travel time on alternative route to travel time on congested freeway: TRUCE assumes that a current freeway motorist who wants to avoid the toll by driving on alternative toll-free routes would face a travel time that exceeds the "base" congested travel time on the freeway (i.e., prior to introduction of pricing) by 25%. Note, however, that travel models generally assume that travel times on alternative routes are equal on congested networks, due to equilibration, so at first glance one might assume that travel time on an alternative route during rush hours should not be much higher than on a freeway. The 25% higher travel time assumed as a default in TRUCE for toll-free routes accounts for the possibility that some motorists for whom the freeway is the most direct route may have to drive far out of their way to use a toll-free arterial. The TRUCE model user may adjust this parameter if local conditions make it necessary.

Estimation of daily benefits: Social benefits are estimated as follows:

Net user cost savings on the priced freeway. Benefits to those who continue to travel on the freeway are estimated based on the benefits per vehicle trip calculated in Table 2. Motorists with lower values of time would perceive disbenefits if they had to pay a toll, and respond to new congestion tolls by diverting to other modes, routes, or times of the day. Losses of consumer surplus by motorists who shift from driving alone on the freeway are estimated based on the rule of half, i.e., number of diverted motorists times half the difference between: (a) monetary value of motorists' travel time plus toll cost on priced freeways; and (b) monetary value of motorists' travel time cost prior to pricing. Given typical observed distributions of values of time of motorists (10), it is reasonable to assume that the 10% of all motorists who shift from driving alone in the peak periods on the freeway would have a value of time equal to about 50% of the average value of time.

Toll revenue "transfers" from motorists to the system operator. (Note that tolls paid by motorists are subtracted in computing net user cost savings above).

Changes in government fuel tax receipts due to changes fuel consumption are estimated to be zero since the fuel cost savings estimated to compute user cost savings above does not include fuel taxes based on an assumption that congestion pricing revenues will replace fuel taxes.

Several components of social benefits are not included in the above social benefit calculations:

Benefits from an increase in trip time reliability. With more predictable trip times, travelers will be able to reduce the "buffer" time that they build into their schedules. Surveys of travelers who use priced lanes in San Diego and in Orange County, California, suggest that travelers perceive that they save almost twice the amount of time that they actually save. This may simply reflect a reduction in the amount of "buffer" time that they allocate for their trips, due to the reliability of their trip times. Travel time reliability benefits are estimated in Step 11 (discussed later).
Environmental, health and safety benefits, such as reductions in air pollution, noise and greenhouse gas emissions, and accident cost reductions. Environmental benefits are expected to be positive, since mode shifts will reduce vehicle traffic, and higher traffic speeds will reduce most vehicle emissions. Shortening of response times for emergency personnel may save lives. With reduced traffic, the number of accidents would also be reduced; however, severity of accidents could increase due to higher speeds, raising the average cost per accident. Reduction of the stress of freeway driving will reduce negative health impacts and improve the quality of life of drivers.
Impacts of traffic diversion on congestion levels on parallel toll-free routes: Modal and time of travel choices and the availability of low-income driver toll discounts are expected to limit traffic diversion, so any negative impacts are expected to be minor, and positive impacts may occur due to increased vehicle and person throughput in the freeway corridor (3).
Benefits to businesses and the economy, including productivity benefits from reduced freight delays and increased reliability of deliveries, and reduced distortions in the housing market.
Increase in energy security, due to reduced fuel consumption.

Step 3: Annual Benefits on Highway Network

As shown in Table 4, daily benefits and toll revenues for a 10-mile highway segment are annualized based on an annualization factor of 330 (to account for lower existing congestion levels, and therefore lower benefits, on Saturdays and Sundays) and extrapolated to the entire highway network based on number of lane miles. Key performance measures, i.e., change in delay and fuel consumption per peak period traveler, are also estimated. TTI study data is used to get number of existing peak period travelers. To allow comparison of toll revenues with fuel tax receipts from the current system of financing, total fuel tax receipts from each region are also estimated, based on a total State and Federal fuel tax of 40 cents per gallon and average fuel economy of 20 mpg, which results in an average fuel tax per VMT of 2 cents.

Step 4: Estimate Daily Transit Benefits

Table 5 presents estimates of transit benefits. Travel time savings for existing bus passengers are assumed to be equivalent to those accruing to motorists. Operating cost savings for existing bus services are computed by combining driver time savings and bus fuel cost savings.

Fuel consumption: Fuel cost savings are based on AASHTO estimates of fuel consumption per minute of delay for a single-unit truck (9).

Defaults for existing bus services: Existing bus service is estimated at 6 buses per hour, with 40 passengers per bus, resulting in an estimated 240 riders per rush hour in each freeway corridor. This amounts to 2% of travelers on a 6-lane freeway carrying 12,000 people per rush hour (i.e., 2,000 people per lane, based on 1,800 vehicles per lane per hour and an average vehicle occupancy of about 1.10). Where some express buses may have previously used HOV lanes on the freeway, the estimated benefits may not be as high as shown in the table. However, it can be expected that a priced highway would provide additional benefits to such express bus trips because: (a) unlike HOV lanes, high speeds are guaranteed on priced highways; and (b) buses are not faced with the extra time and inconvenience involved in merging into and out of HOV lanes. Benefits to new transit riders are estimated based on the rule of half, i.e., half of the change in travel time costs, times the estimated number of new riders.

Estimation of transit benefits: Benefits to new transit riders are estimated based on the rule of half, i.e., half of the change in travel time costs, times the estimated number of new riders.

Step 5: Estimate Annual Transit Benefits for 10-Mile Highway Segment

Since new express bus services are assumed to be provided only on weekdays, benefits for the 10-mile highway segment are annualized based on 250 working days per year (excluding holidays). See Table 5.

Step 6: Estimate Annual Transit Benefits for Highway Network

To estimate network wide benefits, average benefits per lane are estimated by dividing the output from step 5 by the number of lanes on the highway segment. The result is then extrapolated for the entire network based on number of lane miles. See Table 5.

Step 7: Estimate Highway Capital Costs

To estimate capital costs for toll collection, an open road electronic toll collection system was assumed, with toll gantries installed at 5-mile intervals. See Table 6.

Defaults for capital costs: Unit capital cost estimates were provided by Mitretek (personal communication from Paul Gonzalez in September 2006). Total capital costs are annualized based on a 7% discount rate and 30-year life.

Step 8: Estimate Highway Operating Costs

Average operating costs per vehicle trip for toll collection and traffic management combined are multiplied by number of peak period highway trips to get total highway operating costs. See Table 6.

Toll collection and traffic management operations costs: Average operating costs for toll collection are estimated at 8.5 cents per trip, based on an estimate of 5 to 10 cents per trip by ITS Decision, Service and Technologies (11). Since toll collection costs will decrease with large-scale implementation, this is a conservative estimate. In addition to toll collection costs, highway operations will involve costs for traffic management, such as operation of variable message signs, traffic monitoring equipment, and communications. Annual costs for both traffic management and toll collection on the dynamically priced I-15 HOT facility in San Diego were about $0.7 million in fiscal year 2005 (source: I-15 FasTrak budget and expenditure data for FY 2005). The facility carried about 5 million vehicles during that year, about 75 percent of them non-tolled HOVs. The remaining 25 percent were tolled vehicles. Subtracting costs for tolling (at 10 cents per trip), traffic management costs for the year are estimated at $575,000, or 11.5 cents per vehicle served. Based on these cost estimates, a total cost of 20 cents per vehicle trip was estimated for tolling and traffic management combined.

Step 9: Transit Subsidy Costs

The express bus system would need to have the capability to carry all travelers who would shift from driving on the freeway to transit, i.e., the percentage of peak period freeway demand that is expected to shift to transit, as discussed earlier. See Table 7.

Transit subsidy costs per passenger mile: Transit subsidy needs were estimated at 50 cents per passenger mile, based on nationwide subsidies of $23.5 billion supporting 50 billion passenger miles annually (12).

Transit trip length: An average bus passenger trip was estimated at 12 miles, based on work trip length data (7).

Step 10: Park-and-Ride Subsidy Costs

Most of the new park-and-ride spaces will be needed in exurban or suburban locations. At these locations, it is more likely that a public agency will own land within existing rights-of-way near interchanges or along the freeway. It may therefore be possible to build new park-and-ride facilities on surface lots, adjacent to express bus stations. Also, it may be possible to use existing parking spaces at shopping centers near the freeway, reducing new construction costs. Parking costs are estimated by multiplying cost per parking space per day, by the number of transit round-trips (i.e., half of total transit trips). See Table 7.

Park-and-ride costs: Parking costs are estimated at $2.00 per parking space per day, based on annualized costs for construction and maintenance of surface parking spaces in outer suburbs US DOT's Characteristics of Urban Transportation Systems (13), adjusted for inflation.

Step 11: Assess Net Benefits and Financial Feasibility

Table 8 summarizes estimates of toll revenues, benefits and costs of the multi-modal pricing package from the 10 prior model steps for all five metro areas. Based on a UK study (14), highway benefits based only on travel time and fuel savings may be increased by 20% to account for benefits of trip time reliability. These benefits are added in Table 8.

REFERENCES

1. Chen, Chao and Varaiya, Pravin. The Freeway-Congestion Paradox. Access. Number 20, Spring 2002: pp 40-41.

2. Krusee, Mike. A User Fee is Better Than A Tax. Tollways. Autumn 2006. Journal of the International Bridge, Tunnel and Turnpike Association.

3. DeCorla-Souza, Patrick. High Performance Highways. Public Roads. May/June 2007. Vol.70, No. 6. Federal Highway Administration.

4. Cambridge Systematics, Inc. Mn/DOT Ramp Meter Evaluation. Minnesota Department of transportation, May 10, 2002.

5. Glaister, Stephen and Lewis, David (1978). An Integrated Fares Policy for Transport in London. Journal of Public Economics, Vol. 9.

6. Richard H. Pratt, Consultant, Inc. et al (2000). Traveler Response to Transportation System Changes, Interim Handbook. TCRP Web Document 12, Chapter 2, HOV Facilities, Table 2-22 on page 2-68. Available at: http://www.trb.org/trbnet/projectdisplay.asp?projectid=103315.

7. U.S. Department of Transportation. 2004. National Household Travel Survey 2001. Available at: http://www.bts.gov/publications/highlights_of_the_2001_national_household_travel_survey/html/table_a12.html

8. Texas Transportation Institute. 2005 Urban Mobility Study. Texas A & M University, College Station, TX. 2005.

9. ECONorthwest and Parsons Brinckerhoff Quade & Douglas, Inc. TCRP Report 78: Estimating the Benefits and Costs of Public Transit Projects: A Guidebook for Practitioners. Transportation Research Board, Washington DC, 2003. Available on the web at: http://gulliver.trb.org/publications/tcrp/tcrp78/index.htm

10. Steimetz, Seiji and David Brownstone. Estimating Commuters' 'Value of Time' with Noisy Data: a Multiple Imputation Approach, Transportation Research Part B (39), 865-889, 2005.

11. ITS Decision, Service and Technologies. Web site accessed May 6, 2005: http://www.calccit.org/itsdecision/serv_and_tech/Electronic_toll_collection/electron_toll_collection_report.html

12. Taylor, Jerry and Peter VanDoren. Pricing the Fast Lane. Washington Post, July 12, 2002, page A21.

13. U.S. Department of Transportation. Characteristics of Urban Transportation Systems. Publication Number DOT-T-93-07. Revised Edition. September 1992.

14. Department for Transport, UK. Feasibility Study of Road Pricing in the U.K. 2004.

15. U.S. Department of Transportation. 2002. Memorandum on revised departmental guidance for valuation of travel time in economic analysis. Washington, DC. 2002. Available at: http://ostpxweb.dot.gov/policy/safety/VOT_Guidance_Revision_1.pdf.

16. Transportation Research Board. Highway Capacity Manual. 2000.

You will need the Adobe Acrobat Reader to view the PDFs on this page.