Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
Planning · Environment · Real Estate

HEP Events Guidance Publications Glossary Awards Contacts

Vermont Agency of Transportation (VTrans) Statewide Travel Model Peer Review Report

Overview of the Vermont Statewide Model

The following appendix summarizes the version of the VTrans model at the time of the peer review, along with the data sources used in the development of the model.

Model Components

The following sections summarize models components from the model documentation current to the timing of the model review. The model is made up of two models, freight and passenger, which are each comprised of four primary modules that are shown in the figure below:

4.0.1 Trip Generation

The trip-generation module combines TAZ-based land-use characteristics with town-based fractions of number of workers divided by number of workers cross-classifications to calculate home-based trips produced by each internal TAZ. The module then calculates trip attractions for each internal TAZ by purpose and trip-productions for the non-home-based (NHB) purpose using purpose-specific regression equations, each of which utilizes a different set of employment and/or population field(s) from the TAZ characteristics table. For example, the equation for home-based work (HBW) trips attracted is based on all of the employment fields in the TAZ characteristics table, but the equation for home-based shopping (HBSHOP) trips is based solely on the retail employment field. Truck (TRUCK) productions and attractions are calculated simply by multiplying the truck percentages from the TAZ characteristics table by the production and attraction totals for the other four trip purposes.

Productions and attractions for zones external to Vermont are calculated by first applying external TRUCK trips as the ADT for the external zones listed in the TAZ characteristics table (presumably taken from traffic counts) multiplied by the truck percentages from the TAZ characteristics table. These values are split evenly as productions and attractions. The total for other external vehicle-trips is taken as the remaining fraction of the ADT for each external zone listed in the TAZ characteristics table. The external vehicle occupancy rate (as an input) is applied to this total to derive non-TRUCK external person-trips. Total non-TRUCK external person-trips are then subdivided by the other trip purposes using the fractions in the external trip-fractions table.

Ultimately, this process outputs a table of productions and attractions for each of the five trip purposes in the model (HBW, HBO, HBSHOP, NHB, TRUCK) for each of the 936 internal and external zones. However, because the production and attraction estimates for the internal TAZs came from different sources for each of the four home-based trip purposes, they do not match. This mismatch is typical for most demand-forecasting models where separate regression models are estimated for production and attraction across a full study area with unique predictor variables. Balance factors are calculated as the ratio of trip productions destined for internal zones to the corresponding trip attractions in internal zones by trip purpose. Balancing is accomplished by zone by multiplying the balancing factors to the internal trip attractions only so that they match total productions (internal and external) by trip purpose. The end result is a table of balanced productions and attractions for each of the five trip purposes in the model for each zone. Figure D1 provides a visual summary of the trip generation process.

The figure provides a flow chart summary of the VTrans' model trip generation process. The entire process stems from TAZ-based characteristics, including truck percentages, number of households, jobs, daily traffic counts, and area type. Household characteristics (e.g. persons/workers) by TAZ stem from TAZ based-characteristics through a trip-rate table, town-based household characteristics, and non-home-based production/attraction regression equations to calculate trip productions for non-truck trip purposes for internal TAZs. Meanwhile, households, jobs, and area type stem through both home-based regression attraction equations and non-home-based production/attraction regression equations to calculate trip attractions by trip purpose for internal TAZs. These trip attractions by trip purpose for internal TAZs, as well as the trip productions for non-truck trip purposes, feed into the trip table of all productions and attractions by trip purpose for all TAZs. To calculate external truck trips, truck percentages and AADT stem from the TAZ based-characteristics and are used to determine truck productions and attractions for external TAZs. Applying a 50% rate for each truck productions and truck attractions, it is assumed that the remainder AADT for non-truck purposes for external TAZs. Fractions for non-truck purposes for external TAZs are then applied to determine productions and attractions by trip purpose for external TAZs. To calculate internal truck trips, truck percentages stem from the TAZ-based characteristics to origin-destination estimation to achieve truck purpose trip productions and attractions for internal TAZs. These trip productions and attractions for internal TAZs and productions and attractions by trip purpose for external TAZs then feed into the same trip table of all productions and attractions by trip purpose for all TAZs as the non-truck trips to complete trip generation.

Figure D1: Trip Generation Process

4.0.2 Trip Distribution

The trip-distribution sub-module takes the balanced trip table, a matrix of free-flow travel times between TAZs and a set of impedance functions to develop a matrix of productions and attractions between all zones. The result of this step is a matrix of productions and attractions between all zones. The final step in the trip-distribution application is to convert this matrix into a matrix of origin-destination (O-D)-based trips. Since the model is a daily model, all trips are assumed to return, indicating that all trips originating in one zone and destined for another must also originate in the destination zone and terminate in the origin zone. This assumption requires that the final matrix be diagonally symmetric. To accomplish this, the matrix is transposed, added to the original, and then all cells are halved. The result is a diagonally-symmetric O-D matrix of person-trips. Figure D2 provides an illustration of the tip distribution process.

The figure provides a flow chart summary of the VTrans' model trip distribution process. The process begins with the trip table of all productions and attractions by trip purpose for all TAZs. Next, balancing factors are calculated by trip purpose to adjust internal attractions up or down. The balancing factor equation is: the total of internal productions plus external productions minus external attractions divided by internal attractions. Once the balancing factors are applied, the result is the balanced trip table, which includes total productions equal to total attractions by trip purpose for all TAZs. Free flow travel times between TAZs are then applied to the balanced trip table, as well as trip distribution equations by trip purpose. The trip table then feeds into trip distribution using a production-constrained gravity model. Application of this gravity model results in an origin matrix of productions and attractions by TAZ for each trip purpose. This matrix is then transposed, and the original matrix is added to the transposed matrix. These combined matrices are then divided by two to achieve a diagonally-symmetric daily person-trip matrices for all trip purposes, including home-based work, home-based shopping, home-based other, non-home-based, and truck.

Figure D2: Trip Distribution Process

4.0.3 Mode Choice

In the past, the O-D matrix of person-trips was reduced by the expected transit demand before allocating the remaining trips to passenger vehicles. However, the existing matrix of transit demand may date back as far as 1997, and no definable data source for the transit demand could be located, and the 2009 NHTS does not support the development of full O-D matrix of transit demand statewide. Therefore, transit demand is no longer considered directly in the Model. Instead, the full O-D matrices resulting from the trip-distribution step are divided by a vehicle-occupancy to convert them from person-trips to passenger vehicle-trips.

4.0.4 Trip Assignment

The final matrix, including all external vehicle-trips, is assigned to the road network in the traffic assignment sub-module. Free-flow travel speed on each link is assumed to be the five miles per hour over the speed limit, and the user-equilibrium traffic assignment is used.

Model Comparison Summaries

Table D1: Comparison of the Classification of Vermont Households by Size

No. of People in Household NHTS Model
1 28% 28%
2 38% 38%
3 15% 15%
4 13%  
5 5% 19%
6 1%  

Table D2: Comparison of the Classification of Vermont Household by Number of Workers

No. of Workers in Household NHTS Model
0 25% 25%
1 40% 36%
2 30% 33%
3 4% 6%
4 1%

Table D3: Comparison of the Relationship Between Link Volumes and Traffic Counts

Roadway Category Model Result for Volume/Count Acceptable Standard for Volume/Count
Freeway -15.2% +/- 7%
Divided Arterial +1.0% +/- 15%
Undivided Arterial -13.7% +/- 15%
Collector -3.0% +/- 25%

Model Data Sources

The following section provides brief descriptions of the sources of data used in the model.

4.2.1 Demographic Data

US Census - 2010 Population and Household Data

Bureau of Economic Analysis/Vermont Department of Labor - 2009 Employment Estimates

American Community Survey - 2006 to 2010 Number of Workers and Number of Household Members by Town

National Household Travel Survey - 2009 Household Travel Information (Person Trip Table)

4.2.2 Highway Network/Traffic Volume Data

Vermont Center for Geographic Information - Speed Zone Layer along Interstates and State and Federal Highways in Vermont

VTrans Project Information - Preliminary List of Major Roadway Projects whose Construction Began Prior to 2011

VTrans - 2009 Statewide Traffic Counts - Average Annual Daily Traffic (AADT)

Updated: 03/28/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000