ITTS Freight Economic Analysis Tool - ittsresearch.orgittsresearch.org/feat/ITTS FEAT - Methodology_Ver 1_11302016 .pdf · reduction in shipping costs as a result of a more efficient

Novemebr 30, 2016 www.camsys.com

ITTS Freight Economic Analysis Tool

Methodology

report


Methodology

prepared for

Institute for Trade and Transportation Studies

prepared by

Cambridge Systematics, Inc. 730 Peachtree Street, NE, Suite 500 Atlanta, GA 30308

date

November 30, 2016


Cambridge Systematics, Inc. i

Table of Contents

Introduction ....................................................................................................................................................... 1

1.0 Data Requirements ............................................................................................................................. 1-1

1.1 User Provided Data .................................................................................................................... 1-1

1.2 Travel Demand Model Output .................................................................................................... 1-2

1.3 External Data ............................................................................................................................. 1-3

1.4 Summary of Data Requirements ............................................................................................... 1-4

2.0 Travel Demand Models ...................................................................................................................... 2-1

2.1 The SHIFT Model ....................................................................................................................... 2-2

3.0 Roadway Project Module ................................................................................................................... 3-1

3.1 Travel Time Savings .................................................................................................................. 3-2

3.2 Vehicle Operating Cost Savings ................................................................................................ 3-3

3.3 Safety Benefits ........................................................................................................................... 3-4

3.4 Air Quality Benefits .................................................................................................................... 3-5

4.0 Operational and Safety Improvements ............................................................................................ 4-1

4.1 Operational Only Projects .......................................................................................................... 4-1

4.2 Safety Projects Only .................................................................................................................. 4-4

4.3 Operational and Safety Projects ................................................................................................ 4-5

5.0 Freight Rail Project Module ............................................................................................................... 5-1

5.1 Shipping Cost Savings ............................................................................................................... 5-1

5.2 External Benefits of Truck Diversion .......................................................................................... 5-2

5.2.1 Congestion .................................................................................................................... 5-3

5.2.2 Roadway Wear and Tear .............................................................................................. 5-3

5.2.3 Safety Externalities ....................................................................................................... 5-4

5.2.4 Air Quality Impacts ........................................................................................................ 5-4

5.3 Benefits of Grade Crossing Elimination ..................................................................................... 5-5

5.3.1 Travel Time Savings ..................................................................................................... 5-5

5.3.2 Vehicle Operating Cost Savings ................................................................................... 5-6

5.3.3 Safety Benefits .............................................................................................................. 5-6

5.3.4 Air Quality Benefits ....................................................................................................... 5-7

5.4 Benefits of Rail Upgrades (up to 286,000 lbs.) .......................................................................... 5-7

5.5 Benefits of Double Tracking ....................................................................................................... 5-8

6.0 Pavement Conditions Impact Module .............................................................................................. 6-1

6.1 Impact of Changing Pavement Conditions ................................................................................ 6-1


Cambridge Systematics, Inc. ii

6.1.1 Pavement Condition Impacts on Vehicle Operating Costs ........................................... 6-2

6.1.2 Pavement Condition Impacts on Travel-Time Costs .................................................... 6-4

Travel-Time Delay......................................................................................................... 6-5

Value of Time ................................................................................................................ 6-6

Calculating Impacts to Travel-Time Costs .................................................................... 6-6

6.1.3 Pavement Condition Impacts on Motor Vehicle Crash Costs ....................................... 6-7

Increase in Crash Rates ............................................................................................... 6-7

Cost per Crash .............................................................................................................. 6-8

Calculating Impacts to Crash Costs.............................................................................. 6-8

6.1.4 Total Direct Impacts ...................................................................................................... 6-9

6.2 Impacts of Changing Congestion Levels ................................................................................... 6-9

6.2.1 Changes in Travel Time Costs ................................................................................... 6-10

6.2.2 Changes in Vehicle Operating Costs .......................................................................... 6-11

6.2.3 Changes in Safety Costs ............................................................................................ 6-12

6.2.4 Changes in Air Quality Costs ...................................................................................... 6-12

7.0 Logistics Costs Impact Module ........................................................................................................ 7-1

7.1 The Change in Carrier Costs ..................................................................................................... 7-1

7.2 The Change in Shipper Costs .................................................................................................... 7-3

8.0 Integration with Economic Models ................................................................................................... 8-1


Cambridge Systematics, Inc. iii

List of Tables

Table 1.1 Data Requirements ............................................................................................................... 1-5

Table 4.1 Delay per Vehicle Based on Roadway Level of Service ...................................................... 4-2

Table 4.2 Operational Project Type and Impact Level Assumptions.................................................... 4-2

Table 4.3 Default Assumptions for Operational Projects ...................................................................... 4-3

Table 4.4 Safety Project Types and Impact Assumptions .................................................................... 4-4

Table 6.1 Effects of Pavement Condition on Vehicle Operating Cost .................................................. 6-3

Table 6.2 Economic Cost per Crash by Type ....................................................................................... 6-8


Cambridge Systematics, Inc. v

List of Figures

Figure 2.1 Process of Origin-Destination Matrix Estimation .................................................................. 2-2

Figure 3.1 Benefits of Roadway Investment .......................................................................................... 3-1

Figure 5.1 Benefits of Freight Rail Investment ....................................................................................... 5-1

Figure 5.2 Benefits of Grade Crossing Elimination ................................................................................ 5-5

Figure 5.3 Benefits of Rail Upgrades ..................................................................................................... 5-8

Figure 5.4 Benefits of Double Tracking ................................................................................................. 5-9

Figure 6.1 Benefits of State of Good Repair Investment ....................................................................... 6-1

Figure 6.2 Impact of Deteriorating Transportation Infrastructure ......................................................... 6-10

Figure 7.1 Change in Logistics Costs .................................................................................................... 7-1


Cambridge Systematics, Inc. ES-1

Introduction

The methodology for estimating benefits and economic impact associated with freight roadway and railway

projects depends on the improvements that different projects can bring about. For example, capacity

expansion, operational or safety improvements, logistics improvement projects, state of good repair

investments, and railway upgrades, all generate public benefits and give rise to economic impacts from

different sources. Traditionally, economic modeling software packages are not able to capture all the benefits

and impacts from a mixed range of projects without additional front-end processes. This methodological

memorandum seeks to introduce the methodology of the Freight Economic Analysis Tool (FEAT), a flexible

framework that allows users to estimate benefits for a wide range of projects, and that facilitates the

calculation of economic impact using economic modeling packages.

The tool is based on different modules that analyze roadway and railway projects with different

characteristics. First, roadway projects that result in changes to capacity network flow are assessed using

the output from travel demand models (TDM), for example, statewide or regional travel demand models,

such as the ITTS Southern Highway Interactive Freight Traffic (SHIFT) model. Traditional benefit categories,

such as travel time savings, road safety improvements, reduction in vehicle operating costs, and air quality

benefits are calculated and used as inputs for the economic impact analysis. Smaller operational roadway

projects, such as intersection realignments, can be analyzed with a separate planning-level tool that does not

require TDM inputs. Moreover, FEAT also recognizes that some projects have an impact on logistics costs,

for example, by reducing the buffer times required to ensure timely deliveries (reliability), and the inventory

costs associated with freight transit time. Finally, it also allows estimating the benefit from improvement

pavement conditions (roadway state of good repair), which affects transportation users via higher vehicle

operating costs, travel times, and road safety.

Rail is another planning-level module of the tool, and it includes the assessment of different types of railway

projects, such as capacity expansion (including intermodal yards), grade crossing elimination, rail upgrades

(railway capacity up to 286,000 lbs.), and double tracking. The tool is based on the benefits from the

reduction in shipping costs as a result of a more efficient rail system, and on truck-to-rail diversion, which

generates positive externalities on the roadways as a result of fewer truck mileage. Moreover, the tool

includes a grade crossing elimination assessment component which is based on GradeDec methodology, a

benefit-cost tool developed by the Federal Railroad Administration.

The different FEAT modules allow the calculation of benefits to transportation system users through benefit-

cost analysis (BCA) and the calculation of the impact on the economy at large, as a result of the additional

income and jobs, using economic impact analysis (EIA). This methodology describes the data requirements,

the travel demand models, and the different modules of FEAT, outlines the benefits, and shows the

equations underlying the model benefit calculations.

Cambridge Systematics, Inc.

1-1

1.0 Data Requirements

In order to accurately quantify the benefits and costs of transportation investment, a number of data

elements must be incorporated into the model. To ensure the most accurate model results the most

recent data sources and locally or regionally specific data should be used, where available. Some of

the data inputs to the model are provided by external data sources, such as federal or state agencies.

Some of the information, particularly that information related to the specific project(s), must be

provided by the model user. Much of the analysis is driven by information taken from a travel demand

model, which is used to establish baseline level of traffic activity, and then estimate the changes

resulting from some change in the road or rail network.

1.1 User Provided Data

The user must provide information specific to the project. If actual values are not yet know, estimates

may be used. Using a range of estimates (high and low, for example) can produce a range of results,

which may be useful in the context of risk assessment.

Construction Costs – To conduct a cost benefit analysis, the cost of completing the project must

be included. Also, in the short-run, construction spending will boost local employment and

spending.

Maintenance and Operations Costs – The future cost stream necessary for maintaining an

acceptable level of service for the improvements, must be included in the costs of the project.

Project Description – Some attributes of the project must be known. These include, but may not

be limited to,

− Construction Schedule – A timeline for the construction of the project is needed to

understand the timing of the related capital expenditures and economic stimulus. The

construction schedule will also set the date from which the project will be complete and its

expected benefits will begin.

− List of Improvements – A list of the specific improvements included in the project is need in

order to make the required changes in the travel demand model, and select to the

appropriate benefit categories and model modules.

− Future Pavement Conditions – Not all projects will involve changes to pavement conditions,

but those that do will have an impact on future congestion levels, user costs, and crash rates.

Study Timeline – Transportation investment projects create a stream of benefits and costs into

the future. The aggregate of these benefits and costs will differ depending on the future time

horizon set for the analysis.


Cambridge Systematics, Inc. 1-2

Discount Rates – Based on the idea of “the time-value of money”, future benefits and costs are

not as valuable as current benefits and costs. As they move further into the future their current

value decreases. The rate at which future benefits and costs are discounted to convert them to

current values is the discount rate. Discount rates are used to set return on investment hurdles,

and a change in the rate used may make the difference between a positive and negative net

present value for a project.

Truck-Rail Diversion Rate – One of the benefits of investment in freight rail is a more efficient

rail system. While the logistics costs of shipping via rail are reduced, some freight that previously

had shipped via truck is diverted to rail. FEAT calculates total potential diversion by commodity

using commodity-flow data, and assumes that diversion occurs until the rail mode share for a

given commodity in the project area reaches the national average. The user may instead use

other diversion rate assumptions.

1.2 Travel Demand Model Output

The travel demand model is used to simulate the conditions before and after a project is completed.

Several outputs are produced for each scenario and are compared in order to understand how the

completed project will impact traffic in the study region. These are important inputs into the model,

and generally include the following, which in most cases can be provided for the study area in

aggregate or at the sub-geography level.

Vehicle Hours Traveled – The total hours of travel time (usually daily average hours) for the

users of the transportation system. Changes in vehicle hours traveled have an economic impact

that are quantified using wage rates.

Vehicle Miles Traveled – the total distance traveled (usually in daily average miles) for the

network users. Vehicle operating costs, safety costs, and emissions are all dependent on vehicle

miles traveled.

Delay – Vehicle delay (usually in daily average hours) is the difference in vehicle hours traveled

between expected traffic conditions and free-flow conditions. Delay is particularly important in

assessing the impact of improvements to rail crossings.

Trips – Vehicle trips (usually measure in a daily average count) are used to understand total

traffic volumes, and the distribution of these volumes across geography as well as time.

Occupancy – Vehicle occupancy (an average count of people per vehicle) is used in order to

accurately assess the impact of changes in vehicle hours traveled, as each occupant is impacted

by these changes.

Origin and Destination – Vehicle trips, hours traveled, miles traveled can be assigned to specific

origin-destination pairs in most travel demand models. This is of particular importance in

determining which freight trips are internal to the study area, exporting goods from the study area,

importing goods into the study area, or passing through.


1-3

Trip Purpose – Most travel demand models break up vehicle trips, hours traveled, and miles

traveled by purpose (such as leisure, commute, business, and trucking). This division enables the

model to apportion costs and benefits to the correct economic sector (such as households,

businesses, or the freight industry specifically).

1.3 External Data

Most of the data required for the translation of costs and benefits from changes in travel efficiencies

and pavement conditions are available from public sources such as federal and state agencies.

These data are readily available and are updated with regular frequency (in most cases, at least on

an annual basis).

Wage Rates – Wage rates are used to set a financial value for changes in the time spent by

users of the transportation system.

Fuel Costs – Fuel costs are needed to quantify the change in costs of user vehicle operation,

before and after the project is completed.

Non-Fuel Vehicle Operating Costs – The expenses of vehicle operation, other than fuel, must

be known in order to give a value to the change in costs born by system users resulting from

changes in aggregate mileage and/or delay.

Crash Rates – Some projects result in changes in in aggregate mileage travelled. This is

accompanied by a proportionate change in traffic accidents. Crash rates are means by which this

change is calculated. Some projects (pavement improvement projects for example) may result in

a change in the rate of traffic accidents.

Crash Values – The crash rates above can be used to calculate and expected change in the

count of traffic accidents. In order to translate these accidents to a dollar amount, prescribed

values (usually provided by federal agencies) are used.

Emissions Rates – A transportation project may impact pollution levels through changes in user

miles traveled, average speeds, or delay. These factors determine the change in fuel burned.

This increase or decrease in the amount of fuel burned translates to a change in the emission of

pollutants, which is calculated by using emissions rates.

Emissions Values – Emissions values are used to assign a dollar value to the change in

emissions caused by the project.

Shipping Costs – One of the important benefits of diversion of freight from truck to rail is the

price differential between the two modes. In order to estimate these benefits, current shipping

costs for both truck and rail must be used.



Value of Freight – In order to understand impacts to logistics costs, inventory costs must be

considered. In order to estimate how these costs will fluctuate, the value of the freight moving

though the study area must be input into the model.

Volume of Freight – The freight systems are interconnected. While some commodities are more

suited to one mode over another, many may be transported in several ways, depending on the

relative benefits of one mode over another. Information about the volume of each commodity, by

mode and origin-destination pair, is necessary to estimate the systemic effect of freight rail

improvements.

Weight per Truckload – Once the volume of diversion from truck to rail (or vice versa) is

estimated by the model, the volume must be converted into a number of truckloads, so that the

travel demand model can accurately reflect the resulting change in congestion.

Buffer Indices – Freight carriers must allow for uncertainty in turn times by allocating extra time

per turn, in case of unexpected delays. These buffer times increase as the uncertainty increases,

which increases turn times. Buffer indices are measures of buffer time, as a percentage of travel

time.

Maintenance and Operations Costs – If the user does not know the future costs of maintaining

and operating the project improvements, estimated values calculated from historical expenditures

can be used.

Pavement Conditions – Changes in pavement conditions can lead to changes in travel-times,

vehicle operating costs, crash rates, and congestion. To estimate the impact of pavement

improvement projects, accurate measures of the current pavement conditions are needed.

Length of Road Segments – Pavement conditions do not apply to the road network evenly.

Each road segment has its own condition. To calculate the impact of pavement conditions

accurately, the length of each road segment must be known.

Consumer Price Index – Not all values of the inputs will be given in dollar values from the same

year. In order to account for inflation over time the Consumer Price Index (CPI) is employed.

1.4 Summary of Data Requirements

Not all of the data inputs are needed for every project. Some only apply to specific types of projects

(for example, current and future pavement states are only required for the State of Good Repair

Module). Some information is not required but may be useful. For example, maintenance and

operations costs contribute to the accuracy of the analysis, but it can be carried out with or without

their inclusion. The following table (Table 1.1) lists the inputs for the model, their potential data

sources, and whether or not they are required for a specific module.


1-5

Table 1.1 Data Requirements

Required for Module

Description Roadway Operational and Safety

Freight Rail Logistics

State of Good Repair Potential Sources

Construction Costs Yes Yes Yes Yes Yes User

Maintenance and Operations No No No No Yes User, State Financial Records

Construction Schedule Yes Yes Yes Yes Yes User

List of Improvements Yes Yes Yes Yes Yes User

Future Pavement Condition No No No No Yes User

Study Timeline Yes Yes Yes Yes Yes User

Discount Rates Yes Yes Yes No No User

Rail-Truck Diversion Rate No No Yes No No User (based on FAF4)

Annual Average Traffic Data No Yes Yes No No User

Number of train crossings No No Yes No No User

Vehicle Hours Traveled Yes No No Yes Yes Travel Demand Model

Vehicle Miles Traveled Yes No No Yes Yes Travel Demand Model

Delay Yes No No Yes Yes Travel Demand Model

Trips Yes No No Yes Yes Travel Demand Model

Occupancy Yes No No Yes Yes Travel Demand Model

Origin and Destination No No No Yes No Travel Demand Model

Trip Purpose Yes No No Yes Yes Travel Demand Model

Wage Rates Yes Yes Yes No Yes Bureau of Labor Statistics

Fuel Costs Yes Yes Yes No Yes US Energy Information Administration

Vehicle Operating Costs Yes Yes Yes No Yes AAA, ATRI

Crash Rates Yes Yes Yes No Yes Highway Statistics (FHA), State DOT’s

Crash Values Yes Yes Yes No Yes US Department of Transportation

Emissions Rates Yes Yes Yes No Yes EPA, Department of Energy

Emissions Values Yes Yes Yes No Yes US Department of Transportation

Shipping Costs No No Yes No No Bureau of Transportation Studies, TIGER Methodology

Value of Freight No No Yes Yes No FAF4



Volume of Freight No No Yes No No FAF4

Weight per Truckload No No Yes No No FAF4

Buffer Indices No No No Yes No Federal Highway Administration

Current Pavement Conditions No No No No Yes State DOT’s

Length of Road Segments No No No No Yes State DOT’s

Consumer Price Index (CPI) Yes Yes Yes Yes Yes Bureau of Labor Statistics



2.0 Travel Demand Models

A travel demand model (TDM) is a term for a broad category of tools used by transportation planners and

engineers to understand current traffic conditions across a study area, and how these conditions will be

impacted in the future, if changes are made to the transportation network, which may include highways, all

roadways, transit, and/or the rail network. The inputs for these models include the transportation network,

population and employment, as well as household demographic attributes used to determine expected trip

generation choices.

TDM’s fall into several categories:

Sketch-Planning Models – Sketch-planning models are simple and easy to use models that can be

used to quickly produce order of magnitude estimates of traffic conditions and have minimal data

requirements.

Strategic Planning Models – Strategic planning models produce relatively accurate estimates of traffic

conditions under different scenarios, but are restricted to small geographic regions immediately

surrounding the planned project.

Trip-Based Models – Trip-based models (also known as four-step models) are accurate models which

cover large geographic areas. The four steps involved in trip-based model estimation are1:

− The first step is to estimate the number of trips generated by and arriving at each sub-geography in

the study area.

− The second step is to understand how these trips flow from each origination area to each

destination area.

− The third step is to determine the mode by which the trip is taken (for example via private vehicle,

public transit, or active transit).

− The fourth step is to assign each trip to a specific pathway through the transportation network.

In order to create these models accurate traffic data across the study area is required.

Activity-Based Models - Activity-based models are similar to four-step models but use more complex

estimation methodologies, focusing on the actions of individual households rather than sub-

geographies2. These model produce detailed and accurate results but have higher requirements in terms

of both data and level of effort in their estimation.

1 Travel Demand Forecasting: Parameters and Techniques. NCHRP Report 716, Transportation Research Board (TRB),

2012

2 Castiglione J., Bradley M., and Gliebe J. Activity-Based Travel Demand Models: A Primer. SHRP 2 Report S2-C46-RR-

1, Transportation Research Board (TRB), 2015.

(Footnote continued on next page...)



Origin-Destination Matrix Estimation (ODME) Models – ODME models are generally used in cases in

which not all of the data requirements are met for the estimation of a trip-based model.

The goal of ODME models is to produce table of origins and destinations for each sub-geography (also

know as an O-D matrix) which would produce the known traffic counts on for those links in the system for

which there is data3. The O-D matrix is then used to fill in estimated traffic volumes for the rest of the

links in the network.

The modeling of the O-D matrix is an iterative process that starts with a “seed” O-D matrix as a starting

point. This table is used to produce initial traffic volume estimates. These are compared to actual

volumes and the O-D matrix is adjusted based on the differences between the actual and estimated

volumes. Using the updated table, traffic volumes are estimated again, and again compared to actual

traffic volumes, which results in more updates to the O-D matrix. This process is continued until the

differences between actual and predicted traffic volumes are reduced to an acceptable level (called

convergence). The following figure (Figure 2.1) illustrates the process.

Figure 2.1 Process of Origin-Destination Matrix Estimation

Source: Cube Analyst Drive Workshop – Citilabs Training by Heejoo Ham, Ph. D. – June 2015.

Of the listed models above, the trip-based models, activity-based models, and ODME models are able to produce the needed TDM output for the FEAT model. Some but not all strategic planning models are able to produce output with enough detail to be used. It is very unlikely that any sketch-planning models will produce output that meets the requirements of the FEAT model.

2.1 The SHIFT Model

Member states of ITTS may have access to their own regional or state-level TDM. All member states also

have access to the Southern Highway Interactive Freight Traffic (SHIFT) Model. The SHIFT Model is an

3 Analytical Travel Forecasting Approaches for Project-Level Planning and Design. NCHRP Report 765. Transportation

Research Board (TRB). 2014



ODME Model which covers the ITTS region (Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi,

Missouri, Virginia, and West Virginia).

The SHIFT Model is not meant to replace other TDM’s available to states or MPO’s in the ITTS region, but

rather to provide additional information, particularly in the case of regional or national truck traffic4.

The SHIFT Model has the advantage of seamlessly covering all 15 member states of the Latin American

Trade and Transportation Study (Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi,

Missouri, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia), so that

interstate freight travel is more accurately estimated. The model has some limitations that may require the

use of supplemental models in order to provide the data needed to assess the costs and benefits of a

project.

The road network included in the SHIFT Model includes highways, intermodal connectors, principal

arterials, some rural minor arterials, and those urban streets important in the freight system. Therefore

the model does not take all streets into consideration, which may prove to be an over simplification of the

road network for small study areas.

The SHIFT Model does not integrate the rail network. Therefore separate freight rail assumptions must

be used. Because the two are not linked any diversion of freight from rail to truck or truck to rail must be

estimated by the user.

4 CDM Smith. SHIFT ODME Model & Utilities. Prepared for Institute for Trade and Transportation Studies. 2016



3.0 Roadway Project Module

In current practice, travel efficiencies are the largest share of estimated roadway investment benefits. These

efficiencies may arise from travel time savings and vehicle operating costs savings, as well as road safety

and air quality improvements. The benefits accrue to users of the transportation system, such as vehicle

owners, motorists, or other passengers, but they can also accrue to local communities as a result of a

project’s contribution to air quality improvement, for example.

Different roadway improvements may lead to different benefit categories that account for the largest amount

of accrued benefits. For example, roadway capacity enhancements have the largest impacts on travel time

savings, while operational enhancements lead to vehicle operating cost savings and road safety

improvements.

Figure 3.1 shows a schematic representation of roadway investment benefits. For each category, the data or

assumptions required for the benefit estimation is shown in the rounded rectangles (for example, value of

time) and the types included in each benefit category (for example, types of travel or types of vehicle) are

shown in document boxes.

Figure 3.1 Benefits of Roadway Investment



3.1 Travel Time Savings

Travel time savings can arise from reduced traffic congestion as a result of a capacity increase, or reduced

vehicle delay as a result of newly a built overpass, for example. Travel time cost is equivalent to the

opportunity cost of time for passenger trips (either commute or non-work trips), or to the out-of-pocket costs

for truck trips or business passenger trips, and are calculated as a function of total vehicle hours traveled for

both the baseline and build scenarios (as part of the demand model output). It is important to highlight that

travel time savings can be accrued not only by current roadway users, but also by users who divert trips from

alternative routes which are no longer preferred after the roadway investment.

Equation (1) below shows daily travel time savings calculations for each type of travel (commute, passenger

business, leisure and truck travel) and for each time period (peak and off-peak). In short, the equation uses

the value of time, the average vehicle occupancy, and the total amount of travel hours saved (as estimated

by the travel demand model) to calculate the total value of travel time savings for each type of trip in each

period. Summing across periods we obtain daily travel time cost changes, and summing across trip type we

obtain savings for all roadway users.

∆𝑻𝑻𝑪𝒑,𝒕 = 𝑽𝑶𝑻𝒑× ∆𝑽𝑯𝑻𝒑,𝒕× 𝑽𝑶𝑹𝒑 (1)

Where:

∆𝑇𝑇𝐶𝑝,𝑡 = the change in travel time costs of trip purpose p in year t

𝑉𝑂𝑇𝑝 = the value of time for the study region, by trip purpose

∆𝑉𝐻𝑇𝑝,𝑡 = the change in vehicle hours traveled, by trip purpose in year t

𝑉𝑂𝑅𝑝 = the average vehicle occupancy rate by trip purpose

p = subscript indicating the trip purpose

𝑡 = subscript indicating the specific intermittent year over the analysis period

For truck transportation, average vehicle occupancy is assumed to be 1, while for passenger transportation

average vehicle occupancy is reported in the travel demand model, or is assumed at 1.2 if no information is

available. Travel time savings for commuting or business passenger trips are annualized multiplying by 260

working days (52 weeks), while for truck trips and non-work trips they are annualized multiplying by 365

days. Travel time savings for truck travel are considered to be the gross median hourly wage for truck drivers

reported for the region. The hourly value of time for passenger travel is considered to be gross median hourly

wage specific to the study region.

In the economic impact model travel time savings attributable to shorter commute times are considered

savings in household income, and they are valued at half of the gross median hourly wage. Second,

business-related travel time savings are considered cost savings for businesses, and are valued at the total

median hourly wage. These business cost savings are distributed across various industries based on each

industry’s transportation cost per dollar of output. Third, travel time savings for non-work trips are not

considered as an input in the economic impact model, since they in general do not generate benefits in

monetary terms. Finally, travel time savings for truck travel accrue across firms according to the industry’s



reliance on freight trucking (based on data from the U.S. Department of Transportation’s Transportation

Satellite Accounts).

3.2 Vehicle Operating Cost Savings

Vehicle operating cost savings are also proportional to change in total travel. These savings originate from (i)

a reduction in travel delays (as a result of capacity project, for example), (ii) a decrease in distance traveled

(as a result of new routes or from diverting longer trips), or (iii) a reduction in vehicle damage (as a result of

improved pavement conditions or pothole repair). Vehicle operating expenses results from vehicle ownership

and use, and may be divided between non-fuel related costs and non-fuel related costs. Non-fuel related

costs can be calculated as the average cost per mile regardless of speed, and includes maintenance, tires,

repairs, and mileage dependent depreciation costs. On the other hand, fuel related costs depends not only

on mileage but also on average speed and fuel economy.

The difference in total non-fuel vehicle operating costs between the build and no-build scenario can be

estimated with equation (2), using the changes in total vehicle miles traveled (from the demand model

output) and the average non-fuel vehicle operating costs for each vehicle type (auto and truck).

∆𝐀𝑽𝑶𝑪𝒗,𝒕𝒏𝒇𝒄

= ∆𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒏𝒇𝒄

(2)

Where:

∆A𝑉𝑂𝐶𝑝,𝑡𝑛𝑓𝑐

= the annual change in non-fuel vehicle operating costs, for vehicle type v in year t

∆𝑉𝑀𝑇𝑣,𝑡 = the change in vehicle miles travelled, for vehicle type v in year t

𝑉𝑂𝐶𝑣𝑛𝑓𝑐

= the non-fuel operating costs per vehicle mile, for vehicle type v

v = subscript indicating the vehicle type


In a similar calculation, change in fuel vehicle operating costs relies on the reduction in vehicle miles

traveled, and on the average fuel economy by type of vehicle (auto or truck) at the current speed, as shown

in equation (3a)

∆𝐀𝑽𝑶𝑪𝒗,𝒕𝒇𝒄

= 𝑭𝑪𝑹𝒗× ∆𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒇𝒄

(3a)

Where:

∆A𝑉𝑂𝐶𝑝,𝑡𝑓𝑐

= the annual change in fuel costs, for vehicle type v in year t

𝐹𝐶𝑅𝑣 = the average fuel consumption rate for vehicle type v


𝑉𝑂𝐶𝑣𝑓𝑐

= fuel costs per vehicle gallon, for vehicle type v





Another factor affecting vehicle operating cost is the relationship between average speeds and fuel

consumption. If a project changes travel speeds, users will experience a change in fuel consumption even

though vehicle miles traveled may remain constant. Equation (3) below shows this relationship:


= ∆𝑭𝑪𝑹𝒗× 𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒇𝒄

(3b)

Where:


= the annnual change in fuel costs, for vehicle type v in year t

∆𝐹𝐶𝑅𝑣 = the change in fuel consumption rate for vehicle type v

𝑉𝑀𝑇𝑣,𝑡 = the vehicle miles travelled, for vehicle type v in year t





The relationship between fuel consumption and average speeds is calculated using average auto or truck

fuel economy, and the efficiency curve (consumption-by-speed) is obtained from models developed for

assessing such relationship. Annual change in total vehicle operating costs for autos and trucks was

estimated as the product of the daily change in vehicle operating cost and 365 days.

3.3 Safety Benefits

In cost benefit analysis, road safety benefits can be calculated as a function of crash rates, crash severity,

and vehicle exposure to traffic. First, safety benefits may accrue as an externality created by a reduction in

vehicle miles traveled due to shorter trips or trip diversion, and the rationale for this is that the lower the

vehicle exposure, the lower the probability of road accidents, given a constant crash rate. The method to

assess external safety benefits involves applying the local or regional fatality, injury and property damage

only (PDO) rates to the annual changes in vehicle miles traveled, and then, estimating the dollar value by

using the comprehensive cost of motor vehicle crashes by injury level, as follows:

∆𝑨𝑺𝑪𝒄,𝒕 = 𝑪𝑹𝒄×∆𝑽𝑴𝑻𝒕 ×𝑺𝑪𝒄 (4a)

Where:

∆𝐴𝑆𝐶𝑐,𝑡 = the change in annual safety costs, for crash type c in year t

𝐶𝑅𝑐 = the crash rate for crash type c

∆𝑉𝑀𝑇𝑡 = the change in vehicle miles travelled in year t

𝑆𝐶𝑐 = the economic cost of crash type c



c = subscript indicating the crash type


Second, safety benefits can also accrue to projects that directly reduce crash rates. For example, projects

that address specific safety concerns, such as improving pavement conditions (therefore lowering the

probability of accidents in a specific roadway segment), eliminating rail crossings (therefore lowering

vehicle/train crash rates to zero), or adding additional lanes in two lane roads (and reducing the probability of

head-on collisions) have direct safety benefits that accrue on top of the safety externalities related to traffic

exposure. In order to calculate these benefits, the expected change in the before and after crash rates is

necessary, and the benefits are given by:

∆𝑨𝑺𝑪𝒄,𝒕 = ∆𝑪𝑹𝒄×𝑽𝑴𝑻𝒕 ×𝑺𝑪𝒄 (4b)

Where:


∆𝐶𝑅𝑐 = the change in crash rate for crash type c

𝑉𝑀𝑇𝑡 = the vehicle miles travelled in year t




3.4 Air Quality Benefits

Air quality benefits accrue by reducing emissions that affect local or regional air quality such as carbon

monoxide (CO), volatile organic compounds (VOCs), nitrogen oxides (NOx), particular matter (PM), and

sulfur dioxide (SOx), in addition to emissions with impacts at a global level, such as carbon dioxide (CO2),

which are included in cost benefit analyses that follow federal guidelines. Emissions can be avoided by

reducing total vehicle miles traveled, hence air quality impacts are estimated as follows:

∆𝑨𝑨𝑸𝑪𝒆,𝒗,𝒕 = 𝑬𝑹𝒆,𝒗× ∆𝑽𝑴𝑻𝒕 ×𝑨𝑸𝑪𝒆 (5a)

Where:

∆𝐴𝐴𝑄𝐶𝑒,𝑣,𝑡 = the change in annual air quality costs, for emissions type e from vehicle type v in year t

𝐸𝑅𝑒,𝑣 = the rate of emissions for emissions type e from vehicle type v


𝐴𝑄𝐶𝑒= the air quality cost of emissions in dollars per gram for emission type e

e = subscript indicating the emission type





The emission rates by vehicle type (auto or truck) and by emission type are calculated by dividing emission

coefficients (emissions per gallon of fuel) and average fuel economy for trucks and autos (miles per gallon of

fuel at a given average speed). Industry standard parameters are used for this calculation. The dollar cost of

emissions is given by federal or state guidance for economic impact analysis.

As discussed in Section 3.2 above, when a roadway project improves the roadway level of service, and

hence increases average vehicle speed, this leads to savings in fuel consumption for all vehicles. In turn,

these savings also represent a reduction on emissions, which can be calculated using equation (8) below.

∆𝑨𝑨𝑸𝑪𝒆,𝒗,𝒕 = ∆𝑬𝑹𝒆,𝒗× 𝑽𝑴𝑻𝒕 ×𝑨𝑸𝑪𝒆 (5b)

Where:


∆𝐸𝑅𝑒,𝑣 = the change in the rate of emissions for emissions type e from vehicle type v






Finally, emission mitigation is an important component of a cost benefit analysis. In an economic impact

analysis, however, these benefits are not computed as they do not have a job or economic multiplier effect

on the regional economy.



4.0 Operational and Safety Improvements

The traditional benefit categories of roadway expansion may also accrue to smaller operational and safety

roadway improvements. However, operational projects typically provide capacity expansion or safety

improvements at a scale that is not captured by aggregate travel demand models. While travel demand

models are validated for a regional analysis, operational projects seek to provide benefits targeted at a more

microscopic level. In order to estimate the benefits of operational improvements, a planning-level

assessment for operations or safety impacts using limited travel demand information and standard

parameters is provided by the tool.

The operational roadway improvements include:

Increasing Length of Turn Bay

Increasing the Turn Radius

Striping Changes

Signal Timing/Phasing Changes

Prohibiting Left-Turn Movements

Prohibiting On-Street Parking

Adding Turn Lanes

Adding a Through Lane

Adding a Traffic Signal

Adding a Roundabout

Innovative Intersection (Continuous Flow Interchange, etc.)

Bridge Replacement

These operational improvements generate benefits from reduced traffic congestion and improved safety.

However, as they do not provide a sizeable change in vehicle miles traveled, benefits from reduced vehicle

operating costs and from emission mitigation are not considered relevant in this sketch level analysis. The

analysis relies on, therefore, on the expected delay reduction and on the expected reduction in local crash

rates.

4.1 Operational Only Projects

In the sketch level tool, operational only projects are assumed to generate benefits only from travel time

savings. The estimation relies on existing Average Annual Daily Traffic (AADT) information for the base year,

rather than on more a detailed travel demand model output. For each project, the type of improvement must

be specified, such as adding a turn lane or prohibiting on-street parking, as the expected impact may vary by

project category (low, moderate, or high impact).

In the model, total travel time savings are annualized with similar factors than those considered for roadway

expansion projects: 260 days for commute and business related travel, and 365 days for non-work and truck

travel. In order to forecast traffic count values for future years (in order to perform the benefic cost and

economic impact analysis), default traffic growth rates and truck mile percentages are based on the

parameters used in statewide travel demand models. Assumptions for traffic growth and percentage of truck

travel can also be directly inputted into FEAT.



The reduction in delay per vehicle (in seconds) depends on the initial roadway level of service (LOS). If the

user does not directly specify the initial LOS in each project, the tool directly assumes an F level. The LOS

specification is important since it determines the original delay per vehicle, according to the table below.

Therefore, changing initial conditions will also change initial and final delay per vehicle.

Table 4.1 Delay per Vehicle Based on Roadway Level of Service

Level of Service Range of Delay per Vehicle

(Seconds) Default Assumption

A Less than 10 0

B 10 to 20 10

C 20 to 35 20

D 35 to 55 35

E 55 to 80 55

F More than 80 80

Vehicle delay reduction depends on the type of project considered. For each project category (with minor,

moderate or major impact) total delay is assumed to be reduced by a give percentage, based on average

values for the types of projects considered. A minor project, such as an increase in the length of the turn bay

is expected to save 15% of total delay in congested hours, a moderate impact project, such as adding turn

lanes, is assumed to save 30% of total delay in congested hours, and a major project, such as a continuous

flow interchange, is assumed to save 60% of total delay in congested hours. The following table shows the

percentage of total travel time savings that can be achieved with each type of project being considered.

Table 4.2 Operational Project Type and Impact Level Assumptions

Project Type Operational Impact Level Operational Time Reduction

Increase Length of Turn Bay Minor 15%

Increase Turn Radius Minor 15%

Striping Changes Minor 15%

Signal Timing/Phasing Changes Minor 15%

Prohibit Left-Turn Movements Minor 15%

Prohibit On-Street Parking Minor 15%

Adding Turn Lanes Moderate 30%

Adding a Through Lane Moderate 30%

Adding a Traffic Signal Moderate 30%

Adding a Roundabout Moderate 30%

Bridge Replacement Moderate 30%

Interchange Construction Moderate 30%



Innovative Intersection Major 60%

The tool also uses the K-factor (planning analysis hour factor) as the parameter that determines peak-to-

daily ratio of traffic volume. Once multiplied by the daily traffic count (AADT), it gives an estimation of the

traffic volume on an average congested hour of the day. The default value is 0.1, but in more urbanized

areas with higher traffic volume the k-factor may drop since peak traffic is spread over longer periods of time.

This parameters must be adjusted from the default value to conform to the project being assessed, but it

typically ranges from 8% to 10%.

The number of congested hours per day is another important parameter to calculate total travel time savings.

It is usually given by the travel demand model, or it can be adjusted for the specificities of each project. It is

assumed that travel time saving only accrue during hours with delay.

The default parameters used to estimate total travel time savings are shown in the Table below. Each can be

adjusted to better characterize specific projects.

Table 4.3 Default Assumptions for Operational Projects

Parameters Default Value

No-Build Level of Service F

Delay per Vehicle 80

K-Factor 10%

Number of Congested Hours/day 4

Using these assumptions, the total changes in travel time are calculated with the following equation:

∆𝑨𝑨𝑫𝑻𝑻𝑪𝒑,𝒕 = 𝑽𝑶𝑻𝒑 × 𝑽𝑶𝑹𝒑,𝒕 × (𝑨𝑨𝑫𝑻𝒑,𝒕 ×𝑲) × 𝑷𝑯𝒕 × 𝑫𝒕 (6)

Where:

∆𝐴𝐴𝐷𝑇𝑇𝐶𝑝,𝑡 = the change in average annual daily travel time costs for trip purpose p in year t

𝑉𝑂𝑇𝑝 = the value of time for trip purpose p

𝑉𝑂𝑅𝑝,𝑡 = the average vehicle occupancy rate for trip purpose p in year t

𝐴𝐴𝐷𝑇𝑝,𝑡 = the average annual daily traffic volume for trip purpose p in year t

𝐾 = the K-Factor for the project

𝑃𝐻𝑡 = the number of congested (peak) hours per day in year t

𝐷𝑡 = the daily delay per vehicle in hours during year t





The annual changes in travel costs are estimated for autos and trucks, using the specified truck percentage,

and the annual factors for mode and type of trips. Similarly to roadway capacity projects, only those benefits

that accrue to business travel and to truck travel are transferred in full to the economic impact analysis tool,

while benefits for commuting travel are accrued at half the gross median hourly wage of the region where the

project will take place. The percentage distribution of auto vehicle miles travelled (VMT) by three purposes

(commuting, business, and non-work) is obtained for each region from statewide travel demand model.

4.2 Safety Projects Only

Annual safety benefits are estimated when projects are expected to avoid crashes and reduce accident

costs. Three groups of accidents are included in the analysis: fatalities, injuries, and property damage only

(PDO). In this planning level analysis, different types of projects are assumed to have different impacts on

crash rates, which can be minor, moderate or major. The table below shows the assumptions for crash

reduction rates for each type of project considered. The assumptions may be overridden by the tool user

when the project’s safety impact can be more accurately estimated.

Table 4.4 Safety Project Types and Impact Assumptions

Table Header Table Header Table Header

Increase Length of Turn Bay Minor 20%

Increase Turn Radius Minor 20%

Striping Changes Minor 20%

Signal Timing/Phasing Changes Minor 20%

Prohibit On-Street Parking Minor 20%

Adding Turn Lanes Minor 20%

Interchange Reconstruction Moderate 40%

Adding a Through Lane Moderate 40%

Adding a Traffic Signal Moderate 40%

Adding a Roundabout Moderate 40%

Bridge Replacement Moderate 40%

Innovative Intersection Major 80%

Prohibit Left-Turn Movements Major 80%

Important inputs in the benefit estimation are: accumulated distance (in miles) within the project influence

area, and the annual average daily traffic (AADT). Similar to operational projects, existing AADT information

in the area of influence is required for the analysis. When multiplied by the accumulated distance in the

project influence area, this yields the total daily vehicle miles traveled that are subject to safety



improvements. The distance is assumed to be 1 mile, based on the length of typical safety project plus the

areas on approach and departure, however, this parameter can also be changed by the user.

Using average crash rates, the total number of accidents can be estimated for the base year. The numbers

of crashes, by type, are then multiplied by the average crash costs (also by type) estimate the value of safety

benefits. Average crash rates may be overridden by the user if more detailed information for the project

location is available.

The following equation is used to estimate safety benefits per crash type and per trip type:

∆𝑨𝑨𝑫𝑺𝑪𝒄,𝒑,𝒕 = ∆𝑪𝑹𝒄 × 𝑨𝑨𝑫𝑻𝒑 ×𝑳 × 𝑺𝑪𝒄 (7)

Where:

∆𝐴𝐴𝐷𝑆𝐶𝑐,𝑝,𝑡 = the change in average annual daily safety costs, for crash type c and trip purpose p, in

year t


𝐴𝐴𝐷𝑇𝑝 = the average annual daily traffic volume for trip purpose p

𝐿 = the extent of the project influence, in miles





The total annual benefit in dollars per year is then calculated using 260 days for commuting and business

trips, and 365 days for trucks and leisure trips. Finally, savings in accident costs in 2040 are also estimated

based on the default or user specified value of annual growth rate from 2010 to 2040, crash rates by type,

crash costs and crash percentage of reduction by project type.

4.3 Operational and Safety Projects

Roadway projects can generate benefits both from an operational perspective (through travel time savings)

and from a safety perspective (through crashes avoided). In this case, the user will combine the two previous

analysis and estimate the total benefits using the same assumptions and methodology for both benefit

categories.



5.0 Freight Rail Project Module

Estimating railroad project benefits involve, for the most part, similar travel efficiency categories as those for

roadway investments. However, these benefits arise not only from savings within the railroad system, but

also from the interaction between the railroad and roadway systems as a result of the potential for truck-to-

rail diversion. Other rail project benefits include shipping cost savings from stronger supply chain networks or

from transportation productivity gains. Examples of railroad projects include capacity expansion (which

includes new intermodal yards that potential for intermodal freight), track upgrades, at-grade crossing

elimination, and double tracking. Figure 5.1 below shows a schematic representation of the potential benefit

of rail investment, as well as the data required to estimate each benefit category.

Figure 5.1 Benefits of Freight Rail Investment

5.1 Shipping Cost Savings

Shipping cost savings are internal benefits accrued to shippers. The benefits arise from the diversion of

freight from truck to rail, and hence from lower rates on rail relative to truck. Shipping cost savings are a

relevant source of efficiency gains, as they lead to an overall cost reduction in the transportation system.

Moreover, reduced shipping rates can be measured as benefits from different sources: better productivity of

rail vs. truck, or increased access to new markets through lower transportation costs.

In order to quantify shipping benefits as a result of diverted freight from trucks, the net change in total in

shipping costs must be assessed. First, this involves estimating the additional number of ton-miles carried by

rail. As most travel demand models do not integrate railway and roadway, diverted freight must be estimated

by analyzing the potential for truck-to-rail diversion in each region. In the model, this can be either a direct

input when freight diversion has been previously estimated for the project, and should be broken down

between terms of intermodal units, and boxcars, and other. Otherwise, an estimation of the diversion



potential can be calculated using the tool, which has available freight flow data from the Freight Analysis

Framework. Current regional rail mode share is compared to the national rail mode share in order to estimate

rail freight potential for each type of commodity. Total truck miles diverted is then calculated using average

truck carload for each type of commodity

𝑨𝑫𝑷𝒈,𝒕𝒓,𝒔𝒂 = 𝑨𝑭𝑽𝒈,𝒕

𝒔𝒂 × (𝑴𝑺𝒈𝒓,𝒏𝒂 − 𝑴𝑺𝒈

𝒓,𝒔𝒂) (8)

Where:

𝐴𝐷𝑃𝑔,𝑡𝑟,𝑠𝑎

= the annual potential rail diversion in the study area in ton-miles, for good g in year t

𝐴𝐹𝑉𝑔,𝑡𝑠𝑎 = the annual freight volume in ton-miles of good g in the study area in year t

𝑀𝑆𝑔𝑟,𝑛𝑎 = the rail mode share (percentage) nationally for good g

𝑀𝑆𝑔𝑟,𝑠𝑎 = the rail mode share (percentage) in the study area for good g

g = subscript indicating the type of good or commodity


Second, shipping costs are estimated by multiplying additional rail freight with the corresponding difference

between average shipping rates for rail and truck. Shipping rates, in this context, reflect the total shipping

cost – including drayage costs for intermodal shipping – for the rail mode.

Equation 9 below summarizes the changes in shipping costs:

∆𝑨𝑺𝑯𝑪𝒈,𝒕 = (𝑺𝑹𝒓 − 𝑺𝑹𝒕𝒓) × 𝑨𝑫𝑷𝒈,𝒕𝒓,𝒔𝒂

(9)

Where:

∆𝐴𝑆𝐻𝐶𝑔,𝑡 = the change in annual shipping costs for good g in the study area in year t

𝑆𝑅𝑟 = average shipping rates per ton-mile for rail

𝑆𝑅𝑡𝑟 = average shipping rates per ton-mile for truck

𝐴𝐷𝑃𝑔,𝑡𝑟,𝑠𝑎

= the annual potential rail diversion in the study area in ton-miles, for good g in year t



5.2 External Benefits of Truck Diversion

Railroad projects also create benefits on the roadway as a result of truck diversion. Truck-to-rail diversion

produces positive externalities associated with the reduction of truck miles traveled and truck hours traveled.

In this section, as travel demand models do not usually integrate roadway and railway demand, the truck



miles traveled diverted must be estimated using the potential diversion for each commodity type, truck load

factors, and total freight in tons. The externalities include reduced traffic congestion, reduced roadway wear

and tear, improved road safety and better air quality.

5.2.1 Congestion

The rationale behind the impact of truck diversion on road congestion is that fewer vehicles on the road

alleviate traffic and generate travel time savings for other road users. In order to quantify this benefit, the

amount of vehicle miles traveled by truck that are diverted to rail is estimated using potential diversion by

commodity as well as average load factors for each type of commodity. Moreover, the Federal Highway

Administration has estimated the external costs of truck driving, which is the delay cost imposed by a truck

on other users of the road. The external cost is calculated for peak and off-peak periods, as well as for rural

and urban roadways – therefore a weighted average representative of truck miles is calculated. The total

savings in external costs related to road congestion can be calculated using the following equation:

∆𝑨𝑪𝑪𝒕 = 𝑪𝑪𝒕𝒓 × (𝑫𝑷𝒈𝒓,𝒔𝒂 ×

𝑨𝑭𝑽𝒈,𝒕𝒔𝒂

𝑳𝑭𝒈𝒕𝒓 ) (10)

Where:

∆𝐴𝐶𝐶𝑡 = the change in annual congestion cost in year t

𝐶𝐶𝑡𝑟 = the cost of congestion per truck mile


𝐿𝐹𝑔,𝑡𝑟 = the average truckload in tons for good g

𝐷𝑃𝑔𝑟,𝑠𝑎

= the potential rail diversion percentage in the study area for good g

𝑫𝑷𝒈𝒓,𝒔𝒂 = (𝑴𝑺𝒈

𝒓,𝒏𝒂 − 𝑴𝑺𝒈𝒓,𝒔𝒂) (11)

Where:





5.2.2 Roadway Wear and Tear

In addition to congestion, truck-to-rail diversion also reduces roadway wear and tear. Compared to autos,

trucks cause the most damage on the roads, hence a decrease in truck miles traveled also lowers road

maintenance needs. In order to measure total benefits in dollars from avoided physical stress on roadways,



the marginal road maintenance expenditure by type of roadway must be assessed (default estimates by the

Federal Highway Administration are used). Maintenance costs are assumed to be for the average truck and

the average mix of urban and rural roadways. Cost per mile is then multiplied by the estimated total truck

miles diverted, as shown in Equation 12:

∆𝑨𝑯𝑴𝑪𝒕 = 𝑯𝑴𝑪𝒕𝒓 × (𝑫𝑷𝒈𝒓,𝒔𝒂 ×

𝑨𝑭𝑽𝒈,𝒕𝒔𝒂

𝑳𝑭𝒈𝒕𝒓 ) (12)

Where:

∆𝐴𝐻𝑀𝐶𝑡 = the change in annual highway maintenance cost in year t

𝐻𝑀𝐶𝐶𝑡𝑟 = the highway maintenance cost per truck mile


𝐿𝐹𝑔,𝑡𝑟 = the average truckload in tons for good g

𝐷𝑃𝑔𝑟,𝑠𝑎

= the potential rail diversion percentage in the study area for good g

𝑫𝑷𝒈𝒓,𝒔𝒂 = (𝑴𝑺𝒈

𝒓,𝒏𝒂 − 𝑴𝑺𝒈𝒓,𝒔𝒂) (11)

Where:





5.2.3 Safety Externalities

A change in truck miles traveled due to truck-to-rail diversion also generates road safety benefits. This

benefit may be calculated in a similar way as the safety benefits shown in Section 3.3. That is, using specific

information on crash rates, the average costs of road accidents (fatalities, injuries, and property damage), as

well as estimated reduction in total vehicle miles traveled due to modal shift. If the railway in which the freight

will be diverted to also poses safety risks (calculated with historical rail accident rates in the region per train

mile), then those are to be included in the net calculation of the total safety benefits.

5.2.4 Air Quality Impacts

In a similar way to safety benefits, air quality benefits also arise from a reduction in truck miles traveled. The

dollar value of these benefits is estimated using the same equation as in Section 3.4, thus employing

average emission rates for trucks (per mile) as well as the value of the different types of fuel emissions (CO,

VOCs, NOx, SOx, PMs, CO2).



These air quality benefits must also be calculated net of equivalent rail emissions, which also requires

information on average locomotive emissions by pollutant type. It is important to highlight that while truck

emissions are calculated on a per mile basis, locomotive emissions can be approximated on a per ton basis

(following guidance by the Environmental Protection Agency), for which a conversion factor must be used

such as the average truck load by commodity. Overall, rail transportation is estimated to be 11 times more

efficient than truck transportation with respect to carbon dioxide emissions, for example.

5.3 Benefits of Grade Crossing Elimination

Projects that involve rail relocation or grade crossing elimination generate public safety benefits by reducing

the number of predicted accidents and their severity, and by reducing waiting times at grade crossings for

motorists. In a similar way to roadway project evaluation, benefits estimation involves the calculation of

safety benefits, travel time savings, vehicle operating cost savings, and air quality benefits. Figure 5.2 below

shows the associated benefits and the data and assumptions necessary to estimate each benefit.

Figure 5.2 Benefits of Grade Crossing Elimination

5.3.1 Travel Time Savings

In order to estimate the dollar value of the time saved by road users due to reduced delays at a grade

crossings, a queuing theory model is employed. The model calculates average daily delay for autos and

trucks, and follows the same methodology used by GradeDec. It relies on several variables and parameters

that characterize roadway and train traffic at the intersection, some of these variables must be intersection-

specific while others can be default values.

With respect to train crossing information, the model is based on data or estimations of the number of daily

train crossings, the time-of-day distribution of train crossing, and the average duration of the train crossing.



The latter information, if not available, is calculated using the average train length and the average train

speed at crossing, in addition to buffer time prior between the traffic interruption and the train crossing.

With respect to roadway information, the model requires vehicle arrival rates, which can be derived from the

daily traffic counts (available as AADT data) and the time-of-day distribution of traffic (peak, off-peak, or

nighttime). Additional parameters required for this analysis include the percentage of truck trips, the number

of lanes, and the vehicle dispersal rate.

Having estimated average vehicle delay, this average is multiplied by the total number of vehicles arriving at

a blocked grade crossing by vehicle and trip type (commute, business, non-work trips, or truck trips) in a

given year. In turn, this total is multiplied by the corresponding value of time and by average vehicle

occupancy, with the same assumptions as those used in Section 3.1.

5.3.2 Vehicle Operating Cost Savings

Savings from vehicle operating expenses as a result of grade crossing elimination originate from lower fuel

consumption for vehicles idling at grade crossings. The total benefit can be assessed using the amount of

vehicle delay hours saved, as calculated in Section 5.3.1 above, and the average rate of fuel consumption of

a vehicle at idle by vehicle type. The following equation applies to this calculation:

∆𝑨𝑽𝑶𝑪𝒗,𝒕𝒇𝒄

= 𝑭𝑪𝑹𝒗𝒊𝒅 × 𝑨𝑫𝒗,𝒕

𝒈𝒄 × 𝑽𝑶𝑪𝒗

𝒇𝒄 (13)

Where:

∆𝐴𝑉𝑂𝐶𝑣,𝑡𝑓𝑐

= the change in annual fuel costs from grade separation for vehicle type v in year t

𝐹𝐶𝑅𝑣𝑖𝑑 = the fuel consumption rate when idle in gallons per hour, for vehicle type v

𝐴𝐷𝑣,𝑡𝑔𝑐

= the annual delay at the grade crossing for vehicle type v in year t


= the fuel cost per gallon for vehicle type v



5.3.3 Safety Benefits

Road-rail level crossings present a safety hazard to vehicles and to pedestrians, and the elimination of such

crossings is expected to reduce fatalities, injuries, and vehicle crashes. The first step in order to quantify

safety benefits is to estimate the predicted number and the severity of accidents at the grade crossing per

year using the methodology developed in the U.S. DOT Accident Prediction and Severity Model. Since every

grade crossing has very different traffic and safety characteristics, there can be a wide variation in the

expected accident rates between grade crossings.

The model provides a forecast of crash rates and their severity given a series of parameters that

characterize the grade crossing. Required information involves the AADT, average daily train crossings,

maximum train timetable speed, and the number of highway lanes, which is the same data also required for



the calculation of travel time savings. In addition, the pavement type, the highway type, the number of train

crossings during daylight, and the number of switch trains per day may be also determined by the user or

kept at the default value. Finally, the crossing category is another information that must be selected, which

can be either passive, flashing lights or gates.

Having assessed the expected number of fatalities, injuries, and property damage only accidents in a given

grade crossing, the dollar value of these safety benefits can be estimated using the same assumptions as

discussed in Section 3.3 the benefits can be estimating using the equation below:

∆𝑨𝑺𝑪𝒄,𝒕 = 𝑨𝑪𝑹𝒄𝒈𝒄

× 𝑺𝑪𝒄 (14)

Where:

∆𝐴𝑆𝐶𝑐,𝑡 = the change in annual safety cost from the grade separation, for crash type in year t

𝐴𝐶𝑅𝑐𝑔𝑐

= the annual crash rate at the grade crossing for crash type c




5.3.4 Air Quality Benefits

To the extent that rail relocation and grade separation reduce vehicle idling at the grade crossing due to road

blockage, the corresponding fuel savings will also generate a benefit in term of avoided emissions. Using the

same assumptions as Section 3.4 regarding fuel economy, emission rates, and value of emissions by type of

emission (VOCs, SOx, NOx, PMs, CO, CO2) and by type of vehicle (car or truck), we may calculate the total

dollar value of air quality benefits from this type of rail investment.

5.4 Benefits of Rail Upgrades (up to 286,000 lbs.)

Railway upgrades that allow heavier trains on the tracks have a direct impact on rail cost effectiveness.

Heavier loads reduce costs per ton-mile as railroads become more productive, introducing economic

efficiency gains to the rail transportation system, and generating benefits in terms of shipping reliability and

shipping time. Different types of benefits can be realized from such investments.

First, truck-to-rail diversion is a major source of rail upgrade benefits. Mode change is made possible by new

freight rail shipment in industries that prefer rail shipping with higher volumes. A more reliable and faster

transit time also make rail a more viable option vis-à-vis truck. Finally, if there are capacity constraints on the

railway, truck may be a more reliable option for shippers or there is excess rail demand that cannot be met.

Therefore, a more productive rail system reduces logistic costs for shippers that choose the rail mode.

In a similar way than the calculations in Section 5.2 rail diversion potential can be calculated using the

differential between the rail mode shares in the region where the investment is taking place and the national

average. In case the project has already estimated the additional freight expected from the upgrade, then it

should be used directly as an input into the model. Once the number of truck miles diverted from the roads is



calculated, the externalities (alleviation of congestion, safety improvement, roadway wear and tear reduction,

and emission mitigation) from reduced truck travel on roadways can be calculated using the equations in

Section 3.5.2.

Second, shippers also benefit from lower rates on rail compared to truck. The reduction in shipping costs is

calculated as the differential between shipping costs per ton-mile for truck and rail (as calculated in Section

5.1), a reflection of the reduced logistics costs.

Other benefits from rail upgrades that are not calculated in the planning-level tool, but may be relevant for

larger investments include: (i) transit time savings, as the new railway may allowed higher average speeds

(however, this is only relevant for shipments that travel longer distances over rail), (ii) emission savings per

ton-mile for freight that currently uses the railway, and (iii) private benefits for the railroad, which will be able

to realize economies of scale.

Figure 5.3 below shows the potential benefits of rail upgrades and the required data and assumptions to

estimate each category.

Figure 5.3 Benefits of Rail Upgrades

5.5 Benefits of Double Tracking

Double tracking generates benefits from the alleviation of capacity constraints by reducing logistics costs to

shippers currently using the railway and to roadway users through externalities from the diverted use of less

efficient modes.

Since double tracking makes rail ore reliable, efficient and faster, these characteristics reduce the logistics

costs for shippers that previously chose truck transportation. The estimation of the potential diversion from

truck-to-rail is based on the same methodology used in Section 5.2, calculating the differential between the

rail mode shares for each commodity in the region where the investment takes place with national averages.

If the project has already estimated potential diversion, this data can be directly applied in the model. Direct

shipping costs can also be calculated as the differential between truck and rail average shipping costs per

ton-mile.



In addition, time savings will also accrue to shippers, as double tracking reduces the need to finance

inventory while goods are in-transit on rail. Average inventory carrying costs can be estimated based on the

number of ton-miles, the average value per ton of freight rail moved, and the estimated average costs of in-

transit inventory (as calculated by the Federal Highway Administration).

Private benefits for the railroad are also expected to be accrued, however, these benefits are not included as

a public benefit calculation in the model. Cost savings accrue from lower train operating costs of train crews,

maintenance, fuel, etc., which are time related and increase with the amount of transit time.

Figure 5.4 Benefits of Double Tracking



6.0 Pavement Conditions Impact Module

A state's transportation system is vital to its economic health. Transportation serves key industries and

economic development assets in the state, provides for emergency routing, serves local commuters, and is

the gateway into the state for millions of visitors each year. Good roads are vital to the economy because

the quality of transportation impacts the cost of doing business through travel times, reliability of travel times,

and overall transportation costs. These factors directly impact productivity, as well as access to markets and

labor, which impact the region’s and state’s economic competitiveness and overall growth.

Roadway conditions have a direct impact on transportation costs through vehicle operating costs, travel

times, and safety impacts. The change in transportation costs to residents and businesses due to the

condition of highway infrastructure can be estimated for a given time period (i.e., the analysis period). Within

this evaluation are the estimations of the potential travel costs or benefits accruing to highway users due to

changing pavement conditions, measured in terms of travel time, vehicle operating costs and accident/safety

costs. The analysis does not include any construction impacts but instead focuses on long-term lasting

economic competitiveness impacts. The analysis examines the impact of both pavement conditions and

congestion levels on the state's economy. The following figure (Figure 6.1) summarizes the benefits of

investing in the improvement of pavement conditions.

Figure 6.1 Benefits of State of Good Repair Investment

6.1 Impact of Changing Pavement Conditions

Pavement conditions have a direct impact on the users of the transportation system in the following ways:



Vehicle Operating Costs

Travel Time Costs

Vehicle Crash Costs

Each of these impacts are discussed in turn in the following sections.

The performance metric to assess pavement conditions for the economic analysis is the International

Roughness Index (IRI). The IRI, an internationally accepted measure of pavement smoothness reported in

meter/kilometer (m/km) or inches/mile (in/mi), has shown good correlation to the public perception of

pavement quality and the physical condition of the pavement. The lower the IRI value, the smoother the ride

is. According to the Federal Highway Administration (FHWA) criteria, pavement with IRI less than 95

inches/mile are rated in “good” condition while pavements with IRI greater than 170 inches/mile are rated as

poor condition. Between 95 and 170 is considered acceptable or fair condition.

6.1.1 Pavement Condition Impacts on Vehicle Operating Costs

Rough roads can lead to greater fuel consumption, wear and tear, and repair and maintenance cost on

vehicles. A comprehensive research study conducted by the National Cooperative Highway Research

Program (NCHRP) on the effects of pavement condition on vehicle operating costs5 indicated the following:

For fuel consumption - surface roughness is an important factor. An increase in IRI of 63.4 in/mi will

increase the fuel consumption of passenger cars by nearly 2 percent regardless of speed. For

commercial vehicles, the increase is estimated to be about 1 percent at normal travel speeds (i.e., 60

mph) and 2 percent at low travel speeds (i.e., 35 mph).

For repair and maintenance cost - there is no effect of roughness up to IRI of 190.2 in/mi. Beyond this

range, an increase in IRI up to 253.6 in/mi will increase vehicle repair and maintenance costs by 10

percent. Worse pavement conditions will lead to more significant increases. An increase in IRI up to

317.0 in/mi will add repair and maintenance cost up to 40 percent for passenger cars and 50 percent for

heavy trucks. An increase in IRI up to 380.0 in/mi will escalate repair and maintenance cost up to 70

percent for passenger cars and 80 for heavy trucks.

For tire wear - an increase in IRI of 63.4 in/mi will increase the tire wear of passenger cars and

commercial vehicles by 1 percent at 55 mph or higher.

Table 6.1 summarizes the impact of IRI on vehicle operating costs, based on NCHRP Report 7206. The

impacts (as a percent of fuel consumption cost, repair and maintenance cost, and tire wear cost) shown in

this table are used to estimate the economic value of the additional vehicle operating costs due to pavement

conditions.

5 Chatti, K., and Zaabar, I. Estimating the Effects of Pavement Condition on Vehicle Operating Costs. NCHRP Report

720, Transportation Research Board (TRB), 2012.

6 Chatti, K., and Zaabar, I. Estimating the Effects of Pavement Condition on Vehicle Operating Costs. NCHRP Report 720, Transportation Research Board (TRB), 2012.



Table 6.1 Effects of Pavement Condition on Vehicle Operating Cost

Vehicle Operating Cost Type Range

Impact on Passenger Cars

Impact on Commercial

Vehicles

Fuel Consumption An increase in IRI of 63.4 in/mi 2% 2% (Low Speeds)

1% (Normal Speeds)

Repair and Maintenance

Up to IRI of 190.2 in/mi 0% 0%

An increase in IRI up to 253.6 in/mi 10% 10%



Tire Wear An increase in IRI of 63.4 in/mi 1% 1%

Source: Chatti, K., and Zaabar, I. Estimating the Effects of Pavement Condition on Vehicle Operating Costs. NCHRP

Report 720, Transportation Research Board (TRB), 2012.

The change in vehicle operating cost per vehicle due to pavement conditions can be calculated, by using the

following equations.

∆𝑽𝑶𝑪𝒗,𝒔,𝒕 = ∆𝑽𝑶𝑪𝒗,𝒔,𝒕𝒇𝒄

+ ∆𝑽𝑶𝑪𝒗,𝒔,𝒕𝒓𝒎 + ∆𝑽𝑶𝑪𝒗,𝒔,𝒕

𝒕𝒘 (15)

where,

∆𝑽𝑶𝑪𝒗,𝒔,𝒕𝒇𝒄

= 𝑽𝑶𝑪𝒗𝒇𝒄

×∆𝑰𝑹𝑰𝒔,𝒕×𝑷𝒗×𝑳𝒔 (16)

∆𝑽𝑶𝑪𝒗,𝒔,𝒕𝒓𝒎 = 𝑽𝑶𝑪𝒗

𝒓𝒎×∆𝑰𝑹𝑰𝒔,𝒕×𝑷𝒗×𝑳𝒔 (17)

∆𝑽𝑶𝑪𝒗,𝒔,𝒕𝒕𝒘 = 𝑽𝑶𝑪𝒗

𝒕𝒘×∆𝑰𝑹𝑰𝒔,𝒕×𝑷𝒗×𝑳𝒔 (18)

∆𝑉𝑂𝐶𝑣,𝑠,𝑡 = the change in annual vehicle operating cost per vehicle due to pavement conditions, for

vehicle type v on road segment s in year t

∆𝑉𝑂𝐶𝑣,𝑠,𝑡𝑓𝑐

= the change in fuel cost per vehicle due to pavement conditions, for vehicle type v on road

segment s in year t

∆𝑉𝑂𝐶𝑣,𝑠,𝑡𝑟𝑚 = the change in repair and maintenance cost per vehicle due to pavement conditions, for

vehicle type v on road segment s in year t

∆𝑉𝑂𝐶𝑣,𝑠,𝑡𝑡𝑤 = the change in tire wear cost per vehicle due to pavement conditions, for vehicle type v,

road segment s, in year t


= the fuel cost per vehicle ignoring pavement conditions, for vehicle type v

𝑉𝑂𝐶𝑣𝑟𝑚 = the total repair and maintenance cost per vehicle ignoring pavement conditions, for vehicle

type v



𝑉𝑂𝐶𝑣𝑡𝑤 = the total tire wear cost per vehicle ignoring pavement conditions, for vehicle type v

∆𝑰𝑹𝑰𝒔,𝒕 =𝑰𝑹𝑰𝒔,𝒕−𝟔𝟑.𝟒

𝟔𝟑.𝟒 (19)

𝐼𝑅𝐼𝑠,𝑡 = International Roughness Index for road segment s in year t

𝑃𝑣 = percent change in cost for vehicle type v, per Table 5.1

𝐿𝑠 = length of segment s (in miles)

𝑠 = subscript representing each of the examined road segments

𝑣 = subscript indicating vehicle type


Then, the change in annual vehicle operating cost due to pavement condition of segment s in year t is

calculated using the following formulae.

∆𝑨𝑽𝑶𝑪𝒗,𝒔,𝒕 = ∆𝑽𝑶𝑪𝒗,𝒔,𝒕×𝑨𝑨𝑫𝑻𝒗,𝒔,𝒕×𝑾𝑫𝒗,𝒔,𝒕 (20)

where,

∆𝐴𝑉𝑂𝐶𝑣,𝑠,𝑡 = the total change annual operating costs for vehicle type v on road segment in year t

∆𝑉𝑂𝐶𝑣,𝑠,𝑡 = the change in annual vehicle operating cost per vehicle due to pavement conditions, for

vehicle type v, road segment s, in year t

𝐴𝐴𝐷𝑇𝑣,𝑠,𝑡 = the annual average daily traffic, for vehicle type v, road segment s, in year t

𝑊𝐷𝑣 = working days a year for vehicle type v


𝑣 = subscript indicating vehicle type


6.1.2 Pavement Condition Impacts on Travel-Time Costs

Similar to additional user operating costs, this analysis estimates travel delay costs accruing to highway

users due to the deteriorating pavement condition.



A study on the impacts of pavement roughness on free-flow speed conducted by the Pavement Research

Center of the University of California (UC) for the California Department of Transportation (Caltrans)7 shows

that pavement roughness accounts for a very small portion of the total speed variance. When holding all

other variables of the roadway constant, this study shows that one unit of IRI change (63.4 in/mi) only leads

to about 0.30 mph decrease in free-flow speed. Another study that built a regression model of highway

speeds using 72 sites near Ontario, Canada8, reveals a drops of about 1.95 mph when IRI increases from

63.4 in/mi to 128 in/mi and about 1.14 mph when IRI increases from 128 to 190 in/mi when all other variables

of the roadway are held constant.

This analysis uses the findings from the Caltrans study, which is considered a conservative approach, to

estimate the impact of pavement roughness on travel-time delay for highway users.

Travel-Time Delay

Free-flow speeds by highway functional class, which can be provided by the Travel Demand Model, are used

to estimate the changes in free-flow speeds by road segment due to pavement condition. Then, the travel-

time delay for segment s in year t is calculated using the formulae below.

∆𝑻𝑻𝒔 = 𝑻𝑻𝒔𝑹𝑭𝑭 − 𝑻𝑻𝒔

𝑭𝑭 (21)

where,

∆𝑇𝑇𝑠 = travel-time delay per vehicle (passenger or commercial) due to the pavement condition of

segment s, in hours per vehicle

𝑇𝑇𝑠𝐹𝐹 = travel-time (in hours) under free-flow speed conditions, for segment s, such that

𝑻𝑻𝒔𝑭𝑭 =

𝑳𝒔

𝑭𝑭𝒔 (22)

𝑇𝑇𝑠𝑅𝐹𝐹 = travel-time (in hours) under reduced free-flow speed conditions, for segment s, such that

𝑻𝑻𝒔𝑹𝑭𝑭 =

𝑳𝒔

𝑹𝑭𝑭𝒔 (23)

𝐹𝐹𝑠 = free-flow speed (in miles per hour), for segment s

𝑅𝐹𝐹𝑠 = reduced free-flow speed (in miles per hour), for segment s, and

𝑹𝑭𝑭𝒔 = 𝑭𝑭𝒔 − 𝟎. 𝟑×∆𝑰𝑹𝑰𝒔 (24)

𝐿𝑠 = length of segment s, (in miles)

7 Wang, T., Harvey, J., Lea, J., and Kim, C. Impact of Pavement Roughness on Vehicle Free-Flow Speed. Prepared by

the Research Center of the University of California, UC Davis and UC Berkeley. Prepared for the California Department of Transportation. July, 2013.

8 Karan, M. A., R. Kher, and R. Haas. Effects of Pavement Roughness on Vehicle Speeds. Transportation Research

Record, No. 602, 1976, pp. 122-127.



∆𝑰𝑹𝑰𝒔 =𝑰𝑹𝑰𝒔−𝟔𝟑.𝟒

𝟔𝟑.𝟒 (19)

𝐼𝑅𝐼𝑠 = International Roughness Index for segment s


Value of Time

Generally, the economic value of travel-time delay costs are estimated for work and business trips made by

passenger cars and commercial vehicles, because the associated time is considered "on-the-clock" time.

Leisure-related trips are assumed to have zero economic value since travel time for leisure-related trips

represent no on-the-clock time and thus represent opportunity costs as opposed to economic costs.

Historical data of hourly wage rates are applied in the calculations of the dollar value of travel time of

highway users. In forecasting the wage rates for private auto users and freight for the analysis period, the

historical data of hourly wage rates are extrapolated Wage rates forecasts for “all occupations” and “truck

drivers”, calculated based on the hourly wage and the historical average annual (compound) growth rate, are

used in estimating the dollar value of travel time delays of projected trips made by passenger cars and

trucks, respectively, in the analysis period.

Calculating Impacts to Travel-Time Costs

Average vehicle occupancy rates are typically available from travel demand models. These along with the

travel time delay and value of time, described in the sections above, can be used to calculate the change in

annual travel time delay due to pavement conditions, using the formulae below.

∆𝑨𝑻𝑻𝑪𝒗,𝒔,𝒕 = ∆𝑻𝑻𝒔 × 𝑨𝑨𝑫𝑻𝒗,𝒔,𝒕 × 𝑽𝑶𝑻𝒗 × 𝑽𝑶𝑹𝒗 × 𝑾𝑫𝒗 (25)

where,

∆𝐴𝑇𝑇𝐶𝑣,𝑠,𝑡 = change in annual travel-time cost (in nominal dollars) due to pavement condition of segment s,

in year t, for vehicle type v

∆𝑇𝑇𝑠 = travel-time delay per vehicle (passenger or commercial) due to the pavement condition of segment s,

in hours per vehicle

𝐴𝐴𝐷𝑇𝑣,𝑠,𝑡 = the annual average daily traffic, for vehicle type v, road segment s, in year t

𝑉𝑂𝑇𝑣 = value of time for vehicle occupants (in dollars per hour) for vehicle type v

𝑉𝑂𝑅𝑣= average vehicle occupancy rate (in count of people), for vehicle type v

𝑊𝐷𝑣 = working days a year for vehicle type v



𝑣 = superscript indicating vehicle type (passenger or commercial)



6.1.3 Pavement Condition Impacts on Motor Vehicle Crash Costs

A research study prepared for the Road Safety Trust of New Zeeland on the relationships between road

safety and road roughness parameters reveals that by reducing the IRI by 63.4 in/mi, motor vehicle crash

rates can be reduced by about 25 percent on roads with average travel speed of 43.5 mph or greater.9

Another study on the effects of asphalt pavement conditions on traffic accidents in the state of Tennessee

using data from four urban interstates with asphalt pavements, divided median types and 55 mph speed

limits shows that the pavement roughness was a significant variable in all types of “accident models” that

were developed to explain the relationship between accident frequency and pavement roughness.10

Specifically, the “accident models” indicate that the crash frequency could increase by 1.649 times if the IRI

increases from 0-100 in/mi to 101-200 in/mi.

Based on the findings from the literature review this analysis assumes that an increase in IRI of 63.4 in/mi

will increase crash rates by 25 percent on road segments that rate in poor pavement condition and report

average travel speeds equal or greater than 43.5 mph. This assessment also uses the outputs of travel

characteristics generated by the Travel Demand Model to identify the road segments that meet the travel

speed criterion.

Increase in Crash Rates

To estimate the increase in crash rates by road segment, this analysis uses the motor vehicle crash rates,

that is, total motor vehicle crashes/100 million vehicle miles traveled (VMT) at the regional-level. Motor

vehicle crashes fall into one of three categories, and a crash rate is calculated for each. The categories are,

Property Damage Only

Injury Crashes

Fatal Crashes

Where available, county-level data is used to calculate regional crash rates. Where county-level data is not

available, market-level or state-level data is used. Then, the change in traffic crash rates caused by

pavement condition for each road segment is estimated as follow:

∆𝑪𝑹𝒄,𝒔 = 𝑪𝑹𝒄,𝒔 ×𝟎. 𝟐𝟓 × ∆𝑰𝑹𝑰𝒔 (26)

where,

∆𝐶𝑅𝑐,𝑠= Increment in the traffic crash rate (crashes/100 million VMT) for the crash category c of the

region where the segment s is located.

9 Cenek, P. and Jamieson, N. Quantification of Safety Benefits Resulting from Road Smoothing. Opus International

Consultants Limited. August 2011 (Revised June 2012).

10 Chan, C., Huang, B., Yan, X., and Richards, S. Effects of Asphalt Pavement Conditions on Traffic Accidents in Tennessee Utilizing Pavement Management System (PMS). Seed Grand Final Report Submitted to Southeastern Transportation Center University of Tennessee, Knoxville. August, 2008.



𝐶𝑅𝑐,𝑠 = the current traffic crash rate (crashes/100 million VMT) for the crash category c of the region

where the segment s is located.

∆𝑰𝑹𝑰𝒔 =𝑰𝑹𝑰𝒔−𝟔𝟑.𝟒

𝟔𝟑.𝟒 (19)

𝑠 = subscript representing each of the examined road segments with average travel speed equal or

greater than 43.5 mph

𝑐 = subscript indicating the category of vehicle crash

Cost per Crash

The economic valuation of fatalities and injuries is not necessarily a simple endeavor. In this methodology

the values produced by the U.S. Department of Transportation11 are used because they are generally

accepted as reasonable and unbiased. Similarly, the U.S. Department of Transportation economic value of

property damage only crashes12 is used. Table 5.2 displays the costs per crash used in this analysis, in 2015

dollars.

Table 6.2 Economic Cost per Crash by Type

Crash Category Economic Cost

(2015 Dollars per Crash)

Property Damage Only $4,198

Injury Crashes $174,030

Fatal Crashes $9,600,000

Source: U.S. Department of Transportation

Calculating Impacts to Crash Costs

The annual (injury and property damage only) crash cost for passenger cars due to pavement conditions of

segment s in year t is calculated using the formulae below.

∆𝑨𝑺𝑪𝒄,𝒔,𝒕 = ∆𝑪𝑹𝒄,𝒔 ×𝑽𝑴𝑻𝒔,𝒕 ×𝑺𝑪𝒄 ×𝑾𝑫 (27)

where,

∆𝐴𝑆𝐶𝑐,𝑠,𝑡 = the annual change in traffic crash cost for vehicles (in nominal dollars) due to pavement

conditions for crash category c on road segment s in year t.

∆𝐶𝑅𝑐,𝑠 = the change in the traffic crash rate (crashes/100 million VMT) for the crash category c of the

region where the segment s is located.

11 Guidance on Treatment of the Economic Value of Statistical Life (VSL) in U.S. Department of Transportation Analysis – 2015 Adjustment. U.S. Department of Transportation, June, 2015

12 The Economic and Societal Impact of Motor Vehicle Crashes, 2010 (Revised). U.S. Department of Transportation,

National Highway Traffic Safety Administration, May, 2015



𝑉𝑀𝑇𝑠,𝑡 = Daily average vehicle miles travelled (in 100 millions) for the road segment s in year t.

𝑆𝐶𝑐 = the economic safety cost in dollars per crash (from Table 5.2), for crash category c on road

segment s

𝑊𝐷 = working days a year

𝑠 = subscript representing each of the examined road segments with average travel speed equal or

greater than 43.5 mph

𝑐 = subscript indicating the category of vehicle crash

6.1.4 Total Direct Impacts

The total direct user impacts arising from pavement conditions is equal to the sum of changes in vehicle

operating costs, cost of travel time changes, and the cost of changes in crashes.

∆𝑨𝑼𝑪𝒕 = ∆𝑨𝑽𝑶𝑪𝒕 + ∆𝑨𝑻𝑻𝑪𝒕 + ∆𝑨𝑺𝑪𝒕 (28)

Where:

∆𝐴𝑈𝐶𝑡 = Total change in annual user costs resulting from pavement conditions in year t

∆𝐴𝑉𝑂𝐶𝑡= the change in annual vehicle operating cost due to pavement conditions, for all vehicle types and

road segments, in year t

∆𝐴𝑇𝑇𝐶𝑡 = the change in annual travel-time delay cost (in nominal dollars) due to pavement conditions, for all

vehicle types and road segments, in year t

∆𝐴𝑆𝐶𝑡 = the change in annual safety cost for vehicles (in nominal dollars) due to pavement conditions, for all

crash categories on all road segments, in year t


6.2 Impacts of Changing Congestion Levels

Failure to maintain the transportation system’s ability to provide safe, efficient mobility of goods and people

can lead to lost economic activity and opportunities. As shown in Figure 6.2, lack of investment can lead to

worsening conditions, including increased traffic levels and congestion and increases in crashes. In turn, this

leads to increases in travel times and overall transportation costs for residents and business. As

transportation costs increase, the region may become less attractive in terms of business expansion,

retention and recruitment.



Figure 6.2 Impact of Deteriorating Transportation Infrastructure

Changes in pavement condition will impact congestion through changes in VMT, VHT, speed, and delay experienced by users of the road network. These changes can be translated into direct economic impacts using the same methods as show in above in Section 3.0. Like the impacts of roadway projects, the direct user impacts of congestion fall into four major categories,

1. Changes in Time Travel Costs (see Section 3.1)

2. Changes in Vehicle Operating Costs (see Section 3.2)

3. Changes in Safety Costs (see Section 3.3)

4. Changes in Air Quality Costs (see Section 3.4)

Each of these can be calculating using the equations from the references section. The relevant equations are listed below for reference.

6.2.1 Changes in Travel Time Costs

Changes in travel time costs arising from congestion can be calculated using Equation 1 from Section 3.1.

∆𝑻𝑻𝑪𝒑,𝒕 = 𝑽𝑶𝑻𝒑× ∆𝑽𝑯𝑻𝒑,𝒕× 𝑽𝑶𝑹𝒑 (1)

Where:

∆𝑇𝑇𝐶𝑝,𝑡 = the change in travel time costs of trip purpose p in year t

𝑉𝑂𝑇𝑝 = the value of time for the study region, by trip purpose

∆𝑉𝐻𝑇𝑝,𝑡 = the change in vehicle hours traveled, by trip purpose in year t

𝑉𝑂𝑅𝑝 = the average vehicle occupancy rate by trip purpose





6.2.2 Changes in Vehicle Operating Costs

Changes in vehicle operating cost can be calculated using equation 2, 3a and 3b from Section 3.2.

∆𝐀𝑽𝑶𝑪𝒗,𝒕𝒏𝒇𝒄

= ∆𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒏𝒇𝒄

(2)

Where:

∆A𝑉𝑂𝐶𝑝,𝑡𝑛𝑓𝑐

= the annual change in non-fuel vehicle operating costs, for vehicle type v in year t


𝑉𝑂𝐶𝑣𝑛𝑓𝑐

= the non-fuel operating costs per vehicle mile, for vehicle type v




= 𝑭𝑪𝑹𝒗× ∆𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒇𝒄

(3a)

Where:



𝐹𝐶𝑅𝑣 = the average fuel consumption rate for vehicle type v







= ∆𝑭𝑪𝑹𝒗× 𝑽𝑴𝑻𝒗,𝒕× 𝑽𝑶𝑪𝒗𝒇𝒄

(3b)

Where:



∆𝐹𝐶𝑅𝑣 = the change in fuel consumption rate for vehicle type v

𝑉𝑀𝑇𝑣,𝑡 = the vehicle miles travelled, for vehicle type v in year t







6.2.3 Changes in Safety Costs

Changes in safety costs can be calculated using equations 4a and 4b from Section 3.3.

∆𝑨𝑺𝑪𝒄,𝒕 = 𝑪𝑹𝒄×∆𝑽𝑴𝑻𝒕 ×𝑺𝑪𝒄 (4a)

Where:


𝐶𝑅𝑐 = the crash rate for crash type c





∆𝑨𝑺𝑪𝒄,𝒕 = ∆𝑪𝑹𝒄×𝑽𝑴𝑻𝒕 ×𝑺𝑪𝒄 (4b)

Where:







6.2.4 Changes in Air Quality Costs

Changes in air quality cost can be calculated using equations 5a and 5b from Section 3.4.

∆𝑨𝑨𝑸𝑪𝒆,𝒗,𝒕 = 𝑬𝑹𝒆,𝒗× ∆𝑽𝑴𝑻𝒕 ×𝑨𝑸𝑪𝒆 (5a)

Where:


𝐸𝑅𝑒,𝑣 = the rate of emissions for emissions type e from vehicle type v








∆𝑨𝑨𝑸𝑪𝒆,𝒗,𝒕 = ∆𝑬𝑹𝒆,𝒗× 𝑽𝑴𝑻𝒕 ×𝑨𝑸𝑪𝒆 (5b)

Where:


∆𝐸𝑅𝑒,𝑣 = the change in the rate of emissions for emissions type e from vehicle type v








7.0 Logistics Costs Impact Module

Logistics refers to the maintenance of the flow of freight from origins to destinations in a timely and efficient

manner. Logistics costs are those costs incurred while performing this function and include expenses such

as chassis, truck, and storage rental, as well as the payment of fines for late delivery, or even the potential

loss of customers if the firm cannot function competitively. In addition to changes in travel-time and related

vehicle operating costs, freight haulers sustain additional changes in logistics costs. These logistics costs fall

into two categories, those costs borne by carriers and those borne by shippers.

Carrier costs will change with changes in required buffer times. This additional operating time involves costs

in terms of labor and equipment. The changes to travel times and buffer times for the carriers will translate to

a change in costs for the shippers in the form of inventory costs. The following sections will provide a

methodology for calculating both the change in costs to carriers and the change in costs to shippers. The

following figure (Figure 7.1) outlines the methodology for calculating the change in logistics costs.

Figure 7.1 Change in Logistics Costs

7.1 The Change in Carrier Costs

In order to ensure timely delivery, carriers include extra time (buffer time) in their schedule estimates. The

buffer time will fluctuate with the reliability of travel times. A history of reliable travel times will allow carriers to

safely reduce their buffer times, while an increase in volatility of travel times will require an increase in buffer

times. The impact of the change in travel times is already included in the changes in travel efficiencies

calculated in the previous module (the Highway Project Module). The following equations are used to

calculate the additional change in carrier’s costs.

∆𝑨𝑪𝑪𝒕 = ∆𝑨𝑩𝑻𝒕 × (𝑽𝑶𝑻𝒕/𝒉𝒓 + 𝑽𝑶𝑪𝒕/𝒉𝒓) (29)

Where:

∆𝐴𝐶𝐶𝑡 = The change in annual carrier costs in year t



𝑉𝑂𝑇𝑡/ℎ𝑟 = the value of time for trucks, usually a wage rate for truckers

𝑉𝑂𝐶𝑡/ℎ𝑟 = the vehicle operating cost for trucks, in dollars per hour.


b = subscript indicating the dollar year in which the relevant cost is given

∆𝐴𝐵𝑇𝑡 = the change in annual buffer time in year t

∆𝑨𝑩𝑻𝒕 = 𝑽𝑯𝑻𝒕𝒕𝒐𝒅 ×(𝑩𝑰𝒇 − 𝑩𝑰𝒏) (30)

Where:

𝑉𝐻𝑇𝑡𝑡𝑜𝑑 = the annual vehicle hours traveled by trucks with origins and/or destinations within the study

area in year t

BI = the buffer index. This is a coefficient of travel time, used to calculate the time cushion that

travelers must add to ensure on time arrival 95 percent of the time.

𝐵𝐼𝑛 = the buffer index for the study are in the no-build scenario, before the project. Regional buffer

indices are published by the Federal Highway Administration (FHWA) in the National Performance

Measure Research Data Set (NPMRDS).

𝐵𝐼𝑓 = the buffer index for the study area in the full-build out scenario, after the project has been

completed.

Where:

𝑩𝑰𝒇 = 𝑩𝑰𝒏 (𝑫𝒆𝒍𝒂𝒚𝒇

𝑫𝒆𝒍𝒂𝒚𝒏) (31)

Where:

𝐷𝑒𝑙𝑎𝑦𝑛 = the delay for the study area under in the no-build scenario, before the project.

𝐷𝑒𝑙𝑎𝑦𝑓 = the delay for the study area under in the full-build out scenario, after the project has been

completed.

Note that these calculations assume that the change in the buffer index is proportionate to the change in

delay.

The change in direct costs resulting from changes in travel times and buffer times are unlikely to be borne by

shippers in the short-run. Eventually there may be a shift in rates that shift some of the cost from the carriers

to the shippers.



7.2 The Change in Shipper Costs

Changes in inventory costs make up the logistics costs borne by shippers as the result of changes in travel

efficiencies. Inventory costs are time sensitive, in that they will increase as the inventory time increases.

Carrier turn times will impact the inventory time – if turn times increase, inventory times will increase by the

same amount, and if turn times decrease, inventory times will decrease by the same amount. The change in

inventory costs can be computed using the following equation.

∆𝑨𝑰𝑪𝒕 = 𝑰𝑹𝒉𝒓× 𝑨𝑽𝒕𝒊 × ∆𝑨𝑰𝑻𝒕 (32)

Where:

∆𝐴𝐼𝐶𝑡 = the annual change in inventory cost in year t

𝐼𝑅ℎ𝑟 = the hourly rate charged for inventory. This rate is often an annual rate, calculated as some

percentage of the value of the inventory. In this case the hourly rate is calculated from the annual,

using 365 day of 24 hours each.

𝐴𝑉𝑡𝑖 = the annual value of the time sensitive inventory being shipped via truck, that has an origin

and/or destination within the study area during year t


∆𝐴𝐼𝑇𝑡 = the change in inventory time.

∆𝑨𝑰𝑻𝒕 = ∆𝑽𝑯𝑻𝒕𝒕𝒐𝒅 + ∆𝑨𝑩𝑻𝒕 (33)

Where:

∆𝑉𝐻𝑇𝑡𝑡𝑜𝑑 = the annual change in vehicle hours traveled by truck with origins and/or destinations

within the study area during year t

∆𝐴𝐵𝑇𝑡 is the change in annual buffer time in year t, calculated above (in Section 6.1).



8.0 Integration with Economic Models

After having considered transportation efficiency benefits of roadway capacity, operational, state-of-good-

repair, or freight rail improvements with a Cost-Benefit Analysis, the next step is using the direct user

benefits as inputs to estimate the economic impact analysis at the local, regional, or state level. Economic

impact is usually calculated using a model such as REMI, IMPLAN, Transight, or TREDIS. These models

measure the impact of a transportation project by industry on output, value added, taxes, employment, and

personal income from changes in spending patterns or improved market access. These models consider

business productivity and economic development impacts that are not represented in transportation system

efficiency tools.

The inputs required by the economic impact model are summarized into a file that feeds into the software.

These inputs include the project costs, reduced travel time and vehicle operating costs, change in shipping

costs, alleviation of traffic congestion, and improved transportation system reliability. Other benefits such as

reduction in emissions have no multiplier effect in the regional economy.

The model uses input-output tables (or economic multipliers) to estimate the direct, indirect, and induced

effects on the economy and on employment, including short-term (construction) and long-term (operations)

jobs created or maintained, in the case of state-of-good-repair projects:

- Direct Impacts: Are assessed only in the industries immediately affected by the investment. For

example, operation and maintenance jobs in roadways and railways;

- Indirect Impact: It reflects the impact on industries that supply intermediate goods, that is, the

demand created for suppliers and contractors in terms of income and employment.

- Induced Impacts: They consider increased household income and their spending behavior. For

example, as more jobs are added to the economy or more disposable income is made available,

individuals will consume more and generate additional economic activity for industries in general.

ITTS Freight Economic Analysis Tool - ittsresearch.orgittsresearch.org/feat/ITTS FEAT - Methodology_Ver 1_11302016 .pdf · reduction in shipping costs as a result of a more efficient

Documents