I75 NorthSouth Transportation Initiative Truck Model

1
I-75 North-South Transportation Initiative Truck
Model

• By John Gliebe
• PB Consult
• Albuquerque, NM

2
Presentation Overview

• Development of a base-year model
• Growth forecasts
• Issues and lessons regarding data and
  methodologies

3
I-75 NSTI Background

• Major investment study combining two MPOs
• OKI (Cincinnati)
• MVRPC (Dayton)
• Major initiatives under consideration
• Cincinnati regional light rail system
• Dedicated truck lanes

4
NSTI Study Area
5
TruckModel Structure
6
Database Development

• Freight analysis zone (FAZ) system
• Business employment and household data
• Commodity flow database
• Traffic counts with trucks classified
• Network of freeways and principal arterials

7
Freight Analysis Zone System
8
Developing Base Year Truck Tables

• Original Approach
• Use commodity flow data to generate initial truck
  trip table estimate seed matrix
• Apply matrix adjustment procedure to calibrate to
  truck count data

9
Developing Base Year Truck Tables

• Alternative Approach
• Create seed matrix using Quick Response methods
• (Quick Response Freight Manual, USDOT, 1996)
• Apply matrix adjustment procedure to calibrate to
  truck count data

10
Quick Response Method

• Generate truck trip ends
• Apply modified QRFM coefficients to employment
  and households
• Distribute using gravity model
• Use QRFM friction function parameters
• Expanded Ohio DOT external station survey to
  derive EE, IE and EI trips

11
Truck trip generation coefficients
12
Base-Year Truck Tables
13
Synthetic Matrix Estimation

• Single-path algorithm in TransCAD
• Iterative process
• Network assignment of seed matrix
• Calculate difference between flows and counts on
  shortest path between zones
• Re-factor seed matrix and re-assign to network
• Stop when reach min. gap or max. iterations
• 840 link counts used in final calibration

14
Base Year Model Fit to Truck Counts (all trucks,
1039 obs.)
15
Base Year ModelTrip Length Distributions
mean 34.0 min.
mean 27.5 min.
16
Post-processing Base Year Tables

• Convert from FAZ flows to TAZ flows
• Distribute by time of day

17
Forecast Year Truck Tables

• Calculate growth factors
• Apply trip generation coefficients to 2030
  employment and households by TAZ
• Apply industry productivity inflators
• Calculate 2030 productions and attractions
• Fratar base year tables to match 2030 productions
  and attractions

18
Productivity Deflators

• Industry measures of output per worker over time
• Durable Manufacturing
• Non-durable Manufacturing
• Mining
• Construction
• Transportation, Communications Utilities
• Finance, Insurance Real Estate
• Retail
• Wholesale
• Services
• Agriculture, Fishing and Forestry
• Results in region-wide truck trip growth,
  1995-2030 71 (SU trucks), 77 (MU trucks)
• Without deflators 25 (all trucks)

19
Growth in Multi-Unit Truck Trips
20
Data Issues and Lessons

• Commodity flow data may not be sufficiently
  developed to support intra-regional freight
  modeling
• Need local survey data to calibrate truck trip
  generation and distribution models
• Need more frequent vehicle classification of
  counts
• Need to maintain zonal employment by industrial
  classification

21
Methodological Issues

• How to hand four-tire trucks? (pickups, vans,
  SUVs)
• Comprise a large portion of total commercial
  vehicle traffic
• Do not typically haul freight
• 70 are non-commercial (VIUS, 1997)
• Counters do not distinguish from passenger
  vehicles
• Included in home interview surveys? (NHB-work)
• Synthetic matrix adjustment is not trivial
• Might provide great fit to counts, but can create
  nonsensical table interchanges
• Very sensitive to assignment method
• Not good for congested networks