GreenSTEP%20Statewide%20Transportation%20Greenhouse%20Gas%20Model

1
GreenSTEP Statewide Transportation Greenhouse
Gas Model

• Cutting Carbs Conference
• December 3, 2008
• Brian Gregor
• ODOT Transportation Planning Analysis Unit

2
What are models?
3
We Are All Modelers
Any time that logic is used to predict the
consequences of decisions, a model is used to
structure the reasoning process. Most often that
is a mental model.
Whatever shall we do?
4
Conflicts Arise from Differences in Mental Models
Sprawl
Highway Expansion
Highway Expansion
More Travel
More Travel
Sprawl
Youre wrong!
No You Are!
5
Formal Models are Needed for Analyzing Complex
Systems

• Well structured models
• Can account for many complex interactions
• Maintain logical consistency in addressing
  interactions
• Provide a more complete accounting of effects
• Allow policies to be tested in a reasonable
  amount of time
• Help to resolve conflicts between differing
  points of view

6
Steps in Model Development Process

• Design
• Define model scope, structure, and components
• Estimation
• Use data to develop mathematical functions and
  algorithms for model components
• Calibration
• Adjust function parameters and algorithms to
  match observed values (e.g. zero vehicle
  households)
• Validation
• Check that overall model behavior is reasonable
• Models are complex, and may be difficult to
  understand, but should not be black boxes.

7
What is GreenSTEP
8
GreenSTEP Model

• Greenhouse gas State Transportation Emissions
  Planning Model
• Why develop GreenSTEP
• The OGWC needs to be able to analyze the effects
  of transportation and land use strategies for
  reducing GHG emissions statewide.
• Existing land use and transportation models cant
  be used to do the required analysis on a
  statewide basis in the time available.
• Statewide scope of analysis required
• A wide range of factors need to be analyzed
• Run-time issues

9
GreenSTEP Model Requirements

• Develop statewide forecasts of GHG emissions from
  transportation sources in response to various
  policy approaches and other factors (e.g. fuel
  prices).
• Be responsive to regional differences including
  differences between metropolitan areas, other
  urban areas and rural areas.
• Is not a substitute for the use of metropolitan
  transportation and land use models for regional
  planning
• Run relatively quickly so that a number of
  iterations of scenario development and testing
  can occur.

10
Factors for Model to Address

• Demographic changes
• Relative amounts of development occurring in
  urban and rural areas
• Metropolitan and other urban area densities and
  urban form
• Amounts of metropolitan area public transit
  service
• Highway capacity
• Vehicle fuel efficiency
• Electric vehicles
• Fuel prices
• Other vehicle pricing
• Demand management
• Congestion effects
• Vehicle operation and maintenance
• Carbon content of fuels including well to
  wheels impacts
• CO2 production from electrical power generation

11
Model Structure
12
GreenSTEP Model Structure
Population Projection by Age Cohort (OEA County
Forecasts)

• Create Synthetic Households by County to
  Represent Population Projection
• Number of persons by age
• Income

Household Age Structure Model
Household Income Model
State Average Per Capita Income Growth
Census Tract Density Model
Urban Growth Boundary Expansion Rates
Calculate Population Densities and Urban
Mixed-Use Characteristics where Households are
Located
Urban Mixed-Use Model
Proportions of Growth Occurring in Metropolitan,
Other Urban, and Rural Areas
Calculate Vehicle Ownership
Vehicle Ownership Model
Urban Mixed-Use Assumptions

• Calculate Vehicle Fleet Characteristics
• Vehicle ages, types by income
• Average MPG
• Gas and diesel proportions
• Proportion of Mileage that is EV

Models of Age and Type of Vehicle by Income Group
Average Fleet MPG by Type
Primary EV Driving Range and Proportion of VMT in
Range to be EV
Model of Daily Miles Driven by Vehicle by
Population Density
Rate of Transit Revenue Mile Growth
Calculate Freeway and Public Transit Supply Levels
Rate of Freeway Lane-Mile Growth

• Household DVMT Models
• Metropolitan Area
• Other Urban and Rural

Calculate Annual Household VMT
13
Fuel Cost Assumptions
Calculate Household Travel Cost Increase Over
2000 Levels and Adjust Household Income
Electric Power Cost Assumptions
VMT Tax Policy Assumptions
Recalculate Vehicle Ownership Based on Adjusted
Household Income
Vehicle Ownership Model

• Household DVMT Models
• Metropolitan Area
• Other Urban and Rural

Recalculate Annual Household VMT Based on
Adjusted Household Income
Demand Management Assumptions
Calculate Demand Management Adjusted VMT
Demand Adjustment Factors
Statewide Population Projection
Calculate Truck VMT
Truck VMT Model
State Average Per Capita income Growth
Vehicle Fleet MPG MPKwh
Public Transit VMT Calculated from Revenue Miles
( above )
Vehicle Maintenance Operations Assumptions
Calculate Fuel Consumption, Electric Power
Consumption, and Greenhouse Gas Emissions
Average Fleet MPG and MPKwh ( calculated above )
Fuel Lifecycle Carbon Content
CO2 Production per KWH
14
Model Sensitivity Testing
15
Test of Model Sensitivity to Land Use and
Transportation Inputs
16
Test of Model Sensitivity to Land Use and
Transportation Inputs
Approximate density of Los Angeles
Approximate density of Atlanta
17
Test of Model Sensitivity to Land Use and
Transportation Inputs
Approximate freeway supply of New York
Approximate freeway supply of Minneapolis
Approximate freeway supply of Houston
18
Test of Model Sensitivity to Land Use and
Transportation Inputs
Approximate transit supply of Seattle
Approximate transit supply of Miami
Approximate transit supply of Detroit
19
Test of Model Sensitivity to Land Use and
Transportation Inputs
All places are urban mixed-use
No places are urban mixed-use
20
Test of Model Sensitivity to Land Use and
Transportation Inputs
21
(No Transcript)
22
Conclusions

• GreenSTEP can be used to evaluate a large number
  of different policies and other factors on GHG
  emissions land use, transportation, prices,
  vehicle characteristics, fuels, etc.
• GreenSTEP evaluates interactions between factors
  e.g. density -gt vehicle ownership -gt vehicle
  travel
• GreenSTEP includes new modeling components that
  we will be able to combine with other models,
  statewide and urban for modeling GHG emissions.