15. Supplementary Notes Supported by a grant from the Office of the Governor of the State of Texas, Energy Office

Transcription

1 1. Report No. SWUTC/95/ Technical Report Documentation Page 2. Government Accession No. 3. Recipient s Catalog No. 4. Title and Subtitle Quantifying the Benefits of High-Occupancy Vehicle Facilities Using Automatic Vehicle Identification Technology 5. Report Date November Performing Organization Code 7. Author(s) Shawn M. Turner, Gary A. Carlin, and Russell H. Henk 8. Performing Organization Report No. Research Report Performing Organization Name and Address Texas Transportation Institute The Texas A&M University System College Station, Texas Sponsoring Agency Name and Address Southwest Region University Transportation Center Texas Transportation Institute The Texas A&M University System College Station, Texas Work Unit No. (TRAIS) 11. Contract or Grant No Type of Report and Period Covered 14. Sponsoring Agency Code 15. Supplementary Notes Supported by a grant from the Office of the Governor of the State of Texas, Energy Office 16. Abstract This report examines the benefits of high-occupancy vehicle (HOV) lanes in three major freeway corridors in Houston, Texas: the Katy (I-10), Northwest (US 290), and North (I-45) Freeways. The analyses described in this report used eight months of travel time data (April through November 1994) available through Houston s automatic vehicle identification (AVI) traffic monitoring system. The travel time data were used to quantify travel time savings and reliability benefits of the HOV lanes relative to the freeway general-purpose lanes. The travel time data were also used to calibrate a macroscopic freeway simulation model for comparing an HOV lane alternative to other transportation improvements. Various measures of effectiveness, like person delay, fuel consumption, and mobile source emissions, were used to compare alternative improvements. The analyses of 1994 travel time indicated that the three Houston HOV lanes have significant travel time savings, energy consumption savings, and travel time reliability benefits over the freeway general-purpose lanes. The travel time savings were different for each freeway corridor because of different congestion levels, but the day-to-day and month-to-month reliability of speeds in the HOV lane were consistent among all corridors. The extensive amount of travel time data available through Houston s traffic monitoring system provided a clear picture of the daily operation of the three HOV lanes and freeway general-purpose lanes. The computer simulation analyses showed that HOV lanes have significant energy consumption savings over other transportation improvements. The simulation analyses also showed that an HOV lane compared favorably with other transportation improvements, such as providing additional general-purpose lanes or increasing freeway capacity through traffic management practices. 17. Key Words High-occupancy vehicle lane evaluation, automatic vehicle identification, travel time savings and reliability, energy consumption, freeway simulation 18. Distribution Statement No restrictions. This document is available to the public through NTIS: National Technical Information Service 5285 Port Royal Road Springfield, Virginia Security Classif.(of this report) Unclassified 20. Security Classif.(of this page) Unclassified 21. No. of Pages Price Form DOT F (8-72) Reproduction of completed page authorized

2

3 QUANTIFYING THE BENEFITS OF HIGH-OCCUPANCY VEHICLE FACILITIES USING AUTOMATIC VEHICLE IDENTIFICATION TECHNOLOGY by Shawn M. Turner Assistant Research Scientist Gary A. Carlin Graduate Research Assistant and Russell H. Henk, P.E. Assistant Research Engineer Research Report Sponsored by The Office of the Governor of the State of Texas, Energy Office Southwest Region University Transportation Center Texas Transportation Institute The Texas A&M University System College Station, Texas November 1995

4

5 ABSTRACT This report examines the benefits of high-occupancy vehicle (HOV) lanes in three major freeway corridors in Houston, Texas: the Katy (I-10), Northwest (US 290), and North (I-45) Freeways. The analyses described in this report used eight months of travel time data (April through November 1994) available through Houston s automatic vehicle identification (AVI) traffic monitoring system. The travel time data were used to quantify travel time savings and reliability benefits of the HOV lanes relative to the freeway general-purpose lanes. The travel time data were also used to calibrate a macroscopic freeway simulation model for comparing an HOV lane alternative to other transportation improvements. Various measures of effectiveness, like person delay, fuel consumption, and mobile source emissions, were used to compare alternative improvements. The analyses of 1994 travel time indicated that the three Houston HOV lanes have significant travel time savings, energy consumption savings, and travel time reliability benefits over the freeway general-purpose lanes. The travel time savings were different for each freeway corridor because of different congestion levels, but the day-to-day and month-to-month reliability of speeds in the HOV lane were consistent among all corridors. The extensive amount of travel time data available through Houston s traffic monitoring system provided a clear picture of the daily operation of the three HOV lanes and freeway general-purpose lanes. The computer simulation analyses showed that HOV lanes have significant energy consumption savings over other transportation improvements. The simulation analyses also showed that an HOV lane compared favorably with other transportation improvements, such as providing additional general-purpose lanes or increasing freeway capacity through traffic management practices. v

6 ACKNOWLEDGMENTS This publication was developed as part of the University Transportation Centers Program which is funded 50% in oil overcharge funds from the Stripper Well Settlement as provided by the Texas State Energy Conservation Office and approved by the U.S. Department of Energy. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. The authors would like to acknowledge the assistance of the following students in data analysis and preparation of the final report: David Berry - data analysis and computer programming; Monty Poppe - data analysis and computer programming; and Jordon Richard - final report graphics. vi

7 EXECUTIVE SUMMARY Introduction When properly implemented, HOV facilities provide energy, air quality, and travel time savings and reliability benefits relative to many traditional methods of improving urban mobility. These particular benefits have historically been extremely difficult to quantify due to data limitations. A thorough assessment of these benefits using extensive existing data (as opposed to simulations based on hypothetical conditions or small samples of data) has not yet been conducted. Recent advances in automatic vehicle identification (AVI) enable tremendous quantities of travel time and speed information to be collected along instrumented highways. The Texas Department of Transportation (TxDOT) is developing a traffic monitoring system in Houston, Texas that uses AVI technology to collect real-time travel times and speeds. The traffic monitoring system will eventually include all major freeway facilities in the Houston area, including HOV lanes located in the median of most radial freeway corridors. The detailed travel times and speeds gathered by the Houston system provide an opportunity to analyze HOV facility benefits with greater levels of confidence than has been possible in the past. Study Objectives and Scope The primary objective of this study was to quantify the energy consumption, air quality, and travel time savings and reliability benefits of the Houston HOV lane system relative to traditional transportation system improvements (e.g., addition of general-purpose lanes to existing freeways). Other transportation improvements examined in this study include: 1) freeway geometrics that existed before HOV lane implementation (no-build scenario); 2) adding one general-purpose lane in each direction without an HOV lane; 3) adding two general-purpose lanes in each direction without an HOV lane; and, 4) implementing traffic system management (TSM) practices to increase freeway capacity by ten percent. These four scenarios and the existing case of a reversible HOV lane were evaluated under the existing level of person demand currently in the three freeway study corridors: Katy (I-10) Freeway, Northwest (US 290) Freeway, and North (I-45) Freeway. Overview of the Study Design Travel time data from Houston s traffic monitoring system were used for two basic purposes in this study: Determine current travel time savings and reliability of selected Houston HOV lanes throughout the calendar year; and, Calibrate speeds in FREQ10PL base models to ultimately simulate the energy, air quality, and travel time savings benefits of HOV lanes and other transportation improvements. vii

8 For the first purpose, travel time data for eight months in 1994 were summarized to determine the travel time savings and reliability benefits for three HOV lanes relative to the adjacent freeway general-purpose lanes. These benefits are based on actual travel times collected by the AVI traffic monitoring system in Houston. The second purpose for the travel time data--calibrating FREQ10PL computer simulation base models--ensured that the computer model outputs of energy consumption, air quality, and travel time savings were accurate and comparable. For the FREQ10PL model outputs, the estimated HOV corridor benefits are compared to the simulated effects of other transportation improvements. Findings The travel time savings and reliability analysis produced the following findings: Speeds in the freeway general-purpose lanes vary significantly between days of the week. Day-to-day standard deviations range from 4.0 to 11.9 mph for the peak hour, and 3.8 to 9.9 mph for the peak period. Speeds in the HOV lanes do not vary significantly between days of the week. Dayto-day standard deviations range from 0.4 to 4.1 mph. Speeds for similar weekdays (e.g., all Mondays) do not vary as much as between different weekdays (e.g., Monday through Friday). In particular, Fridays exhibit a tendency to have higher speeds than other days of the week. Figure 14 (in the body of this report) illustrates the typical day-to-day average peak hour speed variation for the Northwest Freeway and HOV lanes for the month of July The other two freeway corridors had similar day-to-day variations. The average number of probe vehicles on the freeway general-purpose lanes during a fifteen-minute period in the peak hour is approximately 15 to 20, or about one vehicle every minute. The probe vehicle coverage on the HOV lanes for similar conditions is approximately 5 to 10, or about one vehicle every 1-1/2 to 3 minutes. The probe vehicle coverage varies by corridor, most likely because of the proximity of tollway facilities utilizing AVI transponders at automated toll booths. North Freeway had the best probe vehicle coverage, presumably because the Hardy Toll Road is a parallel corridor (separated by three to four miles) in the northern suburbs, and the Sam Houston Tollway intersects North Freeway in northern Houston suburbs (Figure 1). Katy Freeway had the least amount of probe vehicle coverage among the three freeway corridors examined, averaging about one vehicle every three minutes during the peak hour. Katy Freeway can be considered the most congested corridor, as the HOV lane experiences average travel time savings between 9 and 15 minutes during the viii

9 morning peak hour, and between 12 and 17 minutes during the evening peak hour. The Northwest Freeway HOV lane experiences average travel time savings between 5 and 13 minutes during the morning peak hour, and between 5 and 8 minutes during the evening peak hour. The North Freeway HOV lane experiences average travel time savings between 2 and 10 minutes during the morning peak hour, and between 1 and 3 minutes during the evening peak hour. The FREQ10PL computer simulation analyses also quantified other benefits of HOV lanes. For each freeway corridor, the addition of a single HOV lane results in the greatest reduction in fuel consumption compared to the addition of general-purpose lanes, TSM applications, or doing nothing. For severely congested corridors like the Katy Freeway, the HOV lane alternative provided delay savings comparable to adding two freeway lanes in each direction. For moderately congested corridors like the North and Northwest Freeways, the HOV lane alternative compared favorably (in terms of delay reduction) to the addition of one freeway lane in each direction. It should also be noted that the addition of two general-purpose lanes is typically not a feasible option in most freeway corridors due to cost and right-of-way considerations. Conclusions and Recommendations Extensive travel time data were used in this study to quantify travel time savings and reliability benefits of three Houston HOV lanes relative to the freeway general-purpose lanes. The travel time data were also used to calibrate computer simulation models, which were, in turn, used to simulate the benefits of HOV lanes relative to other traditional transportation improvements. The major conclusions and recommendations are presented below. The travel time data analyses illustrated that the HOV lanes have substantial travel time savings over the freeway general-purpose lanes. The savings are in direct proportion to the level of congestion in the adjacent freeway lanes. The data available through Houston s traffic monitoring system provides extensive information about day-to-day speeds and variability in speeds (as reflected by standard deviations and confidence intervals). The data summaries show that speeds are higher and more reliable in the HOV lane than the adjacent freeway general-purpose lanes. The freeway speeds were found to vary dramatically between weekdays, while some patterns could be established between similar weekdays (e.g., Fridays). The speed variability information provided in this report should be useful in planning detailed data collection or other activities. The computer simulation analyses indicated that, for severely congested corridors like the Katy Freeway, the HOV lane alternative provided delay savings comparable to adding two freeway lanes in each direction. For moderately congested corridors like the North and Northwest Freeways, the HOV lane alternative compared favorably (in terms of delay reduction) to the addition of one ix

10 freeway lane per direction. However, the addition of two general-purpose lanes per direction is typically not a realistic option due to funding and right-of-way constraints. The HOV lane alternatives also reduced delay by a lower amount than the add-one-lane alternatives for all but the most congested corridor. The Katy Freeway HOV lane analysis is indicative of the point at which travel time savings are greater due to the emphasis on person movement, than an alternative that focuses on vehicle movement. Finally, it should be realized that while the HOV alternative may not currently offer the best overall solution on the less congested freeways, it will become much more attractive as time goes on and congestion levels increase on those roadways. It is recommended that future research should: Merge an adequate incident data base with the AVI travel time data base to better quantify the effects of incidents on freeway general-purpose lane and HOV lane operation; Quantify typical values for speed variability along a wider range of operating levels, from free-flow to severely congested; and, Recognize the usefulness of the vast amounts of operational data now available through many intelligent transportation system (ITS) technologies for planning applications. Information from probe vehicles, loop detectors, and other vehicle detection technologies can be used to better quantify traffic patterns, trends, and characteristics. x

11 TABLE OF CONTENTS Page ABSTRACT...v ACKNOWLEDGMENTS... vi EXECUTIVE SUMMARY... vii LIST OF FIGURES... xiii LIST OF TABLES... xii CHAPTER I. INTRODUCTION...1 Problem Statement...1 Opportunities Afforded by Automatic Data Collection Technology...1 Study Objectives and Scope...2 Organization of Report...2 CHAPTER II. BACKGROUND...3 Summary of HOV Evaluation Techniques...3 Washington State Department of Transportation...3 Texas Department of Transportation...4 Overview of the Houston Traffic Monitoring System...8 CHAPTER III. STUDY DESIGN...11 Overview of Study Design...11 Description of Data Source...13 Analysis of Travel Time Data...17 FREQ10PL Simulation Analysis...20 Development of Models in FREQ10PL...20 Temporal and Spatial Redistribution of Existing HOV Traffic...21 Outputs Available for Comparison using FREQ10PL...22 CHAPTER IV. FINDINGS...31 Travel Time Savings and Reliability Analysis...31 FREQ10PL Simulation Analysis...46 xi

12 TABLE OF CONTENTS (Continued) Page CHAPTER V. CONCLUSIONS AND RECOMMENDATIONS...51 Conclusions...51 Travel Time Savings and Reliability Analyses...51 FREQ10Pl Simulation Analyses...51 Recommendations...52 REFERENCES...55 APPENDIX A: Guide to the Appendices... A-1 APPENDIX B: Katy Freeway (I-10) and HOV Lane Travel Time Data...B-1 APPENDIX C: Northwest Freeway ( US 290) and HOV Lane Travel Time Data...C-1 APPENDIX D: North Freeway (I-45) and HOV Lane Travel Time Data... D-1 xii

13 LIST OF FIGURES Page Figure 1. Calculation of Travel Times Using AVI Technology...9 Figure 2. Houston AVI System Reader Sites...10 Figure 3. Overall Approach for Quantifying HOV Benefits...12 Figure 4. Example of Travel Time Data in ASCII-Text Format...14 Figure 5. Example of Travel Time Daily Summaries...18 Figure 6. Example of Fifteen-Minute Period Summaries by Freeway Section...19 Figure 7. Example of Fifteen-Minute Period Summaries by Freeway Corridor...19 Figure 8. Example of FREQ10PL Speed Contour Diagram...23 Figure 9. Example of FREQ10PL Queue Length Contour Diagram...24 Figure 10. Example of FREQ10PL V/C Ratio Contour Diagram...25 Figure 11. Example of FREQ10PL User-Supplied Speed-Flow Curve...27 Figure 12. Example of FREQ10PL Vehicle Occupancy Distribution...28 Figure 13. Example of FREQ10PL Freeway Summary Table Figure 14. Day-to-Day Speed Variation along the Northwest Freeway, July Figure % Speed Confidence Interval for the Katy Freeway Mainlanes and HOV Lane (Morning Peak Hour)...42 Figure % Speed Confidence Interval for the Northwest Freeway Mainlanes and HOV Lane (Morning Peak Hour)...43 Figure % Speed Confidence Interval for the North Freeway Mainlanes and HOV Lane (Morning Peak Hour)...44 xiii

14

15 LIST OF TABLES Page Table 1. Suggested Objectives and Measures of Effectiveness...6 Table 2. Suggested Data Collection Efforts for Corresponding Objectives and Measures of Effectiveness...7 Table 3. Summary of Phase One Freeway and HOV Lane Sections...15 Table 4. Vehicle Occupancy Rates Before and After Implementation of HOV Lanes...22 Table 5. Travel Time Savings and Reliability for Katy Freeway HOV Lane, Peak Hour...34 Table 6. Travel Time Savings and Reliability for Northwest Freeway HOV Lane, Peak Hour...35 Table 7. Travel Time Savings and Reliability for North Freeway HOV Lane, Peak Hour...36 Table 8. Travel Time Savings and Reliability for Katy Freeway HOV Lane, Peak Period...37 Table 9. Travel Time Savings and Reliability for Northwest Freeway HOV Lane, Peak Period...38 Table 10. Travel Time Savings and Reliability for North Freeway HOV Lane, Peak Period.. 39 Table 11. Summary of Travel Time Savings and Reliability for Selected Houston HOV Lanes...40 Table 12. Vehicle and Person Volumes for Katy, Northwest, and North Freeways and HOV Lanes Table 13. Daily Peak Hour Delay Savings for Katy, Northwest, and North HOV Lanes...46 Table 14. Delay, Fuel Consumption and Emissions for Freeway Corridor Alternatives...48 Table 15. Percent Change in Daily Delay, Fuel Consumption and Emissions for Freeway Corridor Alternatives...49 Table 16. Preferred Corridor Alternatives by Congestion Level...50 Table B-1. Katy Freeway Peak Hour, April B-1 Table B-2. Katy Freeway Peak Period, April B-2 Table B-3. Katy Freeway Peak Hour, May B-3 Table B-4. Katy Freeway Peak Period, May B-4 Table B-5. Katy Freeway Peak Hour, June B-5 Table B-6. Katy Freeway Peak Period, June B-6 Table B-7. Katy Freeway Peak Hour, July B-7 Table B-8. Katy Freeway Peak Period, July B-8 Table B-9. Katy Freeway Peak Hour, August B-9 Table B-10. Katy Freeway Peak Period, August B-10 Table B-11. Katy Freeway Peak Hour, September B-11 Table B-12. Katy Freeway Peak Period, September B-12 Table B-13. Katy Freeway Peak Hour, October B-13 Table B-14. Katy Freeway Peak Period, October B-14 Table B-15. Katy Freeway Peak Hour, November B-15 Table B-16. Katy Freeway Peak Period, November B-16 xv

16 LIST OF TABLES (Continued) Page Table C-1. Northwest Freeway Peak Hour, April C-1 Table C-2. Northwest Freeway Peak Period, April C-2 Table C-3. Northwest Freeway Peak Hour, May C-3 Table C-4. Northwest Freeway Peak Period, May C-4 Table C-5. Northwest Freeway Peak Hour, June C-5 Table C-6. Northwest Freeway Peak Period, June C-6 Table C-7. Northwest Freeway Peak Hour, July C-7 Table C-8. Northwest Freeway Peak Period, July C-8 Table C-9. Northwest Freeway Peak Hour, August C-9 Table C-10. Northwest Freeway Peak Period, August C-10 Table C-11. Northwest Freeway Peak Hour, September C-11 Table C-12. Northwest Freeway Peak Period, September C-12 Table C-13. Northwest Freeway Peak Hour, October C-13 Table C-14. Northwest Freeway Peak Period, October C-14 Table C-15. Northwest Freeway Peak Hour, November C-15 Table C-16. Northwest Freeway Peak Period, November C-16 Table D-1. North Freeway Peak Hour, April D-1 Table D-2. North Freeway Peak Period, April D-2 Table D-3. North Freeway Peak Hour, May D-3 Table D-4. North Freeway Peak Period, May D-4 Table D-5. North Freeway Peak Hour, June D-5 Table D-6. North Freeway Peak Period, June D-6 Table D-7. North Freeway Peak Hour, July D-7 Table D-8. North Freeway Peak Period, July D-8 Table D-9. North Freeway Peak Hour, August D-9 Table D-10. North Freeway Peak Period, August D-10 Table D-11. North Freeway Peak Hour, September D-11 Table D-12. North Freeway Peak Period, September D-12 Table D-13. North Freeway Peak Hour, October D-13 Table D-14. North Freeway Peak Period, October D-14 Table D-15. North Freeway Peak Hour, November D-15 Table D-16. North Freeway Peak Period, November D-16 xvi

17 CHAPTER I. INTRODUCTION The priority treatment of high-occupancy vehicles (HOVs) has been widely reported to be one of the most effective measures to manage traffic congestion and its related effects, such as vehicle delay, excess energy consumption, and mobile source emissions. The reported benefits of HOV priority treatments commonly include: Shift to more space- and energy-efficient transportation modes (e.g., carpools, vanpools, buses); Increase in the person-carrying capacity of transportation corridors; Reduction in total trip times, energy consumption and mobile source emissions; and, Improvement in the efficiency and economy of public transit operations. These benefits are often predicted (via computer models) for planned HOV treatments and serve as justification for construction of the project. In some cases, the predicted benefits are based on limited data. Recent legislation such as the Clean Air Act Amendments of 1990 and the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991 has resulted in an increased emphasis of urban transportation systems making more efficient use of limited resources. Facilities such as HOV lanes have been receiving growing attention as a means of addressing increasing travel demand on urban area transportation networks. This added attention has heightened the need for comprehensive asessments of HOV facilities as a part of alternatives analyses in urban travel corridors. In particular, the focus of the recent legislation is related to the impacts of alternative transportation improvements on energy consumption and air quality. This area of emphasis has created a need for HOV facilities (as well as other transportation improvements) to be analyzed on such a basis. Computer simulations are commonly used to estimate the energy and air quality benefits of the HOV facility. These computer simulations are often based upon hypothetical conditions or small samples of data. Problem Statement When properly implemented, HOV facilities provide energy, air quality, and travel time savings and reliability benefits relative to many traditional methods of improving urban mobility. These particular benefits have historically been extremely difficult to quantify due to data limitations. A thorough assessment of these benefits using extensive existing data (as opposed to simulations based on hypothetical conditions or small samples of data) has not yet been conducted. Opportunities Afforded by Automatic Data Collection Technology Recent advances in automatic vehicle identification (AVI) enable tremendous quantities of travel time and speed information to be collected along instrumented highways. The Texas Department of Transportation (TxDOT) is developing a traffic monitoring system in Houston, Texas that uses AVI technology to collect real-time travel times and speeds. The traffic monitoring system 1

18 will eventually include all major freeway facilities in the Houston area, including HOV lanes located in the median of most radial freeway corridors. The detailed travel times and speeds gathered by the Houston system provide an opportunity to analyze HOV facility benefits with greater levels of confidence than has been possible in the past. Study Objectives and Scope The primary objective of this study was to quantify the energy consumption, air quality, and travel time savings and reliability benefits of the Houston HOV lane system relative to traditional transportation system improvements (e.g., addition of general-purpose lanes to existing freeways). Other transportation scenarios examined in this study include: 1) freeway geometrics that existed before HOV lane implementation (no-build scenario); 2) adding one general-purpose lane in each direction without an HOV lane; 3) adding two general-purpose lanes in each direction without an HOV lane; and, 4) implementing traffic system management (TSM) practices to increase freeway capacity by 10 percent. These four scenarios and the existing case of a reversible HOV lane were evaluated under the existing level of person demand currently in the three freeway study corridors: Katy (I-10) Freeway, Northwest (US 290) Freeway, and North (I-45) Freeway. Organization of Report This report is organized into five major chapters: Chapter One, Introduction, provides an introduction to the research topic and presents the research objectives and scope. Chapter Two, Background, provides general information about other HOV lane evaluations in Texas and across the United States. This chapter also presents an overview of the Houston freeway traffic monitoring system that provided the travel time and speed data used in this research. Chapter Three, Study Design, contains a summary of the procedures used to quantify the energy, air quality, and travel time savings and reliability benefits of selected Houston HOV lanes. This chapter presents the source of data, the techniques used to reduce and analyze the data, and the simulation model used to examine HOV lane benefits relative to other transportation improvements. Chapter Four, Findings, presents the major findings for the research study, concentrating on the energy consumption, air quality, and travel time savings and reliability benefits for selected HOV lanes in Houston, Texas. This chapter includes the detailed results of the data analyses and computer simulations. Chapter Five, Conclusions and Recommendations, presents the conclusions of the study based upon the research findings presented in Chapter Four. The conclusions of the research relate to the benefits of properly implemented HOV lanes. This final chapter also provides recommendations of this study regarding HOV lane evaluations and future research. 2

19 CHAPTER II. BACKGROUND This chapter provides general information about HOV lane evaluations in Texas and across the United States. Common procedures and methodologies for evaluating HOV lanes are presented, and the use of various performance measures is discussed. This chapter also presents an overview of the Houston traffic monitoring system that provided the travel time and speed data used in this research. The components of the traffic monitoring system are described, and the available information is presented. Summary of HOV Evaluation Techniques Many techniques for evaluating the effectivenes of HOV facilities have evolved since the first freeway HOV lane was opened on the Shirley Highway in Northern Virginia/Washington, D.C. in Similarly, many measures of effectiveness have been devised to determine whether an HOV facility is fulfilling intended goals and objectives. The following sections present methodologies and measures of effectiveness that have been developed by two state departments of transportation (DOTs), Washington DOT and TxDOT, to evaluate HOV facilities. Washington State Department of Transportation The Washington State DOT (WsDOT) is developing a regional HOV system in Seattle that includes concurrent flow facilities on several of the area s major freeways and arterial streets. Approximately 33 miles (53 kilometers) of HOV lanes are included in the current system, with additional expansion of the system continuing over the next several years. In the Washington State Freeway HOV System Policy, WsDOT established several objectives for the HOV system (1): Improve the capability of congested freeway corridors to move more people by increasing the number of persons per vehicle; Provide travel time savings and a more reliable trip time to high occupancy vehicles that use the facilities; and, Provide safe travel options for high occupancy vehicles without unduly affecting the safety of the freeway general-purpose mainlanes. WsDOT also uses the following measures of effectiveness to determine whether these HOV system objectives are being satisfied: person throughput; average vehicle occupancy; comparative and absolute general-purpose and HOV lane travel times; travel time reliability; and, accident rates. No specific values or thresholds are given for these measures of effectiveness except for travel speed and travel time, for which the WsDOT HOV system policy (1) states that HOV lane vehicles should 3

20 maintain or exceed an average speed of 45 mph or greater at least 90% of the times they use that lane during the peak hour (measured for a consecutive six-month period). Several HOV lane evaluations conducted for specific corridors in Washington State have used these HOV system objectives and measures of effectiveness. An evaluation of Seattle s South I-5 Interim HOV lane used these measures to determine that the interim HOV lane had been relatively ineffective, and that the general-purpose lanes required geometric improvements to reduce congestion and safety problems created by the HOV lane (2). The evaluation collected peak period traffic data for the general-purpose and HOV lanes, including lane volumes, accidents before and after implementation of the HOV lane, travel times, and average vehicle occupancies. The South I-5 evaluation also included a traveler opinion survey that was distributed to commuters in the HOV lane corridor. An evaluation of the I-5 North HOV lane in Seattle used similar measures of effectiveness to determine that lowering the occupancy requirement from three or more persons per vehicle (3 + ) to two or more persons per vehicle (2 + ) negatively impacted the HOV lane (3). The evaluation indicated that HOV lane travel times increased and HOV travel time reliability decreased, presumably from congestion caused by allowing additional two-occupant vehicles on the HOV lane. The evaluation noted the importance of consistent, accurate data for the measures of effectiveness and HOV lane evaluation. The study noted that traffic and public survey data collected before implementation of an HOV facility was essential to characterize the HOV facility impacts. Texas Department of Transportation TxDOT and the Metropolitan Transit Authority (METRO) of Harris County has committed to developing a 103-mile (166-kilometer) network of HOV lanes in Houston. Dallas Area Rapid Transit (DART) has joined with TxDOT in planning and programming 37 miles (59 kilometers) of HOV lanes in Dallas (4). The existing HOV lane mileage in Houston totals 64 miles (102 kilometers) in five separate radial freeway corridors, whereas the Dallas HOV system currently consists of a single, 5-mile (8-kilometer) contraflow HOV lane. The Texas Transportation Institute (TTI) has been monitoring and evaluating the HOV facilities in Houston and in Dallas both before and after implementation for all HOV facilities. The most recent annual evaluation report (4) states that:... the primary goal of the HOV lanes in Texas is to cost effectively increase the person-movement capacity of the freeways. Achieving this should: 1) enhance bus transit operations; 2) improve air quality; and, 3) reduce fuel consumption. Implementation of the HOV lanes should not unduly impact the operation of the freeway general-purpose lanes. That implementation should have general public support. 4

21 The following specific objectives were also listed: Increase the effective person-movement capacity of the freeway; Improve the efficiency of bus transit operations; HOV lane implementation should not unduly impact freeway mainlane operation, and its implementation should increase overall roadway efficiency; The HOV lane project should be cost effective; Development of the HOV facility system should have public support; and, High-occupancy vehicle facilities should have favorable impacts on air quality and energy consumption. The annual study conducted by TTI evaluates the HOV lanes in Texas based upon these general objectives and a related matrix of measures of effectiveness. Data collection for the study occurs on a quarterly basis during the year for the HOV lanes and on a semi-annual basis for the general-purpose freeway lanes. As mentioned previously, traffic data along existing and potential HOV corridors has been collected prior to HOV lane implementation for most facilities. TTI has also developed general guidance and generic procedures for evaluating the effectiveness of HOV facilities (5). The suggested procedures include objectives and the respective measures of effectiveness, general threshold ranges, and data needs for each specific objective. Tables 1 and 2 summarize the suggested objectives, measures of effectiveness, and related data collection efforts. 5

22 Table 1. Suggested Objectives and Measures of Effectiveness (Adapted from Ref. 5) Objective The HOV facility should improve the capability of a congested freeway corridor to move more people by increasing the number of persons per vehicle. The HOV facility should increase the operating efficiency of bus service in the freeway corridor. The HOV facility should provide travel time savings and a more reliable trip time to HOVs utilizing the HOV facility. Measures of Effectiveness Actual and percent increase in the person movement efficiency. Actual and percent increase in the average vehicle occupancy rate. Actual and percent increase in carpools and vanpools. Actual and percent increase in bus riders. Improvement in vehicle productivity (operating cost per vehicle-mile, operating cost per passenger, operating cost per passenger-mile) Improved bus schedule adherence (on-time performance). Improved bus safety (lower accident rate). The peak period, peak direction travel time in the HOV lane(s) should be less than the travel time in the adjacent freeway lanes. Increase in travel time reliability for vehicles using HOV lane(s). 6 The HOV facility should have favorable impacts on air quality and energy consumption. Reduction in emissions. Reduction in total fuel consumption. Reduction in the growth of vehicle-miles of travel and vehicle-hours of travel The HOV facility should increase the per lane efficiency of the total freeway facility. The HOV facility should not unduly impact the operation of the freeway mainlanes. The HOV facility should be safe and should not unduly impact the safety of the freeway general-purpose mainlanes. The HOV facility should have public support. The HOV facility should be a cost-effective transportation improvement. Improvement in the peak-hour per lane efficiency of the total facility. The level of service in the freeway mainlanes should not decline. Number and severity of accidents for HOV and freeway lanes. Accident rate per million vehicle-miles of travel. Accident rate per million passenger-miles of travel. Support for the facility among users, non-users, general public, and policy makers. Violation rates (percent of vehicles not meeting the occupancy requirement). Benefit-cost ratio.

23 Table 2. Suggested Data Collection Efforts for Corresponding Objectives and Measures of Effectiveness (Adapted from Ref. 5) Data Collection Efforts Objective Vehicle and Occupancy Counts Travel Time Runs Surveys 1 Measures of Effectiveness (MOEs) 3 Freeway HOV Lane Freeway 2 HOV Lane Freeway HOV Lane Other Increase vehicle occupancy Bus operating efficiency Travel time savings * * ** ** ** 4 Actual and percent increase in peak hour, peak direction person volume, increase in average vehicle occupancy, modal shift * * * 5 Improved vehicle productivity, bus schedule adherence, and bus safety * * ** ** ** 6 Amount of travel time savings by HOV users, reliability of trip time for HOV users Energy and air * * * * ** ** ** 7 Reduction in emissions and energy consumption Per lane efficiency * * * * Increase in peak hour lane efficiency of total freeway facility 7 Freeway operations * * ** Maintain or improve level of service on freeway mainlanes Safety ** ** * 8 Number and severity of accidents, accident rate per million vehiclemiles of travel and per million passenger-miles of travel Public support ** * * ** 9 Percent of users, non-users, and general public who approve of HOV facility, violation rates Cost-effective ** * * * Benefit-cost ratio * Indicates the top priority data collection efforts needed to evaluate the objectives. ** Indicates data collection efforts which should ideally be conducted, but are not absolutely necessary to evaluate the objectives. 1 Involves the periodic use of surveys of HOV users (bus riders, carpoolers, and vanpoolers), non-hov users in the general-purpose traffic lanes, and in some cases, the general public. 2 It is strongly suggested that these data be collected for both the freeway lanes adjacent to the HOV facility and the control freeway. 3 Some, but not necessarily all, of the suggested MOEs associated with gauging the attainment of the objectives are shown. 4 Vehicle/occupancy counts on alternate arterial routes to identify corridor throughput changes, counts at park-and-ride lots, and vehicle and occupancy counts on a control freeway. 5 Before and after bus service levels, vehicle productivity, schedule adherence, number and severity of bus accidents, vehicle operating costs, and changes in other costs. 6 Monitoring bus on time performance and schedule adherence before and after implementation of the HOV lane(s). 7 Monitoring air quality levels along the corridor and use of simulation models to estimate impact. 8 Monitoring freeway accident rates and types before and after implementation of the HOV lane(s), as well as obtaining accident rates on the HOV facility. 9 Identifying violation rates for the HOV lane (i.e., those vehicles not meeting the minimum occupancy requirement). Monitor complaints, media, and policy actions.

24 Overview of the Houston Traffic Monitoring System The Texas Department of Transportation (TxDOT) is developing a traffic monitoring system for the Houston area that includes 227 miles (365 kilometers) of freeway and 70 miles (113 kilometers) of HOV lanes (6). The traffic monitoring system utilizes automatic vehicle identification (AVI) technology to collect travel times for 1- to 5-mile (1.6- to 8-kilometer) sections of the freeways and HOV lanes. The AVI equipment for the monitoring system includes the following: vehicle transponders (tags) attached to the interior surface of the windshield; reader antennas mounted on overpass bridge structures or sign bridges; roadside reader cabinets that relay transponder identification numbers to the traffic control center via phone lines; and, computer hardware and software at the traffic control center. Figure 1 illustrates the equipment and procedures used to calculate travel times with the AVI system. The control center computer system matches transponder idenfication numbers at consecutive reader sites and calculates a travel time for that particular section of freeway or HOV lane. A screening algorithm is used to eliminate erroneous travel times that occur when a transponder-equipped vehicle leaves and re-enters the freeway between reader sites. The system calculates and displays the travel time information in real-time, and distributes the information to the media and transportation agencies. A color-coded speed map for the Houston area is also provided on the World Wide Web at Figure 2 illustrates the freeways included in the traffic monitoring system and also shows the checkpoint locations at which reader antennas are located. A portion of the vehicle transponders used to collect travel times have been distributed free of charge by TxDOT to commuters wishing to participate in the traffic monitoring system and act as probe vehicles. A significant percentage of the transponders read by the traffic monitoring system have been distributed by the Harris County Toll Road Authority (HCTRA). Transponders distributed by HCTRA are used for electronic toll collection on two area toll roads: the Sam Houston Tollway and the Hardy Toll Road (Figure 2). The Metropolitan Transit Authority (METRO) equipped their buses on the HOV lane with transponders, and the city of Houston Aviation Department also uses compatible transponders to monitor commercial vehicle traffic. The number of vehicle transponders in the Houston area has been estimated at approximately 50,000, with other compatible transponders being read from Oklahoma toll roads or Louisiana toll bridges. TxDOT installed the monitoring system in three separate phases, with the last phase to be completed by the end of Houston s traffic monitoring system provides travel time data every day for the area s major freeways and HOV lanes. The following chapter presents the methodology used to evaluate three selected HOV lanes and freeway corridors with this travel time data. 8

25

26

27 CHAPTER III. STUDY DESIGN This chapter presents the methodology and analysis techniques that were used to quantify the energy consumption, air quality, and travel time savings and reliability benefits of HOV lanes. The overall study approach is first described, with the following sections presenting details on the travel time data used to quantify existing travel time savings and reliability benefits of several Houston HOV facilities. Subsequent sections discuss the use of the same travel time data for calibrating computer simulation models. This chapter concludes with a discussion of the FREQ10PL computer simluation program used to evaluate the energy, air quality, and travel time savings benefits of HOV lanes relative to alternative transportation improvements. Overview of Study Design Figure 3 presents an illustration of the approach used in this study to quantify the energy, air quality, and travel time savings and reliability benefits of HOV lanes. One of the primary motivations for this study was the large amount of travel time data available through Houston s traffic monitoring system. This travel time data was used for two basic purposes in this study (see Figure 3): Determine current travel time savings and reliability of selected Houston HOV lanes throughout the calendar year; and, Calibrate speeds in FREQ10PL base models to ultimately simulate the energy, air quality, and travel time savings benefits of HOV lanes relative to other transportation improvements. For the first purpose, travel time data for eight months in 1994 were summarized to determine the travel time savings and reliability benefits for three HOV lanes relative to the adjacent freeway general-purpose lanes. These benefits are based on actual travel times collected by the AVI traffic monitoring system in Houston. The second purpose for the travel time data--calibrating FREQ10PL computer simulation base models--ensured that the computer model outputs of energy consumption, air quality, and travel time savings were accurate and comparable. For the FREQ10PL model outputs, the estimated HOV corridor benefits are compared to the simulated effects of other transportation improvements. The next section discusses the travel time data that were used for these two study purposes. 11

28

29 Description of Data Source Travel time data from the Houston traffic monitoring system were used to quantify the travel time savings and reliability benefits of existing HOV lanes relative to the freeway mainlanes. The travel time data were also used to calibrate the computer models used to simulate alternative transportation improvements. At the time of this study, only Phase One of the Houston traffic monitoring system was fully operational. Phase One of the system includes three corridors to the north and west of downtown Houston: Katy Freeway (I-10), Northwest Freeway (US 290), and North Freeway (I-45). Within each of these corridors, transponder reader sites are located at 1.8- to 5.1-mile (2.9- to 8.2-kilometer) spacings on both the freeway mainlanes and the HOV lanes located in the median of the freeway. Based upon the AVI reader sites (see Figure 2), there were a total of sixty directional freeway and HOV lane segments analyzed for this study. These segments represent a wide range of operating conditions, from free-flow on the HOV lanes to severely congested on some freeway segments. A summary of these freeway and HOV sections are presented in Table 3. The first month of complete operational data became available in April The data used in the study includes the following number of weekdays for eight months in 1994: April: 20 weekdays (excludes April 1, Good Friday); May: 20 weekdays (excludes May 10, unavailable data, and May 30, Memorial Day); June: 22 weekdays; July: 20 weekdays (excludes July 4, Independence Day); August: 21 weekdays (excludes August 22 and 23, unavailable data); September: 20 weekdays (excludes Sept. 5, Labor Day and Sept. 29, unavailable data); October: 21 weekdays; and, November: 15 weekdays (excludes Nov , Thanksgiving Day and Nov. 11, 14, and 28, unavailable data). The data were stored in ASCII-text format, and included the following variables: Start Checkpoint: numeric code for first consecutive reader site; End Checkpoint: numeric code for second consecutive reader site; Start Time: military clock time at the Start Checkpoint ; Read Date: month, day, and year of observation; Travel Time: travel time from Start Checkpoint to End Checkpoint, in hours:minutes:seconds format; and, Segment Length: distance in miles from Start Checkpoint to End Checkpoint. A typical ASCII-text file contained between 25,000 and 35,000 travel time observations for each weekday, and the file sizes for each weekday ranged from 1.5 to 2.0 megabytes. Figure 4 shows an example of the travel time data as it was obtained from the traffic monitoring system computers. 13

30 Start Checkpoint End Checkpoint Start Time Read Date Travel Time Segment Length ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ ÃÃÃÃÃÃÃÃÃÃÃ Figure 4. Example of Travel Time Data in ASCII-Text Format 14

31 Table 3. Summary of Phase One Freeway and HOV Lane Sections Consecutive Section No. Section Limits Section Length (miles) Katy Freeway (I-10) Eastbound (Inbound) Barker Cypress to Eldridge Eldridge to Sam Houston Tollway Sam Houston Tollway to Blalock Blalock to I-610 I-610 to T.C. Jester Katy Freeway (I-10) Westbound (Outbound) T.C. Jester to I-610 I-610 to Blalock Blalock to Sam Houston Tollway Sam Houston Tollway to Eldridge Eldridge to Barker Cypress Katy Freeway (I-10) HOV Eastbound (Inbound)--Morning Only SH 6 to Sam Houston Tollway Sam Houston Tollway to Bunker Hill Bunker Hill to Silber Silber to HOV Endpoint Katy Freeway (I-10) HOV Westbound (Outbound)--Evening Only HOV Endpoint to Silber Silber to Bunker Hill Bunker Hill to Sam Houston Tollway Sam Houston Tollway to SH Northwest Freeway (US 290) Eastbound (Inbound) Barker Cypress to FM 1960 FM 1960 to Sam Houston Tollway Sam Houston Tollway to Fairbanks Fairbanks to Pinemont Pinemont to 34th Street 34th Street to Dacoma Northwest Freeway (US 290) Westbound (Outbound) Dacoma to 34th Street 34th Street to Pinemont Pinemont to Fairbanks Fairbanks to Sam Houston Tollway Sam Houston Tollway to FM 1960 FM 1960 to Barker Cypress