The use of a regional travel demand model in combination with the ELToD Assignment Model is the preferred method to prepare project traffic forecasts for Express Lanes. ELToD is a stand-alone application that follows the FSUTMS standards and works in conjunction with all Florida’s travel demand models. ELToD was initially calibrated using 2011 data from I-95 Express Phase 1. Since then, ELToD has been re-calibrated against observed data in Broward and Palm Beach Counties on both I-95 Express Phase 1-4 and I-595 Express. Florida’s Turnpike Enterprise has used ELToD for Project Traffic Forecasting for numerous projects around the state in the South Florida, Central Florida, and West Central Florida regions.

ELToD estimates the volume of traffic by hour on both general use and Express Lanes by applying Equation 9-1 to O-D matrices from any travel demand model, including models that already have Express Lanes functionality in the highway assignment process. ELToD also identifies the Express Lanes dynamic toll and congested speeds by hour based on traffic conditions. The ELToD model has a robust choice model component in a controlled environment that can account for various factors that are involved in Express Lanes choice, such as reliability, comfort, safety, lane preference and avoidance of trucks. ELToD provides robust results with quick run times; yet is detailed enough for Project Traffic Forecasting and for Express Lanes ingress/egress traffic operations testing.

Table 9-1 Express Lanes Project Traffic Forecasting Methods Used in Florida
Method Activity/Project
Phase		 Data
Requirements		 Pros Cons
Manual
Estimation Using
Peak Hour O-D		 Sketch-Level
  • O-D data
  • EL Access Points
  • Estimated EL Share
  • Quick estimation method
  • Provides expected volume range for EL Segments and Access Points
  • Supply/demand equilibrium not considered
  • Aggregate effect of multiple O-Ds not considered
Travel Demand Model Based Regional Travel Demand Model (TDM) with Dynamic Toll Function or VTTS Curve Assignment Corridor Planning
/ PD&E / Design
  • EL Access Points
  • Number of EL Lanes
  • Speed & Capacity Info
  • Dynamic Toll Function/ VTTS Curve
  • Estimates Daily or period demand directly from TDM without needing to use another model
  • Provides a systemwide EL evaluation
  • Typically uses generalized cost or predefined share
  • Dynamic toll calculation typically at period level and not at 15-min or hourly levels
  • Toll amount not reported or used
  • VOR not considered
  • Long model run times for alternative testing
  • Post processing needed
Regional Travel Demand Model with ELToD Assignment Model Corridor Planning / PD&E / Design
  • O-D (from TDM)
  • EL Access Points
  • Number of EL Lanes
  • Speed & Capacity Info
  • Toll Choice Parameters (VTTS, VOR, Toll Constant)
  • Proven to be efficient
  • Quick turnaround time for alternatives testing
  • Consistent results in controlled environment
  • Incorporates value of reliability
  • Calculates tolls, congested speeds, and volumes by hour
  • Does not account for queue spillback
  • EL choice selection at first entry only
Microsimulation Model PD&E / Design
  • O-D (from TDM or mesoscopic model)
  • Concept Plans (with EL Access Points and Number of Lanes)
  • Signal Timings
  • Pricing Model Parameters
  • Decision Model Parameters (VTTS, Toll Constant)
  • Pricing model customized to match the Statewide Express Lanes Software tolling algorithm
  • Accounts for complex weaving and geometry
  • Accounts for queue build-up, spillback and dissipation
  • Can be integrated with other multi-resolution tools
  • Extensive time and effort for model development and validation
  • Default model does not account for VOR and requires customized scripting
  • EL choice selection at first entry only
  • Coding difficulties for unconventional ingress/egress ramp combinations
  • EL module cannot be run concurrent with the network-wide microscopic dynamic traffic assignment


In ELToD, the percentage of traffic using the Express Lanes, or Express Lanes share, is calculatedusing the equation below. The equation reflects the latest representation of the toll choice model.

$$Express\:Lanes\:Share = \frac{1}{1+e^{(-1*(\beta_{constant}+\beta_{Time}*Time+\beta_{Toll}*Toll+\beta_{Reliability}*Reliability-Distance \:Penalty))}}$$
Equation 9-1


Where:

FDOT Project Traffic Forecasting Handbook
  • $\beta_{Constant}$ - This parameter determines the Express Lanes share when time, toll, and reliability have a net zero effect.
  • $\beta_{Time}$ - This parameter is for the travel time coefficient in the choice model equation defined in the ELToD Model as the Travel Time Coefficient (with units of 1/min). This is the disutility of increasing travel time by one minute.
  • $\beta_{Toll}$ - This parameter is for the toll cost coefficient in the choice model equation defined in the ELToD Model as the Toll Coefficient (with units of 1/$). This is the disutility of increasing the toll by one dollar.
  • $\beta_{Reliability}$ - This parameter is calculated from a Reliability Ratio (defined in the ELToD Model as the Reliability Coefficient Ratio) and the travel time coefficient. It indicates the disutility of one unit (one minute) of standard deviation.
  • $Distance\: Penalty$ - This parameter is a penalty applied to trips that may use the Express Lanes for a short distance to discourage short Express Lanes trips.

The ELToD Model procedure uses four primary sets of inputs:
  • Total estimated subarea project traffic (in a matrix layout) at a period or daily level.
  • Hourly distribution of total traffic within the project corridor (by direction), based on observed traffic data.
  • Geometric configuration of the subarea network links: link lengths, free flow speed, lane capacity, and link facility type.
  • Toll costs: Pricing policy curve, including toll rate limits (minimum and maximum toll rates).

The steps to use ELToD for project traffic forecasting are as follows: FDOT Project Traffic Forecasting Handbook
  1. Obtain inputs (1, 2, and 3 above) from the validated regional travel demand model.
  2. Obtain input (4 above) from the current FDOT toll pricing policy.
  3. Create a subarea model with Express Lanes and apply ELToD.
  4. Update the ELToD model parameters, consistent with guidance in the ELToD – User Guide and Documentation.
  5. Run the ELToD model.

The following model output data is extracted from the link data, and summarized by hour and direction:
  • Volume
  • Time of Day Percent
  • Express Lanes Share
  • V/C Ratio
  • Congested Speed
  • Tolls
  • Revenue
For more detailed information about the ELToD Assignment Model, please refer to the ELToD – User Guide and Documentation. This report can be requested through Florida’s Turnpike Enterprise, Toll Studies & Forecasting Department.
It should be recognized that there have been some research efforts to evaluate the use of mesoscopic dynamic traffic assignment (DTA) models, such as Cube Avenue and DTALite, for Express Lanes forecasting. However, these methods would require significant programming and calibration efforts to include the ELToD toll choice model methodology. An ELToD DTA version has been developed by Florida’s Turnpike Enterprise, Toll Studies & Forecasting Department and can be made available upon request.