Task Order 5315
Traffic Operations Research

Click here to view: Evaluation_of_HOV_project_112806.ppt

Click here to view: i5hovmixedflowdata.pdf


Evaluation of Incorporating Hybrid
Vehicle Use of HOV Lanes

Will Recker
Institute of Transportation Studies
University of California, Irvine

Introduction

High-Occupancy Vehicle (HOV) lanes have been regarded as a cost-effective and environmentally friendly option to help move people along congested routes. The Federal Highway Administration (FWHA) encourages the installation of HOV lanes as an important part of an area-wide approach to help metropolitan areas address the needs they have identified for mobility, safety, productivity, environmental, and quality of life. In California, since having the first HOV toll bypass on San Francisco-Oakland Bay Bridge in 1970, HOV systems have been extended statewide and currently there are more than 1,100 miles of HOV lanes.

In spite of wide adaptation of policies relating to HOV facilities by many states, MPOs and cities, there still remain questions on the effectiveness of HOV systems. Questions include HOV facility's cost-effectiveness and its impact on air quality. These questions arise from the concern that the conversion to a HOV lane might cause more congestion and higher emission by worsening the traffic condition on general-purpose lanes. In reality, HOV lanes may not be an appropriate option for every situation, and their benefit and impact may vary by the location and situation. However, the benefit of HOV systems has not been well quantified mainly due to the lack of analysis tools quantifying the benefit of HOV systems. In particular, air quality impact has not been accurately measured.

Recently, the Governor of California announced a plan to improve the state's air quality by allowing efficient, gas-electric "hybrid" vehicles on California's more than 1,100 miles of HOV lanes. This HOV/hybrid proposal is expected to cut down the amount of air pollution by encouraging drivers to use less fuel as well as ease traffic congestion through more efficient use of the reserve capacity on the HOV lanes. However, its performance impact both on HOV lanes and on general-purpose lanes has not been analyzed yet. According to one statistic¹, HOV lanes in Orange County reached almost their capacity (1,650 vehicles per hour) by carrying an average of 1,568 vph in 1998 while HOV lane utilizations in LA County and San Francisco Bay Area are 1,013 veh and 930 veh, respectively. This implies that understanding of current and future demand on HOV lanes is the key to success of HOV/hybrid proposal. There is also a need to investigate appropriate changes in HOV operations for the success of HOV/hybrid systems.

The main purpose of this project is to investigate the impact of deploying such HOV/hybrid use systems in California. This project includes three major modeling components:

  • microscopic simulation modeling,
  • emission modeling for HOV/hybrid system, and
  • demand modeling for future hybrid vehicles.

Methodology

The basic approach is to incorporate microscopic traffic simulation models for accurate measures of the system. Even though simplified methods have been widely used in quantifying the impact of HOV system, they have limitations in capturing effects of traffic operational changes. This study will employ a microscopic traffic simulation model that is capable of evaluating the HOV/hybrid system and providing detailed outputs that are not available in conventional static models. In this study, an improved HOV behavior model will be developed to reflect HOV/hybrid lane choice behavior that reflects previous experience as well as driver's perception of traffic conditions in HOV lanes and general lanes. This remedies existing programs' limitations in replicating HOV's driving maneuvers and allows more accurate system performance.

This study also includes detailed emission modeling in order to estimate accurate emissions by integrating emission models into microscopic simulation models. Current computer models lack sufficient detail required to properly predict emissions inventories at different scales. Shortcomings of these models/data include inaccurate characterization of actual driving behavior and a disregard of important vehicle operating parameters that affect emissions. The emission model to be employed in this study is a new generation of models that can accurately predict the energy and air quality impacts of transportation systems, operating at the micro-, meso-, and macro-scale levels-of-detail. Known as the Integrated Transportation/Emissions Modeling (ITEM) system, this suite of models closely integrates both transportation simulation models and advanced modal emissions models. ITEM consists of a hybrid framework of macroscopic/microscopic transportation models that can produce emission inventories for both small, detailed operations as well as large regional areas. This feature will enable the estimation of the environmental impact of HOV/hybrid system.

An important aspect of this study is to predict future hybrid demand. The HOV/hybrid system is expected to promote the use of hybrid vehicles by providing travel time saving and travel reliability. How these benefits affect the hybrid vehicle market is a key component in predicting future hybrid vehicles on HOV lanes. While benefit from HOV/hybrid system is estimated by supply side analysis using microscopic simulation model, the hybrid vehicle demand requires an automobile market analysis. In this study, hybrid demand models are developed based on consumers' automobile choice behavior analysis.

Allowing hybrid vehicles to use HOV lanes can affect traffic congestion, overall fuel usage and vehicle emissions through at least four channels. First, most consumers who already have (or who would have otherwise purchased) hybrids will switch from regular to HOV lanes whenever they can as part of their normal driving. This could affect emissions from other vehicles by reducing congestion in the regular lanes.

Second, hybrid owners might change the routes or destinations of their existing trips so that they can take advantage of HOV lanes. This would shift patterns of traffic congestion among routes. Increased congestion could occur on routes with HOV lanes, both in the HOV lanes and in the regular lanes used to access HOV lanes.

Third, hybrid owners might make more or longer trips since using HOV lanes would reduce their travel times. If these new trips were substituting for trips that would have been made in a non-hybrid household vehicle, then this would lead to reduced emissions and fuel usage. UCI researchers have modeled such effects in terms of household purchase of alternative-fuel vehicles, using stated preference data. Previous results showing a substitution effect for limited-range electric vehicles could be greater for hybrid vehicles without range and refueling restrictions.

Finally, households would be more inclined to purchase a hybrid vehicle by the prospect of reduced travel times from HOV lane usage. This can lead to the largest impact in traffic congestion and reduction in emissions and fuel usage, since the hybrid will replace many trips previously made in gasoline and diesel vehicles. We will also need to account for a rebound effect, whereby reduced operating costs and reduced travel times can lead to an increases in the number of miles driven by the household.

The first two channels can be modeled with standard network calculations if the network assignments were sensitive to time savings from HOV lane use. We would also need to know the location of households owning hybrid vehicles and the O-D matrices for the hybrid drivers. DMV registration data can be used to give the location of current hybrids.

The last two channels require more complex modeling of household vehicle demand and utilization. UCI researchers have specified and estimated relevant models using data from the mid 1990s, but their models were complicated by their emphasis on limited-range electric and natural gas vehicles. Assuming that consumers treat hybrids identical to other high capital and low operating cost vehicles, the models required are similar to those being developed at UCI and UC Davis to model the effects of California's Greenhouse Gas law. We can use these new results to modify the existing UCI models to enhance their accuracy for hybrid vehicles. The updated UCI models can then be applied to data from the recent Caltrans 2000-2001 Statewide Household Travel Survey and the 2001 National Household Travel Survey (NHTS). These survey data allow us to locate the households and trip destinations of likely hybrid vehicle owners. Results from UCI studies of demand for toll lanes have established monetary values of saved travel time that can be applied to estimated time savings from network simulations to forecast incentives for purchase of hybrid vehicles. We will also need to access or develop a supply-side model to estimate availability and prices of hybrid vehicles by body type and manufacturer and price in order to forecast penetration of hybrid vehicles.

Throughout this project, the direct and indirect benefit from the HOV/hybrid system will be estimated and operational suggestions will be provided for the success of the HOV/hybrid systems. A couple of indirect benefits are also expected in this project. This project will also answer to the question on the effectiveness of HOV lanes by providing a formalized tool to quantify the benefit of HOV systems as well as other HOV operational alternatives. The incorporated emission models in this study will also play a role as a California's standardized model in estimating emissions using microscopic simulation models.

¹Legislative Analyst Office, HOV Lanes in California: Are They Achieving Their Goals? January 2000