Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

High occupancy vehicle monitoring and evaluation framework Bracewell, Dale J. 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-0195.pdf [ 8.52MB ]
Metadata
JSON: 831-1.0063791.json
JSON-LD: 831-1.0063791-ld.json
RDF/XML (Pretty): 831-1.0063791-rdf.xml
RDF/JSON: 831-1.0063791-rdf.json
Turtle: 831-1.0063791-turtle.txt
N-Triples: 831-1.0063791-rdf-ntriples.txt
Original Record: 831-1.0063791-source.json
Full Text
831-1.0063791-fulltext.txt
Citation
831-1.0063791.ris

Full Text

HIGH OCCUPANCY VEHICLE MONITORING AND EVALUATION FRAMEWORK by  DALE J . BRACEWELL B. Eng., McGill University, 1995  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CIVIL ENGINEERING  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 1998 © Dale J. Bracewell, 1998  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of Civil Engineering The University of British Columbia Vancouver, Canada  Date  ABSTRACT This thesis describes a framework for monitoring and evaluating the effectiveness of a high occupancy vehicle (HOV) facility based on a set of quantifiable objectives. The framework is developed specifically to evaluate the effectiveness of the arterial and freeway HOV facilities within the Greater Vancouver area of British Columbia. However, it should be applicable to other jurisdictions as well.  As a contribution to existing HOV evaluation procedures, the HOV evaluation objectives are categorized into three groups by their relationship to the goals of an HOV program: primary, supporting, and operational. Primary objectives directly relate to the HOV goals of improving the person throughput and encouraging higher occupancy. Characteristics of an HOV facility that contribute to the attainment of the HOV goals are evaluated as supporting evaluation objectives. Attributes that protect the performance of the HOV facility (and therefore its attractiveness to users) are evaluated as operational objectives. Quantifiable measures of effectiveness (MOEs) are selected to determine if an objective is met. The data required to evaluate each of the MOEs is summarized. In addition, a data collection methodology and comprehensive procedures for analyzing and presenting the MOEs are described. To enhance current practices of evaluating HOV facilities, the framework includes a procedure to calculate the statistical reliability and relative uncertainty of the data collected.  Finally, a multi-criteria evaluation  procedure to evaluate the cost-effectiveness of an HOV facility is presented. Only the quantifiable benefits and costs of an HOV facility are considered in performing a benefitcost analysis.  Page ii  The framework is applied to evaluate the effectiveness of the HOV facility on the Barnet/Hastings corridor. A multi-criteria evaluation is performed and the results are presented. The objectives of the framework that are considered to be effective include: 'Person Throughput', 'Travel Time Savings', and 'Impact on GP Lanes'. The travel speed standard deviations of the HOV lane are less than the GP lane. However, it was determined that the HOV lane is not providing a statistically significant more reliable trip at the 95% confidence level. The Ministry of Transportation and Highways desired compliance rate of 85% is not presently being attained on the Barnet/Hastings corridor. No public opinion survey data relating to the operation of the HOV facility is available for the application and the safety results are not conclusive. Finally, a preliminary analysis indicated that the HOV facility on the Barnet/Hastings corridor is cost-effective.  Page iii  TABLE OF CONTENTS  ABSTRACT  ii  T A B L E OF CONTENTS  iv  LIST OF TABLES  xi  LIST OF FIGURES  xiii  ACKNOWLEDGEMENTS  xiv  CHAPTER 1. INTRODUCTION  1  1.1 BACKGROUND  1  1.2 PURPOSE OF THESIS  2  1.3 SCOPE OF THESIS  3  1.4 ORGANIZATION OF THESIS  5  1.5 SUMMARY  6  CHAPTER 2. LITERATURE REVIEW  7  2.1 INTRODUCTION  7  2.7.7 The Evaluation of HOV Facilities  7  2.7.2 Evaluation Objectives  8  2.1.3 Thresholds and Measures of Effectiveness  9  2.2 OBJECTIVES  11  2.2.7 People Moving Efficiency  77  2.2.2 Travel Time Savings  14  2.2.3 Safety  75  2.2.4 Compliance  ;  77  Page iv  2.2.5 Impact of HOV an GP Lanes  20  2.2.6 Impact of HOV on Parallel Routes  20  2.2.7 Overall Corridor Efficiency  22  2.2.8 Air Quality/Environmental  23  2.2.9 Public Support  24  2.2.10 Transit Efficiency  25  2.2.11 Travel Time Reliability  26  2.2.12 Cost-Effectiveness  27  2.3 DATA COLLECTION  28  2.3.1 Data Collection Techniques  28  2.3.2 Travel Time Runs  29  2.3.3 Vehicle Classification and Occupancy Counts  30  2.3.4 Public Opinion Surveys  31  CHAPTER 3. H.Q.V. MONITORING AND EVALUATION FRAMEWORK  32  3.1 INTRODUCTION  32  3.2 THE H.O.V. EVALUATION FRAMEWORK  32  3.2.1 Objectives  33  3.2.2 Measures of Effectiveness  36  3.2.3 Data Requirements  37  3.2.4 Data Collection Methodology  38  3.2.5 Data Analysis  38  3.2.6 Arterial vs. Freeway Analysis  40  3.2.7 Parallel Routes Analysis  40  3.2.8 Statistical Reliability  41  Page v  3.2.9 Statistical Significance  41  3.2.10 Discussion of Results  41  3.3 THE H.O.V. MONITORING FRAMEWORK  CHAPTER 4. OBJECTIVES AND EVALUATION PROCEDURES  42  44  4.1 INTRODUCTION  44  4.2 PERSON THROUGHPUT  44  4.2.1 Measures of Effectiveness  45  4.2.2 Data Requirements  46  4.2.3 Data Collection Methodology  47  4.2.4 Data Analysis  47  4.2.5 Parallel Routes Analysis  49  4.2.6 Statistical Reliability  50  4.2.7 Statistical Significance  50  4.3 TRAVEL TIME SAVINGS  50  4.3.1 Measures of Effectiveness  50  4.3.2 Data Requirements  57  4.3.3 Data Collection Methodology  57  4.3.4 Data Analysis  52  4.3.5 Arterial vs. Freeway Analysis  53  4.3.6 Statistical Reliability  53  4.3.7 Statistical Significance  53  4.4 TRAVEL TIME RELIABILITY  53  4.4.1 Measures Of Effectiveness  54  Page vi  4.4.2 Data Requirements  54  4.4.3 Data Collection Methodology  54  4.4.4 Data Analysis  .  55  4.4.5 Arterial vs. Freeway Analysis  56  4.4.6 Statistical Reliability  56  4.4.7 Statistical Significance  56  4.5 PUBLIC SUPPORT  57  4.5.1 Measures of Effectiveness  57  4.5.2 Data Requirements  58  4.5.3 Data Collection Methodology  58  4.5.4 Data Analysis  58  4.5.5 Statistical Reliability  59  4.6 COMPLIANCE  59  4.6.1 Measures of Effectiveness  59  4.6.2 Data Requirements  62  4.6.3 Data Collection Methodology  62  4.6.4 Data Analysis  63  4.6.5 Arterial vs. Freeway Analysis  64  4.7 SAFETY 4.7.1 Measures of Effectiveness  64 ,  64  4.7.2 Data Requirements  65  4.7.3 Data Collection Methodology  65  4.7.4 Data Analysis  66  4.7.5 Statistical Significance  67  4.8 IMPACT ON GP LANES  68  Page vii  4.8.1 Measures of Effectiveness  68  4.8.2 Data Requirements and Data Collection Methodology  69  4.8.3 Data Analysis  69  4.8.4 Arterial vs. Freeway Analysis  70  CHAPTER 5. MULTI-CRITERIA EVALUATION 5.1 INTRODUCTION 5.1.1 The Multi-Criteria Evaluation Procedure  71 71 71  5.2 BENEFITS AND COSTS  73  5.2.1 Travel Time Savings  73  5.2.2 Vehicle Operating Cost Savings and Air Quality Savings  74  5.2.3 Safety Savings 5.2.4 Costs  5.3 COST- EFFECTIVENESS 5.3.1 Sensitivity Analysis  CHAPTER 6. APPLICATION OF THE FRAMEWORK 6.1 INTRODUCTION  75 77  78 79  81 81  6.1.1 Data Collection Summary  81  6.1.2 Description of the HOV Facility  81  6.1.3 Data Analysis  84  6.2 PERSON THROUGHPUT  85  6.2.1 Per-Lane Efficiency  86  6.2.2 Average Vehicle Occupancy  87  Page viii  6.2.3 HOV Market Share 6.2.4 Person Throughput Summary  6.3 TRAVEL TIME SAVINGS 6.3.1 Travel Time Difference 6.3.2 Travel Time Savings Summary  6.4 TRAVEL TIME RELIABILITY 6.4.1 Travel Speed Standard Deviation 6.4.2 Travel Time Reliability Summary  6.5 COMPLIANCE  91 93  93 94 97  98 98 102  102  6.5.7 Compliance Rates  702  6.5.2 Violation Data  707  6.5.3 Enforcement Data  108  6.5.4 Compliance Summary  770  6.6 SAFETY 6.6.7 Accident Rates 6.6.2 Safety Summary  Ill 113 775  6.7 IMPACT ON GP LANES  115  6.7.7 GP Travel Speeds  77S  6.7.2 GP Vehicular Volumes 6.7.3 Impact on GP Lanes Summary  6.8 MULTI-CRITERIA EVALUATION  77<S 720  120  6.S.7 Project Parameters  727  6.8.2 Travel Time Savings  727  6.8.3 Vehicle Operating Cost Savings  722  6.8.4 Air Quality Savings  723  Page ix  6.8.5 Construction Costs  123  6.8.6 Maintenance Costs  124  6.8.7 Summary  124  CHAPTER 7. CONCLUSION  125  BIBLIOGRAPHY  128  APPENDIX A - STATISTICAL ANALYSIS  130  APPENDIX B - EFFICIENCY DATA  138  APPENDIX C - AVO DATA  144  APPENDIX D - HOV M A R K E T SHARE DATA  156  APPENDIX E - T R A V E L TIME SAVINGS DATA  164  APPENDIX F - T R A V E L TIME RELIABILITY DATA  173  APPENDIX G - COMPLIANCE DATA  175  APPENDIX H - SAFETY DATA  177  APPENDIX I - IMPACT ON GP LANES DATA  182  APPENDIX J - MULTI-CRITERIA EVALUATION  184  Page x  LIST OF TABLES Table 2-1: HOV Evaluation Objectives Identified in the Lilterature Review  9  Table 2-2: Summary of Measures of Effectiveness (MOEs)  12  Table 2-3: 'People Moving Efficiency' Objective MOEs  14  Table 2-4: 'Travel Time Savings' Objective MOEs  15  Table 2-5: 'Safety' Objective MOEs  17  Table 2-6: 'Compliance' Objective MOEs  20  Table 2-7: 'Impact of HOV on GP Lanes' Objective MOEs  20  Table 2-8: 'Impact of HOV on Parallel Routes' Objective MOEs  22  Table 2-9: 'Overall Corridor Efficiency' Objective MOEs  22  Table 2-10: 'Air Quality/Environmental' Objective MOEs  23  Table 2-11: 'Public Support' Objective MOEs  25  Table 2-12: 'Transit Efficiency' Objective MOEs  26  Table 2-13: 'Travel Time Reliability' Objective MOEs  27  Table 2-14: 'Cost-Effectiveness' Objective MOEs  28  Table 2-15: Summary Of Data Collection Surveys  29  Table 3-1: Evaluation Framework Methodology  33  Table 3-2: Evaluation Framework Objectives  34  Table 3-3: Objectives and Their MOEs  37  Table 5-1: Multiple-Criteria Evaluation Benefits and Costs  73  Table 5-2: Accident Costs  77  Table 6-1: Summary of the Data Collection Surveys and Records  82  Table 6-2: Parker/Curtis Street 24-Hour Traffic Volume Counts  85  Table 6-3: Westbound AVO 95% Confidence Limits and Margins of Error  88  Page xi  Table 6-4: Eastbound AVO 95% Confidence Limits and Margins of Error  90  Table 6-5: Westbound Travel Time Difference 95% Confidence Limits  94  Table 6-6: Westbound Peak Hour Travel Time Differences by Section  95  Table 6-7 Westbound Travel Speed Margins of Error  95  Table 6-8: Eastbound Travel Time Difference 95% Confidence Limits  96  Table 6-9: Eastbound Peak Hour Travel Time Differences by Section  97  Table 6-10: Eastbound Travel Speed Margins of Error  97  Table 6-11: Westbound Travel Speed Standard Deviations  99  Table 6-12: Westbound Travel Speed Margins of Error  100  Table 6-13: Eastbound Travel Speed Standard Deviations  101  Table 6-14: Eastbound Travel Speed Margins of Error  101  Table 6-15: Westbound Compliance Rates by Location  105  Table 6-16: Eastbound Compliance Rates by Location  106  Table 6-17: ICBC HOV Violation Ticket Information  108  Table 6-18: Burnaby RCMP Special Enforcement Man-Hours  110  Table 6-20: Summary of Intersections and Sections  112  Table 6-21: Holdom Avenue Traffic Volume Data (1992 -1995)  112  Table 6-22: Westbound "Before-and-After" GP Vehicular Volumes  119  Table 6-23: Eastbound "Before-and-After" GP Vehicular Volumes  119  Table 6-24: Peak Hour Travel Time Differences  121  Table 6-25: Westbound Travel Time Savings Sensitivity Analysis  122  Table 6-26: Vehicle Operating Cost Savings Sensitivity Analysis  123  Table 6-27: Air Quality Savings Sensitivity Analysis  123  Table 6-28: Maintenance Costs Sensitivity Analysis  124  Page xii  LIST OF FIGURES Figure 4-1: Sample Public Opinion Survey - Part 1  60  Figure 4-2: Sample Public Opinion Survey - Part 2  61  Figure 6-1: The Barnet/Hastings HOV Corridor  83  Figure 6-2: Peak Hour Per-Lane Efficiencies  86  Figure 6-3: Westbound Peak Hour AVO  87  Figure 6-4: Westbound Peak Period AVO  87  Figure 6-5: Westbound "Before-and-After" and Minimum AVOs  89  Figure 6-6: Eastbound Peak Hour AVO  89  Figure 6-7: Eastbound Peak Period AVO  89  Figure 6-8: Eastbound "Before-and-After" and Minimum AVOs  91  Figure 6-9: WB - "Before" Market Share  92  Figure 6-10: WB - "After" Market Share  92  Figure 6-11: EB - "Before" Market Share  92  Figure 6-12: EB - "After" Market Share  92  Figure 6-13: Westbound Travel Time Differences  94  Figure 6-14: Eastbound Travel Time Differences  96  Figure 6-15: Westbound Travel Speed Standard Deviations  99  Figure 6-16: Eastbound Travel Speed Standard Deviations  100  Figure 6-17 : Westbound Compliance Rates by Section  104  Figure 6-18: Eastbound Compliance Rates by Section  106  Figure 6-19: "Before-and-after" Intersection Accident Rates  116  Figure 6-20: "Before-and-after" Section Accident Rates  117  Figure 6-21: Westbound GP Travel Speeds  118  Figure 6-22: Eastbound GP Travel Speeds  118  Page xiii  ACKNOWLEDGEMENTS  I would firstly like to thank the British Columbia Ministry of Transportation and Highways and the National Science and Engineering Research Council (NSERC) for providing me with financial support in this research. In particular, I would like to thank Keenan Kitasaka who was my supervisor at the Ministry and who provided me with continuous feedback and support throughout my research. In addition, I would like to thank Denny Leung for including me on the 'TCH-HOV-TMP Monitoring and Evaluation Program Before Implementation' team and always helping me find what I needed at the Ministry. It was a pleasure to also have worked with Jessie Bains, the HOV Operations Engineer, in collecting additional data along the Barnet/Hastings corridor.  I would also like to thank my thesis advisor, Dr. Tarek Sayed, for his constant direction and supervision through my work. He initiated my professional partnership with the Ministry and kept me on track to finish my thesis in good time. He was patient, encouraging, and provided me with constructive suggestions to my research. Both Dr. Tarek Sayed and Keenan Kitasaka taught me some valuable skills in the technical writing process.  Special thanks to my friends and family for their encouragement, support, and most of all, patience during this time. To those few whom I am really close to and have been especially faithful and committed, I thank you even more. God is good, all the time; all the time, God is good.  Page xiv  1. INTRODUCTION  1.1 BACKGROUND In the South Coast Region of British Columbia, a high occupancy vehicle (HOV) network is being implemented by the Ministry of Transportation and Highways. As part of the implementation, HOV facilities are being developed on identified arterials and freeways within the Greater Vancouver area. The goals of the HOV network are to improve the person throughput and encourage higher occupancy on selected corridors. HOV facilities are anticipated to provide incentives such as travel time savings and reliability. It is expected that these HOV incentives will encourage drivers to shift modes from single occupant vehicles (SOVs) into high occupancy vehicles (HOVs) such as carpools, vanpools, and buses.  The first section of the HOV network was completed in the Fall of 1996 on the Barnet/Hastings HOV corridor (Route 7A), from Renfrew Street in Vancouver to loco Road in Port Moody. The HOV facility on the Barnet/Hastings corridor is comprised of both arterial and freeway sections. The curbside lane of the facility is a part-time HOV lane. The HOV lane is in operation (Monday to Friday) during the peak periods with an occupancy requirement of 2 or more people. The second section of the HOV network is currently under construction on Highway 1, between the Grandview and Cape Horn interchanges. The anticipated completion of the HOV facility is the Fall of 1998. An exclusive median HOV lane will operate throughout the day in each direction of travel. The occupancy requirement of the freeway HOV lane is scheduled to be 3 or more people.  Chapter 1: Introduction  Page 1  A framework was developed for the monitoring and evaluation of these initial and any other future HOV facilities. The purpose of developing the framework, the scope of the framework, and the application of the framework to the HOV facility on the Barnet/Hastings corridor are presented in this thesis.  1.2 PURPOSE OF THESIS In general, there is agreement among transportation professionals that "before-andafter" evaluations and ongoing monitoring be performed on HOV facilities. HOV facilities are to be evaluated in an unbiased, consistent manner that clearly evaluates the attainment of an HOV project's objectives. In British Columbia, however, there is no consistent methodology for evaluating its present and future HOV corridors.  Nonetheless, the performance of the HOV facility on the Barnet/Hastings corridor was evaluated based on certain criteria such as travel time savings and the average vehicle occupancy (AVO) of vehicles using the corridor. The facility's baseline information collected prior to the HOV lane construction was compared with data collected after its opening. The results of the "before-and-after" evaluation were documented in the "Barnet/Hastings People Moving Project Monitoring and Evaluation Program After-  Implementation Interim Report' (MoTH , 1997). It is foreseen that a similar "before-and1  after" evaluation will be performed on the future Highway 1 HOV corridor upon its completion. Therefore, extensive baseline information was collected on Highway 1 in the Fall of 1997, prior to any construction activity.  Chapter 1: Introduction  Page 2  To properly measure the effectiveness of these HOV facilities, it was determined that a HOV monitoring and evaluation framework be developed. Firstly, a comprehensive literature review of current North American HOV evaluation practice was performed to assist in developing the framework. The value of reviewing existing HOV evaluation procedures is that it gives transportation agencies a strong foundation for creating policy to properly monitor and evaluate HOV facilities.  The review of the North American HOV literature indicated that there exists considerable differences between the evaluation practices among jurisdictions. In fact, a standard methodology for the evaluation of HOV facilities does not currently exist. Therefore, there is a need to develop an HOV monitoring and evaluation framework to directly consider the goals of the South Coast Region HOV program. By developing this framework, a set of evaluation objectives tailored to the evaluation needs of the South Coast Region could be selected. The HOV monitoring and evaluation framework is based on the HOV evaluation procedures that currently exist.  1.3 SCOPE OF THESIS The HOV framework is a procedure to monitor and evaluate an HOV facility. The process of measuring the effectiveness of an HOV facility is based on evaluating a set of quantifiable objectives. As a contribution to existing HOV evaluation procedures, the set of evaluation objectives are categorized by their relationship to the goals of the South Coast Region HOV program. They are categorized as either primary, supporting, or operational objectives.  Chapter 1: Introduction  Page 3  The primary evaluation objective of the framework is directly related to the HOV network goals of improving the person throughput and increasing the use of HOVs on selected facilities. Characteristics of an HOV facility that contribute to the attainment of the HOV goals are evaluated as supporting evaluation objectives. Travel time savings is an example of a supporting objective. Attributes that ensure the protection of the HOV goals are evaluated as operational evaluation objectives. Safety is an example of an operational objective.  Each of the primary, supporting and operational objectives are described. Quantifiable measures of effectiveness (MOEs) are selected to determine if an objective is met. For example, one aspect of the primary evaluation objective is to examine if adding an HOV lane to a corridor encourages people to shift modes to an HOV. An MOE used to assist in determining if the objective was met is the average vehicle occupancy (AVO).  The data required to evaluate each of the MOEs is summarized. To enhance current practices of evaluating HOV facilities, the framework includes a procedure to calculate the statistical reliability of the data collected. Statistical procedures were often mentioned in HOV evaluations but little was reported on statistically examining the reliability of the data collection. In addition to the data requirements, a data collection methodology and comprehensive procedure for analyzing and presenting the MOEs are described.  Furthermore, the relative uncertainty of an MOE is quantified as part of the framework, where applicable. The literature review indicated that variances (or standard deviations) were examined in evaluating the travel time reliability of an HOV facility. However, the  Chapter 1: Introduction  Page 4  proposed framework in this thesis examines the statistical reliability of other MOEs such as the AVO.  Besides examining the statistical reliability, the framework describes the testing of the difference between two MOEs for statistical significance. It is important to understand if the difference between two values can be considered to be statistically significant or can be expected by chance.  Finally, a multi-criteria evaluation of an HOV facility is outlined in the framework. Only the quantifiable benefits and costs of an HOV facility are considered in a benefit-cost analysis. In addition, a sensitivity analysis methodology is described.  To obtain a clear understanding of the HOV monitoring and evaluation framework, an application was performed on the HOV facility of the Barnet/Hastings corridor. Data related to the implementation of the Barnet/Hastings HOV lane was analyzed in accordance with the framework. The numerical results of the data analysis are summarized in this thesis. As well, this thesis discusses the results on the effectiveness of the Barnet/Hastings HOV corridor with respect to the framework's quantifiable objectives (wherever data is available).  1.4 ORGANIZATION OF THESIS This thesis is divided into seven chapters. Following this introductory chapter, Chapter Two discusses the findings of an extensive literature review on HOV evaluation practices in North America. Chapter Three outlines the objectives and the evaluation  Chapter 1: Introduction  Page 5  procedures of the HOV monitoring and evaluation framework. Chapter Four summarizes the framework's evaluation objectives' measures of effectiveness, data requirements and data collection methodology as well as a statistical method for analyzing such data. Chapter Five describes the procedure for conducting a multicriteria HOV evaluation. Chapter Six summarizes the application of the framework to the Barnet/Hastings HOV corridor. Finally, a concluding chapter summarizes the findings of the thesis and suggestions for further HOV evaluation research.  1.5 SUMMARY A monitoring and evaluation framework is required to evaluate the HOV facilities of British Columbia's South Coast Region's HOV network. Building upon the experiences of other North American cities' evaluations, a framework suitable to the area was designed to evaluate the region's HOV corridors. The framework consists of a set of quantifiable objectives and measures of effectiveness that are typically included in HOV evaluations.  It expands on current practice by categorizing the evaluation objectives and by including comprehensive procedures for statistically analyzing the reliability and significance of the results. A multi-criteria methodology is presented and the application of the framework to an existing corridor is discussed. The proposed monitoring and evaluation framework is directly applicable for an HOV facility within British Columbia and can also easily be adapted for application in other jurisdictions.  Chapter 1: Introduction  Page 6  2. LITERATURE REVIEW  2.1 INTRODUCTION 2.1.1 The Evaluation of HOV Facilities An extensive literature review of current North American HOV evaluation practice was performed. The review of North American literature indicated that there is no standard practice across Canada and the United States in evaluating HOV facilities. Different HOV projects have their own set of evaluation objectives which reflect the reasons for implementing their respective HOV facility (see for example: Brown and Jacobson, 1996; Pint et al., 1995; RSMI , 1997). 1  Although each HOV evaluation has been unique in its approach to a certain extent, there appear to be some common steps taken in the evaluation process. Most studies identified a set of objectives and some measures of effectiveness (MOEs) to evaluate these objectives. Some studies identified expected target values for the MOEs (Pint et al., 1995) whereas others did not (Brown and Jacobson, 1996).  The information needs of an HOV evaluation were usually identified and the data collected accordingly. However, "before" data collected prior to the opening of an HOV facility was not always available. In other cases, "before" data was available but was not adequate to make statistically meaningful conclusions when compared to "after" data (Turnbull etal., 1991).  An example of a comprehensive examination of HOV lane projects is the evaluation carried out by the Texas Transportation Institute (TTI) which is published in several  Chapter 2: Literature Review  Page 7  reports. The TTI studies (see for example: Turnbull , 1992) examined several cities 1,2  with freeway HOV facilities in detail. The common elements of evaluating freeway HOV facilities were highlighted and a general list of the most effective and representative HOV evaluation objectives was proposed.  One particular TTI study (Turnbull et al., 1991) recommended a specific HOV evaluation procedure as a guideline to measure the effectiveness of a freeway HOV facility. The study encouraged that this suggested procedure be tailored to meet the specific set of objectives of different agencies in implementing their HOV facilities. The TTI studies, however, do not discuss the evaluation of arterial HOV facilities. In fact, very little was found from the North American literature on evaluating arterial HOV facilities, especially when compared to the number of HOV freeway evaluation reports.  2.1.2 Evaluation Objectives Defining clear and measurable objectives was often stated as one of the most important and often neglected steps in the HOV evaluation process. If these objectives are not clearly defined, evaluating an HOV facility becomes more difficult because evaluators cannot identify what they are trying to assess (Turnbull et al., 1991). Many of the objectives reviewed were merely qualitative in nature and as simple as a brief statement.  Identifying objectives that relate well to an HOV project's objectives appeared to be a difficult task. Some HOV evaluation procedures separated objectives into primary and secondary categories (RSMI , 1997). However, most HOV evaluations simply stated the 1  objectives and did not attempt to imply their relative importance.  Chapter 2: Literature Review  Page 8  A wide range of objectives have been adopted to become the standard set of HOV evaluation objectives for any specific HOV project. Table 2.1 summarizes all of the objectives identified in the literature review into twelve broad representative categories. Some objectives are similar but have been reported under various names. For example, one study examines the air quality due to the introduction of an HOV lane as the environmental objective (Turnbull et al., 1991). Each of the HOV evaluation studies reviewed in the literature examined the HOV objectives of people moving efficiency, travel time savings, safety and compliance. These four out of twelve objectives could therefore be considered the core group of objectives to be evaluated for any HOV facility.  Table 2-1: HOV Evaluation Objectives Identified in the Literature Review People Moving Efficiency* Impact of HOV on GP Lanes** Public Support Travel Time Savings* Impact of HOV on Parallel Routes Cost-Effectiveness Safety* Overall Corridor Efficiency Transit Efficiency Compliance* Air Quality/Environmental Travel Time Reliability NOTES:  * indicates an objective that was evaluated in each report reviewed ** GP = general purpose  2.1.3 Thresholds and Measures of Effectiveness The literature review indicated that there is no consensus on the level of improvement necessary for an HOV facility to be considered as a successful and effective transportation improvement (Turnbull et al., 1991). This is likely because thresholds and target values are influenced by local and corridor specific conditions such as the level of congestion and delay. In general, few studies in the literature identified improvement thresholds. The thresholds suggested were influenced by the type of HOV facility being evaluated and by the anticipated improvement the HOV facility is expected to have (Turnbull et al., 1991). Where no thresholds were established, terms like "increase" and  Chapter 2: Literature Review  Page 9  "reduce" were used to indicate at least the expected direction the HOV facility should achieve in relation to a certain objective.  HOV  evaluation objectives are evaluated by measures of effectiveness (MOEs). The  literature review identified some key characteristics of MOEs. For example, MOEs should be meaningful and quantifiable and accurately relate to the HOV project's objectives. MOEs should be realistically quantified within a project's budget and time constraints. Some of the MOEs should evaluate the operational impact of the adjacent general purpose (GP) lanes and parallel routes.  Table 2.2 is a list of the MOEs identified in the literature by several HOV evaluation reports. Because each report's MOEs are particular to its set of objectives, the MOEs were categorized under the twelve broad representative HOV evaluation objectives previously mentioned. For example, the TTI study (Turnbull et al., 1991) examines the average vehicle occupancy (AVO) as an MOE to evaluate its objective of 'People Moving Capacity' while another HOV evaluation study (Pint et al., 1995) examines the AVO  to evaluate its objective of 'Increase the Peak Hour Carpool/Vanpool Modal Split'.  Therefore, for comparative purposes, similar MOEs for different studies are categorized under the same HOV evaluation objective. For the case of the AVO, the MOE is categorized under the 'People Moving Efficiency' objective.  There are some interesting things to note about Table 2.2. Firstly, it is evident that no two list of MOEs for any HOV project were exactly the same. Secondly, although each HOV  procedure reviewed and evaluated the core group of objectives mentioned  previously, the same MOEs were not always selected to evaluate them. For example, to  Chapter 2: Literature Review  Page 10  evaluate the 'People Moving Efficiency' objective, one study selected "percent of transit use" (RSMI , 1997) as an MOE while another selected "percentage increase in transit 1  ridership" (Pint et al., 1995). These differences further reinforce the absence of a standard practice in evaluating HOV facilities and that each evaluation has been customized to reflect a specific HOV project's objectives.  2.2 OBJECTIVES The following sections of this chapter discuss in detail each of the twelve HOV evaluation objectives and their related MOEs. Each section summarizes the most common applicable MOEs identified in the literature for each respective HOV evaluation objective.  2.2.1 People Moving Efficiency The 'People Moving Efficiency' objective reflects the goal that the addition of an HOV lane to a facility should increase the people moving efficiency of a corridor. Alternatively, this objective can be restated as the 'People Moving Capacity' objective (Turnbull et al., 1991) or as the 'Increase CarpoolA/anpool Usage' objective (Pint et al., 1995). The 'People Moving Efficiency' objective best represents the overall goal that an HOV lane should encourage an increase in the number of people that choose to commute in HOVs in general (i.e., carpools, vanpools, buses, etc.). Therefore, it is not surprising to have noted that the 'People Moving Efficiency' objective was stated to be the primary objective of some reports reviewed (RSMI , 1997; Turnbull et al., 1991). Increases in the 1  number of carpoolers or bus riders are usually represented by increases in the MOEs for this objective.  Chapter 2: Literature Review  Page 11  Table 2-2: Summary of Measures of Effectiveness (MOEs) HOV  Suggested Procedures  Shirley Highway  San Bernardino  Implementation  for Evaluating the  HOV Lanes,  Freeway Busway,  Project,  Effectiveness of  Northern Virginia  Los Angeles  Vancouver  Freeway HOV facilities  (Tumbuiietai.,1991)  (RSMi*' ,1997) 3  People  AVO  Moving  % transit use  (Turnbull etai.,  1991)  (Turnbull etal.,1991) AVO  AVO  # carpools/vanpools .  Efficiency*  # bus riders  persons/peak hour  person-trips  transit ridership  modal share vehicle & person-miles  Travel Time  travel times  travel times  Safety*  accident rates  accident rates  traffic conflicts  # accidents  # accidents  Compliance*  compliance rates  violation rates  violation rates  Savings*  travel times  travel times  perceived travel times accident rates  # accidents  violation rates  enforcement level  Impact on GP  travel times  enforcement level LOS  of GP lanes  -  -  -  travel times  Lanes  Impact on  •  -  Parallel Route  volumes  Overall  person-movement  Corridor  per-lane efficiency  Efficiency Air  person throughput LOS  of facility  vehicular speeds  Quality/  Environmental  emissions  auto emissions  air pollutant emissions  fuel consumption  gasoline consumption  gallons of gasoline  general perceptions  perception of facility  vehicles-miles vehicle-hours  Public  public awareness  perception of utilization &  Support  level  whether good improvement  Cost-  B/C  ratio  value of time savings  Effectiveness  user cost savings accidents savings  Transit Efficiency  schedule adherence bus  travel speeds  vehicle productivity  travel time  Reliability  variance  Target Values  yes  bus  travel times  schedule adherence bus  Travel Time  schedule adherence  travel speeds  travel time reliability  yes  -  -  no  no  * shading indicates a core objective that was evaluated in each study • indicates mentioned but MOEs not explicitly discussed  Chapter 2: Literature Review  Page 12  Table 2-2: Summary of Measures of Effectiveness (continued) 1-394 Phase III  HOV Evaluation  An Evaluation of  Toronto Arterial HOV  Evaluation  and Monitoring  High-Occupancy  Lanes:  Interim Report,  Phase III, Seattle  Vehicle Lanes in  Operational Safety,  (Brown and  Texas, 1994  and Enforcement  Minnesota (SRF, (Pint  1993)  Effectiveness,  Jacobson,  1996)  (Henk  etal.,1995)  (Bacquie & Bahar, 1996)  etal.,1995)  auto occupancy rates  People  # carpools/vanpools  Moving  transit ridership  Efficiency*  person-trips  Travel Time  travel times  Savings*  peak hour speed  Safety*  accident rates  ACO bus  ridership  AVO  AVO  # carpools/vanpools bus  modal shift  ridership  modal split  :  :  travel times  travel times  travel times  perceived travel times  perception of safety  accident rates  accident rates  # accidents accident frequencies  accident patterns  Compliance*  compliance rates  violation rates  violation rates  violation rates  travel speeds  -  violation outcomes perception of compliance  Impact on GP  Lanes  -  travel times peak hour speed  Impact on  AVO  Parallel Route  carpool volumes % carpools diverted  Overall Corridor  LOS  person-movement  Efficiency Air  per-lane efficiency  Quality/  -  -  fuel consumption kg of emissions  air quality records  level of satisfaction  support for HOV lanes  perception of whether  support for HOV lanes  Environmental Public  person-movement  per-lane vehicle volume  queue lengths  Support  vehicular stops/starts  good improvement  •  Cost-  -  B/C  -  ratio  Effectiveness Transit Efficiency Travel Time  bus bus  travel times  -  travel speeds  speed variability  90  ,h  percentile speed  bus  travel speeds  bus  travel times  schedule adherence  perceived travel times  speed variability  travel time variation  Reliability Target Values  yes  no  yes  no  shading indicates a core objective that was evaluated in each study • indicates mentioned but MOEs not explicitly discussed  Chapter 2: Literature Review  Page 13  The evaluation of the effectiveness of this HOV evaluation objective is affected by the proportion of existing HOVs on the facility. To attain the 'People Moving Efficiency' objective, a significant amount of new carpools may need to be created. It was identified in the literature that is difficult to measure the amount of new carpools to a facility. Oftentimes, the carpools perceived as new are actually existing carpools diverted from parallel routes taking advantage of the travel time savings being provided for HOVs on the facility (see 'Impact of HOV on Parallel Routes' objective for further discussion).  The MOEs used to evaluate the 'People Moving Efficiency' objective are listed in Table 2.3. The AVO was selected as an MOE for each freeway HOV evaluation procedure reviewed. In addition, most studies examined the actual or percentage increase in the number of carpools/vanpools or bus ridership as supplementary 'People Moving Efficiency' MOEs. Other commonly used MOEs includes modal share, modal split, or modal shift. Table 2-3: 'People Moving Efficiency' Objective MOEs  Most Common MOEs  Similar MOEs  average vehicle occupancy bus/transit ridership modal split person-trips percent transit use number of carpools/vanpools vehicle & person-miles  auto occupancy rates number of bus riders modal share, modal shift persons/peak hour  2.2.2 Travel Time Savings The 'Travel Time Savings' objective addresses the main incentive for commuters to switch to a higher occupancy mode. The HOV lane should be providing a relative travel time savings advantage to the occupants of HOVs. Fuhs (1993) indicates that the HOV  Chapter 2: Literature Review  Page 14  lane's ability to provide such a travel time savings as being one of its most reliable predictors of effectiveness. Fuhs (1993) also indicated that commuters who benefit the most from the travel time savings (i.e., long distance commuters) would most likely be encouraged to switch to a higher occupancy mode.  A threshold of at least 1 minute per mile is commonly accepted to be the minimum travel time savings necessary for a freeway HOV lane to be effective (see for example: Brown and Jacobson, 1996; RSMI , 1997; Turnbull et al., 1991; Henk et al., 1995). This 1  threshold is based on the "Freeway High-Occupancy Vehicle Lanes and Ramp Metering  Evaluation Study" (D. Baugh and Associates, 1979)* and no supportive explanations for this threshold was presented in the literature.  The 'Travel Time Savings' objective was generally evaluated with either the travel times or travel speeds MOEs. A couple of studies identified the perceived travel time of commuters as an additional MOE. Table 2.4 is a list of the travel time savings MOEs. Table 2-4: 'Travel Time Savings' Objective MOEs  Most Common MOEs  Similar MOEs  travel times perceived travel times  travel speeds  * NOTE: as reported by Henk et al. (1995)  2.2.3 Safety The 'Safety' objective assesses whether adding an HOV lane to a facility is a safe transportation improvement.  The TTI studies (Turnbull , 1992) suggest that the 1,2  implementation of an HOV lane should not decrease the level of safety of a facility nor should the accident rate on the HOV lane be greater than that of the GP lanes.  Chapter 2: Literature Review  Page 15  However, it is often the case that current accident reporting practice does not include referencing the accident to either the HOV SRF,  or GP lane specifically (see for example:  1993; RSMI , 1997). 3  An example of a comprehensive evaluation of the safety conditions associated with  HOV  lanes was performed by the Institute of Transportation Studies at the University of California (Golob and Recker, 1988). Two freeway HOV  lane's safety impacts in the  greater Los Angeles area were evaluated. It was concluded in the paper that there was "no  adverse effect on safety conditions" due to HOV operation on one of these HOV  facilities. A slight increase in the number of accidents was observed on the second  HOV  facility. However, it was concluded that the increase was "no greater than 2 percent over and  The  above the level that would be expected from mixed-flow operation of the lane".  accident patterns and locations of the Los Angeles HOV  facilities were also  summarized in the University of California paper (Golob and Recker, 1988). The exposure to accidents migrated downstream to the traffic bottlenecks of both HOV facilities. The particular bottlenecks described were categorized as "severe" and it was mentioned that in other jurisdictions where "the downstream bottlenecks are less severe, overall safety improvements can  be expected".  Often, information on the total number, severity, and type of accidents was not available for  a "before-and-after" comparison. One study recommended that a minimum of 3  years of "before" data be collected to perform a statistically meaningful "before-andafter" analysis (RSMI , 1997). Other studies also suggested that the safety of a 2  representative group be monitored to determine the "after" changes in accident rates  Chapter 2: Literature Review  Page 16  that are common to the region (see for example: Turnbull et al., 1991; Sayed , 1996). A 1  control facility is presently being monitored as part of the on-going evaluation of the safety conditions of the Houston HOV project (Turnbull et al., 1991).  Traffic conflicts at the merging points of HOV and GP traffic and at the terminus points of HOV facilities were occasionally observed (Turnbull , 1992). A direct assessment of 2  the traffic conflicts at intersections and/or terminal points is useful when evaluating the safety of HOV facilities to complement insufficient or inconclusive accident data (RSMI , 2  1997).  Table 2.5 is a complete list of MOEs used to evaluate the 'Safety' objective. The primary MOEs are accident rates and the number of accidents. Other safety MOEs include accident frequencies, patterns and locations as well as vehicle breakdown rates (Henk et al., 1995). One HOV evaluation reviewed in the literature included an MOE based on the public's opinion of the safety of the HOV lane (Brown and Jacobson, 1996). Table 2-5: 'Safety' Objective MOEs  Most Common MOEs accident rates number of accidents accident locations traffic conflicts perception of safety  Similar MOEs accident frequencies accident patterns  2.2.4 Compliance The 'Compliance' objective protects the HOV operation by evaluating if violators of the HOV lane are being sufficiently discouraged by the level of enforcement. It is important  Chapter 2: Literature Review  Page 17  to remain within an accepted range of compliance to protect the travel time savings and reliability benefits for HOV  The  commuters.  violation rate of an HOV  lane is calculated by dividing the total number of violators  of the occupancy requirement by the total number of cars in the HOV  lane, expressed  as a percentage. Similarly, the compliance rate is the percentage of vehicles in the  HOV  lane who meet the occupancy requirement and is equal to 100% minus the violation rate. The violation or compliance rates are determined from either vehicle classification and  occupancy surveys or from the violation data of local enforcement agencies.  Different desired thresholds for compliance were observed in the literature.  A  compliance rate between 85% and 90% was most frequently stated as acceptable for freeway HOV  The  operations.  compliance rate observed is dependent on numerous factors. The number of  officers dedicated to enforcing the HOV  occupancy requirement and the availability of  adequate locations for enforcement are most likely the two most important. Supporting programs such as the institution of a region-wide HERO program (Brown and Jacobson, 1996), adequate penalties (Ulberg and Jacobson, 1993), and proper geometric design of the HOV  lane (Turnbull , 1992) were all identified to be secondary but still very 2  important factors.  A study from Washington State Transportation Center (TRAC) examined in detail  HOV  enforcement and the objective of compliance (Ulberg and Jacobson, 1993). It was concluded that each of the three types of enforcement techniques implemented on their regional HOV  network had a positive effect on the compliance rate and none was  Chapter 2: Literature Review  Page 18  considered to be more effective than the others. The three types of enforcement techniques included: intensive enforcement, once a week saturation enforcement, and once a week stationary enforcement. The Washington State TRAC study also complimented its violation rate data with a public opinion survey. Amongst other questions, respondents were asked if they support the concept of HOV lanes as well as admit if they had ever violated the HOV occupancy requirement.  Enforcement fines and strategies were often discussed in the literature. In California, an increase in the HOV violation fine to $271 (US$) for the first offence and public awareness concerning the magnitude of the fine decreased its violation rate by about 65% (Fuhs, 1993). Other proposed enforcement strategies were to penalize second and third time HOV violators greater than first time offenders and to ticket by mail to maximize the allocated resource time of local enforcement agencies (Ulberg and Jacobson, 1993). The Washington State Department of Transportation is currently researching the outcomes of its HOV violations as a compliance follow-up strategy (Brown and Jacobson, 1996).  It is strongly recommended that the level of enforcement (the enforcement details: # of officers, shifts and hours) be kept in records and correlated to either the compliance rate on the corridor or the number of citations being issued (RSMI , 1997). It is perhaps the 3  most effective way of evaluating the performance of enforcement versus compliance but has yet to be fully researched and documented.  The violation rates and compliance rates are the MOEs most frequently used to evaluate the 'Compliance' objective. Table 2.6 summarizes the 'Compliance' objective MOEs.  Chapter 2: Literature Review  Page 19  Table 2-6: 'Compliance' Objective MOEs  Most Common MOEs  Similar MOEs  violation rate perception of compliance outcomes of HOV violation tickets  compliance rate  2.2.5 Impact of HOV an GP Lanes The 'Impact of HOV on GP Lanes' objective examines the operational impact of adding an HOV lane on its adjacent general purpose (GP) lanes. The addition of an HOV lane to an existing facility should not adversely affect the operation of the GP lanes. In other words, the operating speed and level of service (LOS) being provided on the GP lanes should not be degraded due to the addition of the HOV lane (Turnbull et al., 1991). The 'Impact of HOV on GP Lanes' objective was rarely quantified in the review of HOV evaluations. When evaluated, travel times or speeds of the GP lanes were often compared to on a "before-and-after" basis as an MOE. However, the comparison was often made as part of the evaluation of the 'Travel Time Savings' objective as opposed to a separate 'Impact of HOV on GP Lanes' objective. A comparison of the "before-andafter" LOS on the GP lanes was another MOE identified in the literature (Turnbull et al., 1991). Table 2.7 summarizes the 'Impact of HOV on GP Lanes' objective MOEs. Table 2-7: 'Impact of HOV on GP Lanes' Objective MOEs  Most Common MOEs  Similar MOEs'  travel times LOS of GP lanes  travel speeds  2.2.6 Impact of HOV on Parallel Routes The 'Impact of HOV on Parallel Routes' objective addresses the operational impact of adding an HOV lane on its alternative parallel routes. The most significant impact the  Chapter 2: Literature Review  Page 20  HOV  corridor may have on its parallel routes is its tendency to attract and divert HOVs  onto the HOV lane to benefit from the travel time savings being provided. It is assumed that some HOVs from parallel routes will change their commuting patterns to make use of an adjacent HOV facility.  The 'Impact of HOV on Parallel Routes' objective is related to the 'People Moving Efficiency' objective. An increase in a facility's people moving efficiency is proportional to the increase in a facility's number of carpools. The increase in the number of carpools may be doubly counting benefits as existing HOVs are diverted from alternative parallel routes. To understand the net increase in the number of carpools, diverted HOVs should be taken out of the estimate of the benefits.  Information necessary to evaluate this objective is the "before-and-after" percentage of HOVs on the parallel routes. The percentage difference can be used to approximate the number of vehicles diverted from parallel routes. However, it should be made clear that the suggested amount of diversion is difficult to measure. It is only possible to determine a more accurate number with an extensive survey and data collection effort. Network efforts are most difficult to assess because they are affected by many other factors. Few HOV  evaluation studies discussed the 'Impact of HOV on Parallel Routes' objective.  However, Henk et al. (1995) presented results of a survey that estimated the percentage of carpoolers who were diverted from a parallel route as well as the AVO of the parallel routes. The MOEs identified from the literature for the 'Impact of HOV on Parallel Routes' objective are summarized in Table 2.8.  Chapter 2: Literature Review  Page 21  Table 2-8: 'Impact of HOV on Parallel Routes' Objective MOEs  Most Common MOEs  Similar MOEs  percent of mainline carpoolers diverted AVO of parallel routes travel times of parallel routes  carpool volume counts  2.2.7 Overall Corridor Efficiency The 'Overall Corridor Efficiency' objective focuses on the operational impact of adding on HOV lane on the entire HOV facility, GP lanes included. The person-movement MOE is typically used as an MOE for this objective. It is suggested that the percentage increase in the total (GP plus HOV) person volume should be greater than the percentage increase in the amount of directional lanes added to a facility (Henk et al., 1995). The "before-and-after" level of service (LOS) of the entire HOV facility was also presented frequently as an MOE to evaluate the 'Overall Corridor Efficiency' objective.  Another approach for evaluating this objective is suggested by the TTI studies. A perlane efficiency MOE that is equal to the total person volume of the facility multiplied by the average vehicle operating speed is suggested (Turnbull et al., 1991). It is expected that the per-lane efficiency of a facility should increase with the implementation of an HOV lane. Other MOEs mentioned in the literature were person throughput (Turnbull et al., 1991) and queue lengths (SRF, 1993). The person throughput MOE was expressed in person-miles of travel per hour and the queue lengths were expressed in number of vehicles. Table 2.9 summarizes the MOEs for the 'Overall Corridor Efficiency' objective. Table 2-9: 'Overall Corridor Efficiency' Objective MOEs  Most Common MOEs  Similar MOEs  person movement LOS of facility per-lane efficiency queue lengths  person throughput operating speeds per-lane vehicle volume  Chapter 2: Literature Review  Page 22  2.2.8 Air Quality/Environmental The 'Air Quality/Environmental' objective reflects the attitude that adding an HOV lane should have more favourable impacts on the environment than the "do nothing" or "add a GP lane" alternatives (Turnbull et al., 1991). Few detailed analyses were performed regarding this objective.  Neither a decrease in vehicular volumes nor reduction in emissions is expected with the implementation of an HOV lane. Therefore, simplistic computer models were sometimes used to estimate and compare the environmental impacts of three alternatives: do nothing, add a GP lane, and add an HOV lane (Henk et al., 1995). The modelling was performed by keeping the demand constant and adjusting the AVO appropriately for all three alternatives. The 'Air Quality/Environmental' objective was evaluated by comparing the differences in the environmental impact outputs of the computer simulation. These outputs (kg of emissions, fuel consumption) are the most frequently used MOEs.  One particular study examined vehicular starts and stops and air quality records as alternative  MOEs  (Bacquie  and  Bahar,  1996).  The complete  list  of 'Air  Quality/Environmental' objective MOEs is summarized in Table 2.10. Table 2-10: 'Air Quality/Environmental' Objective MOEs r  Most Common MOEs  Similar MOEs  fuel consumption kg of emissions vehicle-miles vehicular starts/stops air quality records  gasoline consumption auto emissions vehicle-hours  Chapter 2: Literature Review  Page 23  2.2.9 Public Support The 'Public Support' objective recognizes that public support of the concept and operation of HOV lanes is important and should therefore be evaluated. The TTI studies indicated that the majority of the public should perceive HOV lanes as being adequately utilized and as a good transportation improvement (Turnbull ,1992). 1  Telephone or mail out surveys are most frequently used to evaluate the public support of HOV  facilities. They are generally conducted to measure the overall attitudes and level  of satisfaction of all commuters: carpoolers, bus riders, and those who drive alone (see for example: SRF, 1993; Brown and Jacobson, 1996). The preferred method of a public opinion survey is a mail out survey that is sent to actual users of an HOV facility whose, license plates are observed and addresses retrieved.  The percentage of people surveyed who support the concept and/or operation of the HOV  lanes are the most frequently observed MOEs for this objective. The public  awareness level of the HOV lane was also suggested as a similar MOE (RSMI , 1997). 2  The 'Public Support' MOEs are summarized in Table 2.11.  The violation rate is often used to complement public opinion survey MOEs. High violation rates suggest that users of the HOV facility do not support or respect the concept and/or operation of the HOV lane. However, as discussed earlier, violation rates are influenced by more than just public support. All phone calls, letters, and press releases concerning the operation of the HOV facility were sometimes monitored and reviewed as alternative MOEs for this objective (Turnbull ,1992). 2  Chapter 2: Literature Review  Page 24  Table 2-11: 'Public Support' Objective MOEs  Most Common MOEs  Similar MOEs  support for HOV lanes public awareness level violation rates phone calls  perception of HOV lane as a good improvement perception of HOV lane utilization compliance rates letters, press releases  2.2.10 Transit Efficiency The 'Transit Efficiency' objective evaluates any improved performance of the transit system in general. Most reports reviewed examined either the schedule adherence or the travel times of buses as MOEs for this objective. However, HOV projects did not often evaluate the 'Transit Efficiency' objective as its own separate category. Transit system impacts were usually examined as part of the general 'People Moving Efficiency' objective.  The data needed to evaluate the 'Transit Efficiency' MOEs is dependent on a significant amount of co-operation with the in-place transit system agency. Therefore, it is often not possible to quantify the transit MOEs without the help of a transit agency's resources which is the main reason why the objective is not exclusively evaluated.  The complete list of MOEs used to evaluate the 'Transit Efficiency' objective are listed in Table 2.12. One particular study evaluated the actual as well as the perceived bus travel times as MOEs for this objective (Bacquie and Bahar, 1996). The TTI study was the only report reviewed that suggested evaluating the vehicle productivity (operating cost per vehicle-mile, passenger, and passenger-mile) as an MOE in addition to schedule adherence or bus travel speeds (Turnbull et al., 1991).  Chapter 2: Literature Review  Page 25  Table 2-12: 'Transit Efficiency' Objective MOEs  Most Common MOEs  Similar MOEs  schedule adherence bus travel speeds perceived bus travel times vehicle productivity  on-time performance bus travel times  2.2.11 Travel Time Reliability The 'Travel Time Reliability' objective examines the variability of the travel times or speeds of the HOV lane in comparison to its adjacent GP lane(s). It was often mentioned as an objective in the literature yet was rarely quantified (Turner et al., 1994). To measure the travel time variability of both the HOV and GP lanes, continual sampling over an extended period of time is required.  The MOEs used most frequently to evaluate the 'Travel Time Reliability' objective are the variance or standard deviations of either the travel times or travel speeds. However, a paper written specifically on travel time reliability stated that travel speeds provide better estimates of travel time reliability because speed distributions are less skewed (more normal) than travel time distributions (Turner et al., 1994). The paper also presented the use of statistical tests to analyze the differences in the HOV and GP travel speed results. Table 2.13 summarizes the 'Travel Time Reliability' MOEs.  MOEs used to evaluate the travel time reliability particular to Washington State are the speed range observed and the 90 percentile speed (Brown and Jacobson, 1996). th  These MOEs are used to evaluate the Washington State Freeway HOV System Policy that "HOV lane vehicles should maintain or exceed an average speed of 45 mph or  Chapter 2: Literature Review  Page 26  greater at least 90% of the times they use that lane during the peak hour (measure for a  six-month period)" (Brown and Jacobson, 1996). Table 2-13: 'Travel Time Reliability' Objective MOEs  Similar MOEs travel time variance travel speed standard deviations 90 percentile speed  travel speed variance speed range observed  th  2.2.12 Cost-Effectiveness The 'Cost-Effectiveness' objective examines if the implementation of an HOV lane is a transportation improvement whose benefits outweigh its costs. To determine an HOV project's cost-effectiveness,  a benefit/cost analysis is usually performed. The cost-  effectiveness MOE most frequently quantified is the benefit/cost ratio. However, the marginal benefit/cost ratio and marginal net present value (NPV) are two other MOEs that have been used to evaluate the cost-effectiveness of an HOV project (Ulberg and Jacobson, 1988). The 'Cost-Effectiveness' MOEs are summarized in Table 2.14.  To calculate the benefit/cost ratio, many assumptions are made regarding an appropriate discount rate, salvage value of the project, and the dollar value of travel time savings. Each component of the ratio, whether benefit or cost, is either annualized or converted to a net present value. The benefits such as the travel time saved are related to the users of the HOV facility. The costs are related to the transportation agencies responsible for the operation of the HOV lane. The costs typically include the following: construction costs (including right-of-way), maintenance costs, operation costs, and enforcement costs.  Chapter 2: Literature Review  Page 27  It is recommended in the literature that only the travel time savings benefits of the users of the HOV the facility be quantified in the benefit/cost ratio (see for example: Henk et al., 1991; Turnbull , 1992). By not including other benefits such as congestion savings, 2  vehicle operating cost savings, and improved trip reliability, a more conservative benefitcost ratio is produced. One particular paper compared the cost-effectiveness of existing HOV lanes in Seattle with the "do nothing" and "add a GP lane" alternatives (Ulberg and Jacobson, 1988). A sensitivity analysis that included worst case, best case, and base values for the benefits and costs was performed for each alternative.  In another HOV evaluation paper, the minimum mode shift or amount of travel time savings required for the HOV lane to be cost-effective was quantified (Henk et al., 1991). It was presented that a minimum travel time savings of 0.8 minutes per mile is required for a Houston HOV lane to be cost-effective. Alternatively, the paper suggested that a HOV lane's annual total value of travel time savings be equal to or greater than .10 percent of its total construction costs to be cost-effective. Table 2-14: 'Cost-Effectiveness' Objective MOEs  Most Common MOEs  Similar MOEs  benefit/cost ratio net present value  marginal benefit/cost ratio marginal net present value  2.3 DATA COLLECTION 2.3.1 Data Collection Techniques After objectives and appropriate MOEs are established as part of an HOV evaluation and monitoring program, the information needs are identified. The frequency of conducting data collection is dependent on the available resources, the type of HOV  Chapter 2: Literature Review  Page 28  facility being evaluated (arterial vs. freeway), and the maturity of the system (newer facilities require higher frequency) (Turnbull et al., 1991). Data is often collected for the purpose of HOV evaluation as part of other data collection efforts to maximize the available resources (Turnbull et al., 1991). The variety of potential data collection surveys is summarized in Table 2.15. Table 2-15: Summary Of Data Collection Surveys travel time runs enforcement level violations  classification & occupancy surveys transit information/scheduling volume counts  public opinion survey accident records traffic conflict survey  Many surveys are conducted on both the HOV mainline as well as its parallel routes. The tendency to collect excessive data is balanced against obtaining too little data for any statistically meaningful results. Therefore, sample sizes are selected on a project specific basis to ensure that representative conclusions on the effectiveness of the HOV lane are made from the data collection surveys. Some of the more common types of data collection surveys are: travel time tuns, vehicle classification and occupancy counts, and public opinion surveys.  2.3.2 Travel Time Runs Two alternative methods are used to collect data for travel time runs: the 'average' (or 'floating') car technique and the license plate method (Turnbull et al., 1991). The results of the travel time runs are dependent on the individual driver's style (Brown and Jacobson, 1996). The 'average' car technique is used to help identify problem areas such as bottlenecks or excessively long delays at intersections along a corridor. However, a limited number of travel time runs are performed such that each run is considered to be statistically independent of one another. A minimum interval spacing of  Chapter 2: Literature Review  Page 29  15 minutes between trips is recommended (Turnbull et al., 1991). The license plate method is capable of producing a more accurate representation of the total trip time over a range of volume conditions but is much more costly (Brown and Jacobson, 1996). The license plate method to collect travel time runs has been used in Seattle. However, they have but recently switched to the 'average' car technique(Brown and Jacobson, 1996). To assist in collecting the large sample size of travel time runs required for the 'travel time reliability' objective, regular commuters or transit agencies were suggested as possible sources to perform travel time runs (RSMI ,1997). 3  2.3.3 Vehicle Classification and Occupancy Counts To determine the number of manual occupancy counts an agency needs to collect on a specific corridor, the regional variance in occupancy rates should be determined (Turnbull et al., 1991). Occupancy counts should be collected on the HOV mainline as well as on parallel routes if the data collection budget permits.  The number of riders on a bus or in a vanpool are only coarsely estimated during vehicle classification and occupancy surveys.  Because of the higher degree of uncertainty  associated with these estimates, buses and vanpools exhibit a higher variance in their AVOs than regular automobiles (Turnbull et al., 1991). Therefore, TTI strongly recommends using a stratified random sampling technique as opposed to a simple random sampling technique (Turnbull et al., 1991). With stratified random sampling, the number of bus and vanpool occupancy counts is maximized and a more significant estimate of the AVOs is achieved.  Chapter 2: Literature Review  Page 30  2.3.4 Public Opinion Surveys Many types of public opinion surveys for the purpose of HOV evaluation were reviewed in the literature (see for example: Turnbull et al., 1991; Brown and Jacobson, 1996; and SRF, 1993). Each of the surveys reviewed a included a section for comments at the end of the survey. Some of the questions asked of the respondents that were common to all of the surveys are highlighted below: • What is your usual present mode of travel ? • How often do you make this commute ? • How many people (and/or family members) do you regularly commute with ? • Prior to the HOV lane, what was your usual mode of travel ? • Why do you choose to carpool (or not to) ? • What is your home zip code (or city) ? • What is the destination of your trip ? • Do you feel that the HOV lane is adequately utilized ? • What is your age and gender ? • What is the highest level of education you have completed ?  Chapter 2: Literature Review  Page 31  3. H.O.V. MONITORING AND  EVALUATION FRAMEWORK  3.1 INTRODUCTION As previously mentioned in Chapter One, a region-wide high occupancy vehicle  (HOV)  network is being implemented in the South Coast Region of British Columbia. An important step in the implementation of a regional HOV  network is to evaluate the  effectiveness of its existing HOV facilities. Accordingly, an HOV framework was developed for the monitoring and evaluation of British Columbia's HOV  facilities. The  framework is based on the HOV evaluation experience of other North American jurisdictions.  The  HOV monitoring and evaluation framework is comprised of three sections: an  evaluation framework, a monitoring framework, and a multi-criteria evaluation. The first two sections are described in this chapter. The multi-criteria evaluation is presented in Chapter Five of this thesis.  3.2 THE The  H.O.V. EVALUATION FRAMEWORK  evaluation framework described herein is based on existing HOV evaluation  practices. It was developed specifically for application in the Greater Vancouver area of British Columbia. It is also applicable to HOV  projects in other jurisdictions. This section  of the chapter is the framework procedure to evaluate an HOV  corridor based on a set  of quantifiable objectives. The general evaluation methodology of the framework is summarized in Table 3.1.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 32  Firstly, a description of the objective being evaluated and its relation to goals of implementing an HOV corridor is presented. Secondly, the measures of effectiveness (MOEs) that were selected to assist in determining the attainment of the objective are described. The data requirements to evaluate each objective are listed as well as a data collection methodology. The data is analyzed according to the type of HOV facility (arterial vs. freeway). Data collected on the mainline as well as routes parallel to an HOV facility are examined. The statistical reliability of the data is investigated and the differences between two MOEs are tested for statistical significance. Finally, a discussion of the results is introduced. Table 3-1: Evaluation Framework Methodology  OBJECTIVE  "  Measures of Effectiveness Data Requirements Data Collection Methodology Data Analysis Arterial vs. Freeway Analysis Parallel Routes Analysis Statistical Reliability Statistical Significance Discussion of Results  3.2.1 Objectives The set of objectives selected for the HOV evaluation framework are its foundation. The objectives of the framework were selected to strongly reflect the HOV network goals. Table 3.2 lists the set of objectives selected for the HOV monitoring and evaluation framework. The objectives have been categorized as either primary, supporting, or operational objectives based on their relationship to the HOV network goals.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 33  Table 3-2: Evaluation Framework Objectives  PRIMARY OBJECTIVE Person Throughput  SUPPORTING OBJECTIVES  OPERATIONAL, OBJECTIVES  Travel Time Savings Compliance Travel Time Reliability Safety Public Support Impact on GP Lanes  The 'Person Throughput' objective is directly related to the HOV network goals of improving the person throughput and encouraging higher occupancy. Therefore, the 'Person Throughput' objective is the primary objective of the evaluation framework.  Some characteristics of an HOV facility contribute to the attainment of the HOV network goals. For example, a travel time savings benefit for HOVs encourages people to shift modes to an HOV. As a result, it is anticipated that the person throughput would increase as well as the HOV share of the commuting traffic. Therefore, a noticeable travel time savings benefit helps to support the attainment of the HOV network goals. The evaluation framework measures the effectiveness of these characteristics as supporting objectives. They include 'Travel Time Savings', 'Travel Time Reliability', and 'Public Support'.  Other characteristics of a newly constructed HOV facility may not contribute to the success of the HOV project goals, however, they may inherently protect them. For example, a low violation rate will most likely not have an effect on the primary objective of person throughput. However, a high number of violators in the HOV lane could potentially negatively affect the travel time savings and reliability benefits for HOVs. Therefore, operational objectives such as 'Compliance' are also assessed as part of the  Chapter 3: H O V Monitoring and Evaluation Framework  Page 34  evaluation framework. Other examples of operational objectives include 'Safety', and 'Impact on GP Lanes'.  The  evaluation framework set seven objectives is very similar to the twelve broad  categories described in the literature review. However, the HOV  evaluation framework  excludes the following three objectives: 'Transit Efficiency', 'Impact of HOV  on Parallel  Routes', and 'Air Quality'. The 'Cost-Effectiveness' objective is described in the  HOV  monitoring and evaluation framework as part of the multi-criteria evaluation and neither a primary, supporting, nor operational objective.  The  'Transit Efficiency' objective was determined to be redundant and unnecessary to  evaluate as a separate objective. The objective's MOEs directly relate to the productivity and  efficiency of the transit fleet service. The operation of the transit fleet in the Greater  Vancouver area is under the direction of BC Transit. Nevertheless, the number of transit passengers and bus operating speeds are still significant variables that should be examined when evaluating an HOV  lane's effectiveness. Therefore, these variables are  taken into account when evaluating the preferred set of seven objectives.  Similar to the 'Transit Efficiency' objective, it was determined to be nonessential to evaluate air quality as part of the HOV of an HOV  evaluation framework. The effect the construction  facility will have on the region's air quality is an important environmental  aspect. However, improving air quality is neither a primary nor supporting objective for building an HOV  lane. It may also be difficult to quantify the air quality of a corridor.  Nevertheless, it is anticipated that the addition of an HOV  lane to a facility would be less  obtrusive on air quality than other alternatives.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 35  Regarding the 'Impact of HOV on Parallel Routes', the literature review recommended that such an operational impact be evaluated as a separate objective. It is a very important aspect that must be accounted as part of the HOV evaluation. In the framework, the impact an HOV lane will have on its adjacent routes is evaluated as part of each of the preferred set of objectives, if relevant. Therefore, it was not required to evaluate the impact on parallel routes as a separate objective.  3.2.2 Measures of Effectiveness Each of the framework's primary, supporting, and operational objectives are evaluated by mean of measures of effectiveness (MOEs). These MOEs are selected such that they directly relate to the preferred set of objectives. An MOE selected for the evaluation of an objective is used to determine the effectiveness of that objective. Therefore, MOEs are usually quantifiable numerical units representative of the effectiveness of the respective objective being evaluated. An MOE's absolute value, magnitude, and trend over time are used to help determine and interpret the effectiveness of an HOV lane.  For example, the average vehicle occupancy (AVO) is an MOE used to determine if the 'Person Throughput' objective has been met. AVO is calculated by dividing the total number of passenger occupants by the total number of passenger cars. Significant increases in the AVO are indicative that an HOV facility has been effective in increasing the average number of occupants per car along the corridor. This type of trend in the AVO indicates that people are shifting to a higher occupancy mode since the introduction of the HOV lane, buses excluded. If people are shifting to HOVs along a  Chapter 3: HOV Monitoring and Evaluation Framework  Page 36  corridor, the person throughput will most likely improve. Therefore, AVO is a quantifiable and representative MOE for the 'Person Throughput' objective.  The specific MOEs for each of the framework's seven objectives are summarized in Table 3.3. The detailed description of the MOEs are found in Chapter Four of this report. Table 3-3: Objectives and Their MOEs  CATEGORY Primary  Supporting  Operational  OBJECTIVE Person Throughput Travel Time Savings Travel Time Reliability Public Support Compliance Safety Impact on GP Lanes  MOE Per-Lane Efficiency AVO HOV Market Share Travel Time Difference Travel Speed Standard Deviation Support for HOV Lane Compliance Rate Accident Rate GP Travel Speed  3.2.3 Data Requirements The MOEs of the objectives are calculated from data collected on the HOV facility. Therefore, the next step of the HOV evaluation framework is to determine the type of data that is required to evaluate the effectiveness of the objectives. The data requirements are often the same for more than one of the objectives. For example, vehicle classification and occupancy data is required for evaluating both the 'Person Throughput' and 'Compliance' objectives. The data requirements of the HOV evaluation framework are outlined below. They are explained in detail in Chapter Four of this report. • vehicle classification & occupancy counts • travel time surveys  Chapter 3: HOV Monitoring and Evaluation Framework  Page 37  • public opinion survey • accident statistics • traffic volumes  3.2.4 Data Collection Methodology Based on the data requirements of the objectives, a data collection methodology is established. Collecting data with the purpose of evaluating an HOV facility is similar to collecting data for the evaluation of any other "before-and-after" transportation improvement. It is necessary to ascertain that there is an adequate amount of data collected such that statistically reliable conclusions can be made from the results. Information that is uniquely characteristic to the data collection methodology of evaluating an HOV facility is emphasized in the framework.  3.2.5 Data Analysis Before evaluating the effectiveness of the HOV facility, the data should be converted into a presentable format that is studied in finer detail so that conclusions can be made. The first step of the data analysis is to calculate the MOEs based on the samples of data collected. An MOE is calculated for each sample of data. The average of these values is presented as the mean MOE of the HOV facility. However, this mean MOE (based on samples) is only an estimate of the true population mean. In previous HOV evaluations, these MOE sample mean estimates were presented as absolute values with no uncertainty associated with them. In other words, it was assumed that the mean MOE  estimate (calculated from only a sample of data) was in fact the mean of the  population.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 38  However, the mean MOE estimate is not always representative of the true population mean. Therefore, it is insightful to understand how close the mean MOE estimate is to the actual population mean. For this purpose, confidence limits of a MOE are calculated, where applicable. A confidence limit is the interval that has a specific probability of capturing the true population mean. Generally, the probability of capturing the true mean is selected as close to unity as possible. Therefore, a 95% level of confidence was selected for constructing confidence limits for HOV evaluation purposes. In other words, an evaluator is confident that a calculated MOE  will be somewhere in between the upper  and lower bounds of the confidence limits 19 out of 20 times. The methodology for calculating a confidence limit is found in Appendix A - Statistical Analysis.  The upper and lower bounds of an MOE's confidence limits are also important as input into the multi-criteria section of the HOV evaluation framework. The net present value (NPV) of an HOV facility is firstly determined with the mean MOE estimates to produce the most likely value. The NPV is then recalculated for the both the upper and lower bounds of the MOEs as part of a sensitivity analysis to produce the worst and best case scenario values.  The second step of the data analysis is to present the MOEs of the framework in an appropriate format. The presentation of the MOEs as either actual numbers in contrast to percent increases or in tables as opposed to graphs is important, depending upon the audience the report is addressing. The most suitable presentation format of the MOEs is outlined in the data analysis section.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 39  3.2.6 Arterial vs. Freeway Analysis The HOV evaluation framework is a flexible procedure that can be applied to evaluate the effectiveness of either an arterial or freeway HOV facility. Realistically, however, it is sometimes preferred to slightly modify or tailor the application of the evaluation framework to the specific type of HOV facility being evaluated. For example, the HOV facility on the Barnet/Hastings corridor is comprised of three very characteristic sections. After applying the HOV evaluation framework to the corridor, it was determined that each section be examined on its own for many of the objectives. The initial application of the HOV evaluation framework to the Barnet/Hastings corridor was tailored to gain a better insight on the effectiveness of its HOV lane.  However, there are some general characteristics of an arterial HOV facility that are unlike that of a freeway HOV facility. Where applicable, any concerns and lessons learned between the application of the HOV evaluation framework to arterial and freeway HOV facilities is highlighted.  3.2.7 Parallel Routes Analysis As mentioned in the discussion of the HOV evaluation framework's objectives, the operational impact the addition of an HOV lane has on its parallel routes is a very important aspect that should be accounted for. However, it is not recommended that all of the MOEs be calculated for the parallel routes. There is a trade-off between the cost of collecting data on adjacent routes and the need for such data to gain further insight into the effectiveness of an HOV facility. Therefore, it is important to spend time  Chapter 3: HOV Monitoring and Evaluation Framework  Page 40  deciding which, if any, of the MOEs need to be quantified for the parallel routes as well as the HOV facility.  3.2.8 Statistical Reliability The data collected are samples of an entire population. The statistical reliability of these samples should be investigated as part of the HOV evaluation framework. The adequacy of data collected is determined based on its sample size, mean, and variance. For the travel time runs, the statistical reliability of a sample of data is tested at the specified permitted error for "before-and-after" studies as recommended in the 'ITE Transportation and Traffic Engineering Handbook' (ITE, 1992). For AVO data, the  margin of error associated with the samples of vehicle classification and occupancy counts is derived. The AVO margins of error are presented as percentages of the mean.  3.2.9 Statistical Significance A methodology for evaluating the statistical significance of an MOE is described within the HOV evaluation framework. The statistical significance of an MOE is tested to determine if the difference between two values can be explained by random chance. The type of statistical test used is dependent on the method the data was collected and analysed. The results are tested at the 95th percentile. Examples of the framework's suggested statistical tests are found in Appendix A - Statistical Analysis.  3.2.10 Discussion of Results A discussion should be conducted on the results of the MOEs of the objectives. Particular emphasis should be placed in the comparison of the peak hour and peak  Chapter 3: HOV Monitoring and Evaluation Framework  Page 41  period results. Likewise, the results of the HOV mainline and its parallel routes should be compared. Any trends over time should be discussed if the evaluation conducted is part of the ongoing HOV monitoring and evaluation program.  3.3 THE H.O.V. MONITORING FRAMEWORK The HOV evaluation framework is a procedure to assess the effectiveness of an HOV lane. The evaluation of an HOV facility is performed at particular times throughout the lifetime of an HOV project. The frequency of such evaluations is determined as part of an on-going HOV monitoring and evaluation program. The HOV monitoring framework is a guideline that discusses how often and when the effectiveness of an HOV facility should be determined.  Firstly, prior to the construction of an HOV facility, it is recommended to establish the baseline conditions of a corridor as part of an anticipated HOV monitoring program. Statistically reliable baseline data is necessary to perform meaningful "before-and-after" HOV evaluations. The data required to establish the baseline conditions of a corridor is the same as that summarized in the HOV evaluation framework. Similarly, the procedures for collecting, analyzing, and presenting the baseline data are also outlined in the HOV evaluation framework.  The literature review identified that it is significant to regularly evaluate an HOV facility in its first few years of operation. Once an HOV facility is stable in its operation, the frequency of evaluations is of less importance. Consequently, there is no specific time period warranted in between successive evaluations of an HOV facility. However, it is invariably important to collect data that is statistically reliable for each HOV evaluation.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 42  Otherwise, it is impractical to make conclusions and interpretations from the data collected. It is preferred to wait longer in between successive HOV evaluations than to collect data that is statistically unreliable.  In an HOV monitoring program, the seasonal timing of an HOV facility evaluation is as important as the frequency of evaluation. It is preferred to repeatedly conduct an annual HOV evaluation at the same time of each year.  Chapter 3: HOV Monitoring and Evaluation Framework  Page 43  4. OBJECTIVES AND EVALUATION PROCEDURES  4.1 INTRODUCTION As discussed in Chapter One, seven objectives are evaluated as part of the HOV monitoring and evaluation framework. The objectives are categorized according to their relationship with the HOV network goals as either a primary, supporting, or operational objective. The general evaluation methodology of the HOV evaluation framework was summarized in Chapter Three. The evaluation procedure is repeated to measure the effectiveness of each objective of the framework. The remaining sections of this chapter describe the specific evaluation methodology for each of the primary, supporting, and operational objectives in the following order: • Person Throughput • Travel Time Savings • Travel Time Reliability • Public Support • Compliance • Safety • Impact on GP Lanes  4.2 PERSON THROUGHPUT The main purpose for performing an evaluation on an HOV facility is to measure its effectiveness at achieving the HOV project's goals. As described in Chapter One, the goals of the HOV network are to improve the person throughput and encourage higher occupancy on selected corridors. As a result, the framework's primary evaluation  Chapter 4: Objectives and Evaluation Procedures  Page 44  objective is to measure an HOV facility's effectiveness at attaining these regional HOV goals. Accordingly, the primary objective is referred to as the 'Person Throughput' throughout the framework.  4.2.1 Measures of Effectiveness The 'Person Throughput' objective is evaluated with the following three MOEs: Per-Lane Efficiency, AVO, and HOV Market Share.  PER-LANE EFFICIENCY The efficiency of an arterial or freeway lane is calculated by multiplying the lane's average operating speed by the total number of persons on the lane over a period of time. The efficiency of a lane is expressed in 1,000 person-km/h per hour.  Efficiency = average speed x total number of persons/time 1,000 The per-lane efficiency of a facility is calculated by taking an average of the facility's respective lane efficiencies. For example, the efficiency of an HOV facility is a per-lane weighted average of the GP lane efficiency and the HOV lane efficiency. It is desired that the per-lane efficiency of a facility increases with the introduction of an HOV lane. Increases in this MOE indicate that the person throughput of the overall facility has improved from the pre-HOV lane condition.  AVO The average vehicle occupancy (AVO) of a facility is calculated by dividing the total number of passenger car occupants by the total number of passenger cars.  Chapter 4: Objectives and Evaluation Procedures  Page 45  AVO  = # of passenger car occupants # of passenger cars  The AVO is a measure of the average number of persons on a facility on a per car basis. An increase in this MOE indicates that the average number of persons per car has increased. In other words, an increase in the AVO indicates that people are shifting to a higher occupancy mode, buses excluded.  HOV MARKET SHARE The HOV Market Share is calculated by dividing the total number of persons that make their trip on an HOV (carpool, vanpool, or bus) by the total number of person trips.  HOV Market Share = # of HOV person trips # of total person trips The HOV Market Share is a measure of the number of persons that commute by the means of an HOV, buses included. It is an MOE of the 'Person Throughput' objective because it is directly related to evaluating the effectiveness of the regional goal of encouraging higher occupancy. Increases in the HOV Market Share indicate that a greater percentage of the commuting traffic is travelling in HOVs, buses included.  4.2.2 Data Requirements To determine the 'Person Throughput' objective, vehicle classification and occupancy counts as well as travel time data are required. The vehicle classification and occupancy counts are necessary for the calculation of all three primary objective MOEs.  Chapter 4: Objectives and Evaluation Procedures  Page 46  Conversely, the travel time data is only required for the calculation of the HOV facility's per-lane efficiency.  4.2.3 Data Collection Methodology The following issues concerning the desired data collection methodology for the 'Person Throughput' objective should be considered: • The vehicle classification and occupancy counts are to be performed for the entire peak period. This reflects the fact that the MOE results of an HOV facility may significantly differ in the peak hour and the peak period. It is recommended that this effort be continued until it is determined that the HOV facility is mature. At such time, the pattern of the peak period could be predicted based on its peak hour of operation. • The vehicle classification and occupancy counts are to be conducted at representative locations along an HOV corridor. • The vehicle classification and occupancy data is combined directly with the travel time data in the determination of the per-lane efficiency MOE. Therefore, it is preferred to coordinate the two types of surveys at the same time (i.e., same week of calendar year).  4.2.4 Data Analysis For all three of the 'Person Throughput' MOEs, the peak hour of operation is defined to be the hour that GP traffic experienced their longest travel times along the HOV corridor. Therefore, the analysis of the peak hour is consistent between surveys (see 'Travel Time Savings' objective for further explanation).  Chapter 4: Objectives and Evaluation Procedures  Page 47  PER-LANE EFFICIENCY The per-lane efficiency is calculated for the peak direction of traffic. For the purpose of HOV evaluation, it is not necessary to evaluate this MOE for both the peak period and peak hour of operation. Therefore, it is recommended that only the peak hour per-lane efficiency be calculated. Example: Sample calculation of the peak hour per-lane efficiency (Source of data: Appendix B - Efficiency Data ) TYPE  #OF  PERSON  TIME  SPEED  PER-LANE  OF LANE  LANES  VOLUME  (hrs)  Jkm/h)  EFFICIENCY  GP  2  1505  1.0  30  23  HOV  1  2188  1.0  36  79  HOV FACILITY  41  AVO The AVO is calculated in 15-minute interval samples for each location of a vehicle classification and occupancy survey. The mean, standard deviation, and 95% confidence limits of these 15-minute interval samples is determined at each location for both the peak period and peak hour of operation.  Example: Sample calculation of the mean peak hour AVO, standard deviation, and 95% confidence limits (Source of data: Appendix C -AVO Data ) TIME  AVO  7:30-7:45  1.24  7:45-8:00  1.30  8:00-8:15  1.38  8:15-8:30  1.37  AVO =1.32 Standard Deviation = 0.07 95% Confidence Limits = 1.21 to 1.43  Chapter 4: Objectives and Evaluation Procedures  Page 48  The peak hour and peak period AVO and 95% confidence limits of the overall HOV corridor are calculated similar to the example above and presented as part of the results of the data analysis.  HOV MARKET SHARE Unlike the AVO results, the HOV Market Share results are not presented in 15minute interval samples. They are calculated as absolute numbers for each location of a vehicle classification and occupancy survey for the peak period of HOV operation. Example: Sample calculation of the peak period HOV Market Share (Source of data: Appendix D - HOV Market Share Data) TYPE  VOLUME  % OF TOTAL VOLUME  SOVs HOVs  3868  53 %  2365  32%  BUS  1047  15%  TOTAL  7280  100%  HOV MARKET SHARE  3412  47%  Similar to the AVO, HOV Market Share is also calculated for the overall HOV corridor as part of the data analysis. The peak period corridor MOE is computed from all representative survey locations.  4.2.5 Parallel Routes Analysis To examine the HOV lane's impact on its parallel routes, it is recommended that the AVO and HOV Market Share be determined on these routes. The trends in the AVO of the mainline and its parallel routes should be examined together to develop a better understanding of the HOV facility's 'Person Throughput' effectiveness.  Chapter 4: Objectives and Evaluation Procedures  Page 49  4.2.6 Statistical Reliability As mentioned in Chapter Three, the statistical reliability of the mean AVO results are investigated by quantifying the margins of error associated with the AVO samples of data. The margins of error are presented as percentages of the mean AVO results. Both the mainline and parallel routes (if available) AVO results are examined.  4.2.7 Statistical Significance The difference between the baseline or "before" AVO and any '"after" AVO results are tested for statistical significance. The differences in the AVO results are tested using a two sample Mest. An example of a two sample Mest is presented in Appendix A Statistical Analysis. Minimum "after" AVOs are also calculated using an average AVO standard deviation and sample size. The minimum "after" AVOs are estimates of the lowest AVO that would produce statistically significant differences in the "before" AVOs.  4.3 TRAVEL TIME SAVINGS Travel time savings is evaluated as a supporting objective because it contributes to the attainment of the HOV network goals. A travel time savings benefit for HOVs is the main incentive that encourages single occupant commuters to shift to a higher occupancy mode. The evaluation of the 'Travel Time Savings' objective examines the effectiveness of the HOV lane at providing a relative travel time advantage to the occupants of HOVs.  4.3.1 Measures of Effectiveness The 'Travel Time Savings' objective is evaluated with the travel time difference MOE:  Chapter 4: Objectives and Evaluation Procedures  Page 50  TRAVEL TIME DIFFERENCE The travel time difference is calculated by subtracting the HOV travel time from the general purpose (GP) travel time for a pair of HOV and GP travel time runs that began at the same time. The MOE is expressed in minutes.  Travel Time Difference = GP travel time - HOV travel time It is anticipated that the HOV travel times will be less than the GP travel times. Therefore, travel time difference is the relative measure of the number of minutes an HOV occupant saves as compared to a GP occupant. The larger the travel time difference being provided, the larger the incentive for commuters to shift to a higher occupancy mode.  4.3.2 Data Requirements Travel time survey data on both the HOV and GP lanes is required to determine the effectiveness of the HOV lane at providing a relative travel time savings advantage. The travel time survey should be conducted for the entire peak period of HOV operation on representative weekdays, preferably Tuesday through Thursday.  4.3.3 Data Collection Methodology To calculate the travel time difference, two simultaneous travel time runs should be conducted. One test vehicle should drive in the GP lane(s) and another vehicle in the HOV lane. Both test vehicles should depart at the same time from the beginning of the HOV facility.  Chapter 4: Objectives and Evaluation Procedures  Page 51  4.3.4 Data Analysis The travel time difference is calculated from the HOV and GP travel time survey data. The mean, standard deviation, and 95% confidence limits of the travel time differences should be presented as part of the data analysis. The results should be determined for both the peak period and peak hour of HOV operation.  To be consistent between travel time surveys, the peak hour should be defined to be the hour where GP traffic experienced their longest delays along the corridor. The incentive for GP traffic to take advantage of the HOV travel time savings is most likely to occur during this period of time. If the GP travel times are relatively similar throughout the entire peak period, the hour with the largest traffic volume should be defined as the peak hour.  Example: Sample calculation of the peak hour travel time difference, standard deviation, and 95% confidence limits (Source of data: Appendix E - Travel Time Savings Data) TIME 7:15 7:30 7:45 8:00  GP travel time HOV travel time Travel Time Difference (minutes) (minutes) (minutes) 25.5 28.0 28.6 26.1  23.0 24.1 25.1 22.2  2.5 3.9 3.5 3.9  TRAVEL TIME DIFFERENCE = 3.5 minutes Standard Deviation = 0.7 minutes 95% Confidence Limits = 2.4 to 3.6 minutes  In addition to travel time differences, a distance versus time plot could be used to summarize the travel time survey data. The plot would be beneficial in identifying an HOV corridor's "hot spots" and gaining further insight into the effectiveness and operation of the HOV lane versus the GP lane(s).  Chapter 4: Objectives and Evaluation Procedures  Page 52  4.3.5 Arterial vs. Freeway Analysis For either an arterial or freeway HOV facility with distinct operational segments, the travel time differences should be calculated for each segment as well as the entire corridor. The HOV lane may be providing a range of travel time savings for different segments. Therefore, the effectiveness and operation of each segment should be examined separately.  4.3.6 Statistical Reliability As mentioned in Chapter Three, the statistical reliability of both the HOV and GP travel time survey data should be investigated. The samples of data should be tested at the minimum specified error of ± 5 km/h for "before-and-after" studies as recommended in the 'ITE Transportation and Traffic Engineering Handbook' (ITE, 1982).  4.3.7 Statistical Significance The mean travel time differences between the HOV and GP lanes are tested for statistical significance using a paired sample f-test. An example of a paired sample Mest is found in Appendix A - Statistical Analysis.  4.4 TRAVEL TIME RELIABILITY An HOV facility should provide HOVs with a more reliable trip relative to other commuters in the GP lane. By providing a more reliable trip, the HOV facility offers another motive for commuters to switch to a higher occupancy mode. The 'Travel Time Reliability' objective examines the effectiveness of an HOV facility at providing a more  Chapter 4: Objectives and Evaluation Procedures  Page 53  reliable trip to HOV drivers and their occupants. It is also evaluated as a supporting objective because providing HOVs with a more reliable trip contributes to the attainment of the HOV network goals.  4.4.1 Measures Of Effectiveness The 'Travel Time Reliability' objective is evaluated by quantifying and comparing the standard deviation of the travel speeds associated both the HOV and GP lanes. It is expressed in kilometers per hour. The standard deviation of the travel speed of a given lane is a measure of the travel time variability that a commuter will experience travelling on that lane over a period of time. The larger the standard deviation, the less reliable the trip on that lane. It is anticipated that the users of the HOV lane will experience a more reliable trip over time as compared to the users of the GP lanes. In other words, the travel speed standard deviation of vehicles travelling on the HOV lane should be less than the travel speed standard deviation of vehicles travelling on the adjacent GP lanes.  4.4.2 Data Requirements To evaluate the 'Travel Time Reliability' objective, a travel time survey is required in addition to the one previously described for the 'Travel Time Savings' objective. The additional travel time survey should be conducted on both the HOV and GP lanes over an extended period of time on consecutive weekdays.  4.4.3 Data Collection Methodology The following issues concerning the desired data collection methodology for the 'Travel Time Reliability' objective should be considered: • The travel time survey should be conducted daily, Monday through Friday.  Chapter 4: Objectives and Evaluation Procedures  Page 54  • Departure times within the peak period of HOV operation should be selected ahead of time. More than one departure time may be selected. However, at least one departure time should be selected within the peak hour. • For each selected departure time, a pair of HOV and GP test vehicles should depart together from the beginning of the HOV facility.  4.4.4 Data Analysis The HOV and GP travel speed standard deviations are calculated from the travel time survey data. As mentioned in the literature review, travel speeds provide better estimates of travel time reliability than travel time distributions (Turner et al., 1994). Therefore, the travel time survey data is firstly converted into travel speeds. Secondly, the travel speed standard deviations of both the HOV and GP lanes is calculated for each departure time of the survey. The standard deviations should be presented as absolute values (in km/h) as well as percentages of the mean travel speed. Example: Sample calculation of the peak hour travel speed standard deviation of a lane (Source of data: Appendix F - Travel Time Reliability Data DATE  SPEED (km/h)  DATE  SPEED (km/h)  October 16 October 17 October 20 October 21 October 22 October 23  42 43 42 39 43 35  October 24 October 27 October 28 October 29 October 30 October 31  43 45 39 36 37 39  TRAVEL SPEED = 40.2 km/h Standard Deviation (km/h)= 3.3 km/h Standard Deviation (% of the mean) = 8.2 %  Chapter 4: Objectives and Evaluation Procedures  Page 55  For each departure time, a comparison of the HOV and GP travel speed standard deviations should be made. Once again, the peak hour of operation is defined to be the hour that GP traffic experienced their longest travel times along the corridor. In addition, the travel time reliability of both the HOV and GP lanes can be analyzed through the use of a frequency versus travel speed plot. The plots may be of use to graphically assist in demonstrating the smaller variability (or range) of travel speeds experienced on the HOV lane when compared to the adjacent GP lanes.  4.4.5 Arterial vs. Freeway Analysis No additional insight is gained in re-calculating the travel speed standard deviations for the distinct operational segments of either an arterial of freeway HOV corridor because it is the overall travel time reliability that is being evaluated.  4.4.6 Statistical Reliability Similar to the 'Travel Time Savings' objective, the statistical reliability of the travel time survey data should be tested at the minimum specified permitted error of ± 5km/h as recommended in the 'ITE Transportation and Traffic Engineering Handbook' (ITE, 1982).  4.4.7 Statistical Significance The difference between the travel speed standard deviations of the HOV and GP lanes is tested for statistical significance using an F-test. The difference is analyzed at the 95th percentile. An example of an F-test is found in Appendix A - Statistical Analysis.  Chapter 4: Objectives and Evaluation Procedures  Page 56  4.5 PUBLIC SUPPORT The public's perception of the operation and effectiveness of HOV lanes is important to the success of an HOV network. The level of satisfaction and attitude of commuters can positively or negatively contribute to the attainment of the HOV network goals. Therefore, the public's support (or lack of) is evaluated in the framework as a supporting objective.  A public opinion survey is used to evaluate the attitudes and travel preferences of all types of commuters on the HOV facilities including bus riders, carpoolers, vanpoolers, and those who drive alone. The results of the public opinion survey can provide insight into the target areas and level of effort necessary to improve marketing and public education. Therefore, the purpose of the survey is to establish commuter perceptions on a variety of HOV topics such as: travel time savings, travel time reliability, HOV enforcement, overall HOV network issues, and carpool incentives.  4.5.1 Measures of Effectiveness The 'Public Support' objective is evaluated with the support for HOV lane measure of effectiveness. SUPPORT FOR HOV LANE The support for HOV lane MOE is calculated by dividing the number of people who support the use of HOV lanes by the total number of people surveyed. The MOE is expressed as a percentage.  Support for HOV Lane (%) = number of people who support HOV lanes number of people surveyed  Chapter 4: Objectives and Evaluation Procedures  Page 57  It is most likely that the responses of many survey questions could be used to determine the percent support for HOV lanes. For example, commuter's answers to the following three survey questions may all be used to determine this MOE: •  Are the HOV lanes a good idea ?  •  Are the HOV lanes being well-used ?  •  Are the HOV lanes convenient to use ?  4.5.2 Data Requirements To evaluate the 'Public Support' objective, either a mail-out or telephone survey is required. However, it is recommended that the license plates of actual users of the HOV facility be recorded and matched to addresses from the driver's license database and a mail-out survey be used.  4.5.3 Data Collection Methodology When conducting a public opinion survey, it is important that a statistically significant sample of HOV users be obtained. The sample size required is highly dependent on the anticipated response rate of those being surveyed. Therefore, the survey data collection methodology should be established after selecting the type of public opinion survey to be conducted.  4.5.4 Data Analysis The analysis of the public opinion survey data is dependent on the content of the survey and the types of questions asked of the respondents. A sample mail-out public opinion survey is, therefore, presented as part of the evaluation framework. The survey is based  Chapter 4: Objectives and Evaluation Procedures  Page 58  on the HOV public opinion surveys of other jurisdictions and is tailored to the needs of the South Coast Region HOV program. It is comprised of two sections: 'Part 1: Your Commute Trip' and 'Part 2: Your Opinion' (see Figures 4.1 and 4.2). The survey format was reviewed by the Region's Communications section as well as the BCTFA. In addition to the two sections, it is necessary to include an introductory letter explaining the purpose and scope of the survey as well as a section for the respondents' written comments.  4.5.5 Statistical Reliability The statistical reliability of the public opinion survey data should quantified as the margin of error associated with the sample size obtained. The margin of error should be expressed as a percentage and obtained for the 95% confidence level.  4.6 COMPLIANCE A high degree of compliance inherently protects HOV operation and the attainment of the HOV network goals. Therefore, compliance is evaluated as an operational objective. HOV facilities with a low degree of compliance indicate a higher than desirable proportion of violators, thereby compromising the effectiveness of the HOV lane. It is important that the degree of compliance remain within an acceptable range to protect the travel time savings and reliability benefits for HOV commuters.  4.6.1 Measures of Effectiveness The 'Compliance' objective is evaluated with the compliance rate MOE.  Chapter 4: Objectives and Evaluation Procedures  Page 59  PART 1: YOUR COMMUTE TRIP  1. How many times per week (Monday through Friday) do you usually commute on either the Barnet Highway or Hastings St. (or both) ? O once a week or less  02 to 4 times per week  D Monday through Friday  2. What is your P R E S E N T usual mode of travel between 6:00-8:30 AM and 3:30-6:00 PM ? (pick only one please) O drive alone D bus O vanpool  O carpool with one other person D carpool with two or more people O other  3. Prior to the introduction of the Barnet/Hastings HOV lane in September 1996, what was your usual mode of travel? (pick only one please) O drive alone a bus D vanpool  O carpool with one other person O carpool with two or more people O other  4. Prior to the introduction of the Barnet/Hastings HOV lane in September 1996, did you commute on a different route ? DYES  D NO  5. Have you regularly made use of the Barnet/Hastings HOV lane since it has been introduced ? D YES...  Don St.Johns and/or Clarke St. Don Hastings Street  D on the Barnet Highway  DNO  6. Have you ever NOT used the Barnet/Hastings HOV lane even though you had enough people to qualify for it ? DYES... DNO  if yes, why?  7. Have you formed a new carpool in order to be able to use the newly constructed HOV lane ? D YES DNO ...  if no, why not ?  8. How many minutes, if any, do you believe you save by using the HOV lane instead of the regular lanes ? minutes in the morning  minutes in the evening  Figure 4-1: Sample Public Opinion Survey - Part 1  Chapter 4: Objectives and Evaluation Procedures  Page 60  P A R T 2: Y O U R OPINION  Please indicate whether you agree, disagree or are neutral about the following statements. PERCEPTION OF THE BARNET/HASTINGS HOV LANE  The HOV lane is being well-used  D agree  O neutral  O  disagree  The HOV lane is faster than the regular lanes The HOV lane is more reliable than the regular lanes The HOV lane has helped ajl commuters to save time  O agree D agree D agree  D neutral D neutral O neutral  O disagree O disagree O disagree  THE OPERATION OF THE BARNET/HASTINGS HOV LANE  The HOV lane is convenient to use The HOV lane is safe There are too many buses in the HOV lane Many cars weave in and out of the HOV lane  O O O O  agree agree agree agree  O D D D  neutral neutral neutral neutral  D D D D  disagree disagree disagree disagree  ENFORCEMENT ON THE BARNET/HASTINGS HOV LANE  HOV violators are seldom caught There are a lot of violators in the HOV lane Cameras should be used to help enforce the HOV lane There should be higher penalties for HOV violators  D D D O  agree agree agree agree  D D O O  neutral neutral neutral neutral  O O D O  disagree disagree disagree disagree  n D O 0 D O D D  agree agree agree agree agree agree agree agree  O neutral D neutral O neutral 0 neutral D neutral D neutral D neutral 0 neutral  O D D D O D O D  disagree disagree disagree disagree disagree disagree disagree disagree  D 0 O O  agree agree agree agree  0 D n D  D n O O  disagree disagree disagree disagree  OVERALL HOV NETWORK ISSUES  HOV HOV HOV HOV HOV HOV HOV HOV  lanes are a good idea lanes need to be more widespread lanes are unfair to those who need to drive alone lanes should be eliminated lanes should be in operation 24 hours a day lanes should all be 2 or more people occupancy lanes should all be 3 or more people occupancy lanes are good investments of tax payers money  CARPOOL INCENTIVES  1 would be encouraged or motivated to carpool if... ... carpooling was easy to arrange ... HOV lanes saved time ... there were more HOV lanes ... HOV lanes were more reliable than the regular lanes Nothing would motivate me to carpool  D agree  neutral neutral neutral neutral  O neutral  D disagree  Fiaure 4-2: Sample Public Opinion Survey - Part 2  Chapter 4: Objectives and Evaluation Procedures  Page 61  COMPLIANCE RATE The compliance rate is calculated by dividing the total number of cars in the HOV lane that meet the occupancy requirement by the total number of cars in the HOV lane. The MOE is expressed as a percentage.  Compliance Rate (%) = number of cars that meet the occupancy requirement total number of cars in the HOV lane The desired rate of compliance on an HOV facility was determined by MoTH to be 85%. However, MoTH recognizes that achieving this target rate is dependent on the resource allocated by local enforcement agencies for conducting HOV enforcement which are presently limited.  4.6.2 Data Requirements Only vehicle classification and occupancy counts for the HOV lane are required to determine the 'Compliance ' objective. The vehicle classification and occupancy counts should be conducted for the entire peak period of HOV operation on randomly selected weekdays.  4.6.3 Data Collection Methodology The vehicle classification and occupancy counts are to be conducted at representative locations along an HOV corridor. To maximize the accuracy of the data collected, it is preferred to conduct compliance counts at locations with good visibility and low vehicular speeds.  Chapter 4: Objectives and Evaluation Procedures  Page 62  It should be made clear that observing the compliance of high speed vehicles in poor daylight conditions is not an easy task. For example, four out of ten cars in the HOV lane identified as violators by a Port Moody police officer turned out to be legitimate HOVs when pulled over (as observed in an actual police ride-along). Children in the back seat, passengers reclining in their seat, and tinted windows are some of the examples that may be perceived as HOV violators when performing compliance counts.  4.6.4 Data Analysis The compliance rates are calculated from the vehicle classification and occupancy counts. The compliance rate for each location should be presented as well as a corridor average. When calculating the corridor average, only locations that are representative of the typical compliance rate observed are to be included. For example, the compliance rate is sometimes obtained at certain locations for operational purposes only to assist in identifying problem areas. These compliance rates should not be included in the calculation of the corridor average.  Example: Sample calculation of the peak period compliance rate of a corridor (Source of data: Appendix G - Compliance Data ) Location  # of cars that meet the occupancy requirement  Williams Clark Reed Bayview Willingdon Esmond  495 724 820 807 851 873  Total # of cars in Compliance the HOV lane Rate 563 829 953 927 958 1015  88.0 % 87.3 % 86.0 % 87.0 % 88.8 % 86.0%  COMPLIANCE RATE = 87.2 %  Chapter 4: Objectives and Evaluation Procedures  Page 63  In addition to compliance rates, the number of HOV violation tickets issued could be obtained. If records are kept on the level of enforcement, these should be correlated to either the compliance rates or the number of HOV violation tickets being issued (RSMI , 3  1997).  4.6.5 Arterial vs. Freeway Analysis For either an arterial or freeway HOV facility with distinct operational segments, the compliance rates should also be calculated for each segment as well as the entire corridor. The compliance rate should be examined separately to identify the segments where special HOV enforcement should be allocated.  4.7 SAFETY 'Safety' is an operational objective which evaluates an HOV facility's "before-and-after" level of safety. Intersections and road sections are examined separately to determine if the implementation of the HOV facility has adversely affected the level of safety being provided on the corridor.  4.7.1 Measures of Effectiveness The 'Safety' objective is evaluated with the accident rate measure of effectiveness. The accident rates for intersections and sections are quantified and compared separately. ACCIDENT RATE The accident rate of an intersection is calculated for a given period of time by dividing the number of accidents at the intersection by the number of vehicles entering the intersection from all approaches. The MOE is expressed in accidents  Chapter 4: Objectives and Evaluation Procedures  Page 64  per million-entering-vehicles (acc/MEV). Similarly, the accident rate of a road section is calculated by dividing the number of accidents by the number of vehiclekilometers, expressed as accident per million-vehicle-kilometers (acc/MVK).  Intersection Accident Rate = N x 1.000.000 AADT x t x 365 Section Accident Rate = N x 1.000,000 L x AADT x t x 365 where  N L AADT t  = number of accidents observed during a period t = length of segment (km) = average annual daily traffic volume (veh/day) = observation period (years)  It is anticipated that the accident rates of a corridor will not increase with the implementation of an HOV facility. Increases in the accident rates indicate that the safety of the corridor has been somewhat compromised as a result of the HOV lane.  4.7.2 Data Requirements "Before-and-after" accident data on both the HOV and GP lanes is required to determine the effectiveness of the HOV lane at providing safe operations. Usually, at least three years of "before" accident data and at least one year of "after" accident data should be obtained for the safety analysis. In addition, "before-and-after" traffic volume data is also required.  4.7.3 Data Collection Methodology The following issues concerning the desired data collection methodology for the 'Safety' objective should be considered:  Chapter 4: Objectives and Evaluation Procedures  Page 65  • AADT data should be obtained from a minimum of one representative location along the HOV facility. • Any changes in the accident reporting procedure of the local enforcement agencies should be documented to assist in interpreting the results of the "before-and-after" safety analysis. • If possible, the lane in which an accident occurs in (either HOV or GP) should be reported as part of the accident reporting and compiling procedure.  4.7.4 Data Analysis The "before-and-after" intersection and section accident rates of the HOV facility should be calculated from the accident and traffic volume data. Example: Sample calculation of the "before" intersection accident rate at a location (Source of data: Appendix H - Safety Data) N AADT t  = 30 accidents over 4 years = 33,400 (veh/day) = 4 years  Intersection Accident Rate = 30 x 1,000,000 33,4000x4x365 = 0.62acc/MEV  In addition, if the accident data is categorized under the type of lane in which the accident occurred in, the "after" accident rates of both the HOV and GP lane should also be quantified and compared with one another. It is anticipated that the "after" HOV accident rates would be equal to or less than the "after" accident rates in the GP lane.  Chapter 4: Objectives and Evaluation Procedures  Page 66  4.7.5 Statistical Significance A simple comparison between the mean "before-and-after" intersection and section accident rates should be performed. This simple "before-and-after" comparison method assumes that the mean accident rates would not have changed if the HOV lane had not been implemented on the facility. The differences between the mean "before-and-after" accident rates are tested for statistical significance using a paired sample Mest. An example of how to perform a paired sample Mest is found in Appendix A -Statistical Analysis. It should be made clear that a simple "before-and-after" comparison method does not account for the inclusion of unrelated effects nor the regression to the mean (RTM) effect. As previously mentioned in the literature review, it is suggested that the "after" changes in accident rates common to the region also be determined (see for example: Turnbull et al., 1991; Sayed , 1996). 1  To account for unrelated effects, it is suggested that the accident rates of a set of "control group" locations similar to the HOV facility be observed. To account for the regression to the mean effect, it is suggested an Empirical Bayes approach be used. The suggested approach differs from most traditional statistical methods of examining safety records in that it requires the use of a reference group. The reference group represents a similar corridor as that of the HOV facility that is being examined and eliminates the bias in traditional safety improvement evaluations. The Empirical Bayes approach provides "a formal mechanism for combining information drawn from different  sources in a single analysis" (Sayed , 1996). The Empirical Bayes approach is presently 1  being used by the Insurance Corporation of British Columbia (ICBC) in their safety analyses.  Chapter 4: Objectives and Evaluation Procedures  Page 67  4.8 IMPACT ON GP LANES The 'Impact on GP Lanes' objective evaluates the operational impact of the HOV lane on its adjacent general purpose (GP) lanes. It is anticipated that the implementation of an HOV facility will not adversely affect the operation of the GP lanes. However, it should be noted that with other network factors such as latent demand and regional growth, it is probable that the level-of-service being provided to GP users will decrease over time even with the addition of an HOV lane. Therefore, the effectiveness of the 'Impact on GP Lanes' objective is only evaluated on a short term basis.  4.8.1 Measures of Effectiveness The "Impact on GP Lanes' objective is evaluated with the GP travel speed MOE. GP TRAVEL SPEED The GP travel speed MOE is determined by quantifying and comparing the "before-and-after" GP travel speeds. The MOE is expressed in kilometers per hour. GP Travel Speed = "before" GP travel speed vs. "after" GP travel speed  It is anticipated that the "before" GP travel speed will be less than the "after" GP travel speed, at least in the short term. The HOV facility is considered to be effective at not adversely affecting the operation of the GP lanes as long as the "before" vs. "after" GP travel speed difference remains positive. The larger the difference, the more effective the HOV facility has been at attaining this objective.  Chapter 4: Objectives and Evaluation Procedures  Page 68  4.8.2 Data Requirements and Data Collection Methodology "Before-and-after" travel time survey data on the GP lanes is required to determine the effectiveness of the 'Impact on GP Lanes' objective. Similar to the 'Travel Time Savings' objective, "before-and-after" travel time surveys should be conducted for the entire peak period of HOV operation on representative weekdays, preferably Tuesday through Thursday. The travel time data collected on the GP lanes for the 'Travel Time Savings' objective can also be used as the "after" travel time data required for the evaluation of this objective.  4.8.3 Data Analysis Mean GP travel speeds are calculated from the GP travel time survey data. Similar to the per-lane efficiency, it is not necessary to evaluate this MOE for both the peak period and peak hour of HOV operation. Therefore, it is recommended that only the peak hour GP travel speeds be calculated. As previously mentioned, the peak hour should be defined to be the hour where GP traffic experienced their longest delays along the corridor.  Example: Sample calculation of the mean peak hour GP travel speed (Source of data: Appendix I - GP Travel Speed Data) TIME 7:15 7:30 7:45 8:00  OCTOBER 22, 1996 42 38 38 41  OCTOBER 24, 1996  km/h km/h km/h km/h  41 38 41 40  km/h km/h km/h km/h  GP TRAVEL SPEED = 40 km/h  Chapter 4: Objectives and Evaluation Procedures  Page 69  In addition, if vehicle classification and occupancy data is also available, a more insightful analysis of the GP travel speeds should be performed. The vehicular volumes in the GP lanes should be quantified and compared to the GP travel speeds. It is anticipated that the "after" GP travel speeds will decrease over time as the "after" GP vehicular volumes increase. Only the volumes at locations where the HOV lane has been added to an existing facility should be compared to on a "before-and-after" basis. In other words, the "before-and-after" GP volumes should not be compared to locations where a GP lane has been converted into an HOV lane.  Example: Sample comparison of the "before-and-after" peak hour GP travel speeds and GP vehicular volumes at a location (Source of data: Appendix I - GP Travel Speed Data) TIME Before After  DATE  GP TRAVEL SPEED (km/h)  GP VEHICULAR VOLUME  18 45 42  2060 1898 2063  March, 1996 October, 1996 October, 1997  4.8.4 Arterial vs. Freeway Analysis For either an arterial or freeway HOV facility with distinct operational segments, the GP travel speeds should be calculated for each segment as well as the entire corridor. The HOV  lane may be providing a range of GP travel speeds for the different segments.  Therefore, the effectiveness of the 'Impact on GP Lanes' objective should be examined separately for each segment.  Chapter 4: Objectives and Evaluation Procedures  Page 70  5. MULTI-CRITERIA EVALUATION  5.1 INTRODUCTION The multi-criteria evaluation is a procedure for evaluating if the cost-effectiveness of the HOV facility. During these times of limited available funds, it is important that the limited financial resources for various transportation improvements are efficiently utilized (Sayed ,1996). Therefore, an economic analysis should be performed to examine the 1  cost-effectiveness of implementing an HOV facility.  In a benefit-cost analysis, the present value of the benefits and costs of a project are determined and compared. It is recommended that only the quantifiable criteria of an HOV facility be used as inputs into the benefit-cost analysis. However, intangible criteria such as travel time reliability that cannot be readily quantified should still be considered in the economic justification of an HOV facility. For example, an HOV project may be deemed to be more desirable if it links separate HOV corridors into a network. The multi-criteria evaluation procedure of an HOV facility and sensitivity analysis are presented in this chapter.  5.1.1 The Multi-Criteria Evaluation Procedure The first step of the multi-criteria evaluation is to identify the quantifiable benefits and costs of the HOV facility. For example, travel time savings is an example of a quantifiable benefit. The travel time saved can be converted into a dollar value. Secondly, the parameters necessary to calculate the quantifiable benefits and costs are identified. For example, travel time differences, person-volumes, vehicle classification  Chapter 5: Multi-Criteria Evaluation  Page 71  counts, average wage rates, and the value of travel time are the parameters required to calculate the travel time savings benefit. In addition, other parameters such as the discount rate and the expected number of years of the HOV project also need to be identified. For each parameter identified, a most likely, worst case, and best case value is determined, where applicable. For example, discount rates are not fixed but variable. Presently, a discount rate of 8% is being used by MoTH to economically evaluate its roadway projects. However, other Canadian roadway jurisdictions are using discount rates ranging between 6% and 12% (MoTH, 1993). Therefore, the most likely discount rate for evaluating the cost-effectiveness of HOV facilities is determined to be 8%. The worst case and best case discount rates are determined to be 12% and 6% respectively.  The cost-effectiveness of the HOV facility is usually evaluated with either the benefit/cost (B/C) ratio or the net present value (NPV) MOEs. In quantifying these MOEs, the most likely values of the parameters are assumed. Finally, a sensitivity analysis is performed by re-calculating the cost-effectiveness MOEs for the worst case and best case values of the parameters. A summary of the steps of the multi-criteria evaluation procedure is presented below: • Identify benefits and costs • Select quantifiable benefits and costs • Determine the types of parameters for each benefit and cost • Select the most likely, worst case, and best case values for each parameter • Determine the cost-effectiveness • Perform sensitivity analysis  Chapter 5: Multi-Criteria Evaluation  Page 72  5.2 BENEFITS AND COSTS Travel time savings is the main quantifiable benefit of implementing an HOV facility. Other quantifiable benefits of an HOV facility include: vehicle operating cost savings, air quality savings, and safety savings. The main quantifiable cost of an HOV facility is its construction. Other examples of quantifiable costs include: maintenance, enforcement, and marketing. Table 5.1 summarizes the quantifiable benefits and costs of the multicriteria evaluation.  Table 5-1: Multiple-Criteria Evaluation Benefits and Costs BENEFITS Travel Time Savings Vehicle Operating Cost Savings Air Quality Savings Safety Savings  COSTS Construction Costs Maintenance Costs Enforcement Costs Marketing Costs  5.2.1 Travel Time Savings The value of travel time savings experienced by HOV, GP, and bus users is calculated based on the following parameters: VARIABLE PARAMETERS  FIXED PARAMETERS  • travel time differences • value of travel time  • average wage rate • person-volumes • vehicle classification counts  The travel time differences between the "add an HOV lane" alternative and the "do nothing" alternative for HOV, GP, and bus users are multiplied by their respective person-volumes to determine the total amount of travel time saved. The mean travel time differences represent the most likely values. The upper and lower bounds (95%  Chapter 5: Multi-Criteria Evaluation  Page 73  confidence limits) of the travel time differences represent the best and worst case values respectively.  The travel time saved is quantified in terms of number of wage hours saved based on the following "value of travel time" guidelines (Waters, 1992). For example, one hour of travel time saved for an automobile driver is equal to 0.5 wage hours saved: • Automobile driver's value of travel time  =  50% of wage rate  • Automobile occupant's value of travel time =  35% of wage rate  • Bus occupant's value of travel time  =  35% of wage rate  • Commercial driver's value of travel time  =  120% of wage rate  The total number of wage hours saved is converted into travel time savings by multiplying the number of wage hours by the average wage rate. The most likely value of the wage rate for the Greater Vancouver area is $18 per hour (MoTH, 1993). It is recommended that the value of travel time range plus 30 or minus 20 percentage points for a sensitivity analysis (Waters, 1992). Therefore, the best case and worst case values are taken to be $23 per hour and $14 per hour, respectively.  5.2.2 Vehicle Operating Cost Savings and Air Quality Savings The annual value of vehicle operating cost savings and air quality savings are based on the following parameters: VARIABLE PARAMETERS  FIXED PARAMETERS  • unit costs • AVO  • person-volumes • length of HOV facility  Chapter 5: Multi-Criteria Evaluation  Page 74  Quantifying the value of vehicle operating cost savings and air quality savings depends on calculating the number of vehicle-kilometers saved due to the implementation of the HOV facility. Firstly, the number of vehicles required to move the same person-volume is determined with both the "before" AVO and the "after" AVO. The mean AVO is used to calculate the most likely values and the AVO's 95% confidence limits are used to calculate worst and best case values.  For example, 1000 vehicles are required to move 1200 persons with a mean "before" AVO of 1.20. However, only 960 vehicles are required to move the same number of people with a mean "after" AVO of 1.25. Due to an increase in the AVO, 40 less vehicles are required to move the same number of people.  The difference in the number of vehicles required to move the same person-volume is multiplied by the length of the facility to determine the number of vehicle-kilometers saved. The value of vehicle operating cost savings and air quality savings are calculated by multiplying the number of vehicle-kilometers saved by a unit cost ($/km). The range of vehicle operating unit costs ($/km) were selected as follows: worst case, 0.08; most likely, 0.09; and best case, 0.12 (CAA, 1997). Similarly, the range of air pollutant unit costs ($/km) used to calculate the air quality savings are as follows: worst case, 0.02; most likely, 0.04; and best case, 0.08 (Bein, 1994).  5.2.3 Safety Savings The annual value of safety savings is based on the following parameters:  Chapter 5: Multi-Criteria Evaluation  Page 75  VARIABLE PARAMETERS • accident frequency • accident costs Safety savings are represented by the estimated reduction in the number and/or severity of accidents following the implementation of the HOV facility. The reduction in accidents between the "before-and-after" accident frequencies are multiplied by the respective accident cost per accident type (fatality, injury, or PDO) to determine the safety savings.  There are several methods to calculate the reduction in accidents. The easiest method is the simple "before-and-after" comparison which assumes that the frequency of accidents would have remained the same if the HOV lane had not been implemented. However, there are several deficiencies with the simple "before-and-after" comparison method (Sayed ,1996): 1  • the inclusion of unrelated effects • the normal distribution assumption • the regression to the mean (RTM) effect • the traffic exposure effect  A simple "before-and-after" comparison of accident rates addresses the traffic exposure effect. However, to account for each deficiency of the simple "before-and-after" comparison method, an Empirical Bayes approach is suggested. The Empirical Bayes approach was previously discussed in the evaluation procedure of the 'Safety' objective (see section 4.7).  Chapter 5: Multi-Criteria Evaluation  Page 76  There is little agreement in the literature regarding the value assigned to the different accident types. The disagreement on what constitutes the direct and indirect accident costs to society has led to several accident cost models. MoTH's Highway Safety Branch suggests that the discount future earnings (DFE) model represents "the most reasonable and acceptable value for costing exercises" (MoTH, 1995). The accident  costs of the DFE model are the most likely values of the accident costs. ICBC's real costs are used as the lower bound values and the willingness to pay (WTP) model accident costs are used as the upper bound values. Table 5.2 summarizes the range of accident costs for the sensitivity analysis.  Table 5-2: Accident Costs ACCIDENT TYPE  ICBC (MoTH,199S)  FATAL INJURY P.D.O. NOTES:  5.2.4  $ 100,000 $ 17,500 $2,300  DFE '•  <'(MoTH,1995)-  WTP ,  $ 1,222,000 $29,500 $5,000*  (Sayecf ,1996)  $2,900,000 $ 100,000 $ 6,000  - all accident costs are reported in Canadian dollars (CDN$) - DFE P.D.O. accident cost modified from $9,000 to be in between upper and lower bounds  Costs  The costs of the HOV  facility are dependent on the following parameters:  VARIABLE PARAMETERS  FIXED PARAMETERS  • maintenance unit cost • enforcement costs • marketing costs  • construction costs • length of HOV facility  The total construction cost of the HOV facility includes capital costs, project management costs, design costs, and right-of-way acquisition costs. The annual maintenance cost is calculated by multiplying the maintenance unit cost ($/lane-km per  Chapter 5: Multi-Criteria Evaluation  Page 77  year) by the number of lane-kilometers of the HOV facility. The most likely maintenance unit cost should be based on MoTH's most recent maintenance services contract. Over the life of the project, the maintenance, enforcement, and marketing costs will fluctuate from year to year. The lowest reported costs should be used as the lower bound values and the highest reported costs as the upper bound values.  5.3 COST- EFFECTIVENESS To determine the cost-effectiveness, either the benefit/cost (B/C) ratio or the net present value (NPV) is usually used. Both MOEs provide a good measure of the costeffectiveness of an HOV facility and inherently have their own advantages and disadvantages. However, it was identified in the literature review that the B/C ratio is the most frequently used cost-effectiveness MOE.  The B/C ratio is the ratio between the present value of the benefits and the present value of the costs. An HOV project that has a ratio B/C ratio greater than one is considered to be cost-effective. Benefit/Cost Ratio = Z benefits / Xcosts  The NPV of an HOV project is calculated by subtracting the present value of the costs from the present value of the benefits. The present value of the benefits and costs should be based on the most likely values of the parameters.  NPV of HOV facility = I benefits - Icosts  Chapter 5: Multi-Criteria Evaluation  Page 78  It is recommended that the B/C ratio or NPV of the HOV facility be calculated for the base year in which the HOV lane was implemented. In other words, all of the costs accrued prior to the implementation of the HOV lane and all of the benefits and costs anticipated in the future are discounted to the base year. The present value of a known future amount occurring N years from the base year is calculated with the following formula:  PV =  d + if where  PV F i N  : present value of amount • future value of amount • discount rate : number of years.  Similarly, the present value of a known past amount from N years ago is calculated with the following formula: PV= where  5.3.1  PV P / N  Px(1+if = present value of amount = past value of amount = discount rate = number of years  Sensitivity A n a l y s i s  A sensitivity analysis of the cost-effectiveness MOEs is performed as a decision making tool to indicate the sensitivity of the results to the different parameters and the uncertainty associated with the most likely values. The desirability and relative priority of a program of HOV projects is dependent on how changes in the parameters affects the outcome of the cost-effectiveness MOEs. For example, if an HOV project's B/C ratio is calculated to be 2.0 for the most likely values of the parameters, the HOV project is  Chapter 5: Multi-Criteria Evaluation  Page 79  considered to be cost-effective. However, if the B/C ratios of the lower and upper bounds of the parameters are calculated to be 0.5 and 5.0 respectively, the costeffectiveness of the HOV project is sensitive to the different parameters of the benefits and costs. The sensitivity analysis is performed by firstly calculating the costeffectiveness MOEs with the most likely values of the parameters. Secondly, the MOEs are re-calculated using the lower bound (or worst case) values as well as the upper bound (best case) values. A comparison of all three scenarios is performed and the sensitivity of the results is discussed.  Chapter 5: Multi-Criteria Evaluation  Page 80  6. APPLICATION OF THE FRAMEWORK  6.1 INTRODUCTION An application of the evaluation framework was performed to evaluate the effectiveness of the HOV facility on the Barnet/Hastings corridor. The purpose of performing the application was to clearly demonstrate the suggested procedure of the evaluation framework. The Barnet/Hastings HOV corridor was selected for the application because it is the only regional HOV facility currently in operation. In other words, it is the only local HOV facility where "before-and-after" data is available.  6.1.1 Data Collection Summary For the purpose of the application, "before-and-after" vehicle classification and occupancy counts, travel time runs, and accident records were retrieved. In addition, "after" compliance data was also obtained as well as "before-and-after" 24-hour traffic counts on a parallel route. No public support data relating to the operation of the Barnet/Hastings HOV corridor was available at the time of the application. The "beforeand-after" data was obtained through numerous agencies. The sources and dates of the Barnet/Hastings HOV corridor data collection surveys and records are summarized in Table 6.1.  6.1.2 Description of the HOV Facility The HOV facility on the Barnet/Hastings corridor, otherwise known as Route 7A, is the first HOV corridor constructed by British Columbia's Ministry of Transportation and Highways (see Figure 6.1). The curb-lane of the corridor operates as an HOV lane in the peak direction of travel. The Barnet/Hastings HOV lane was constructed along certain  Chapter 6: Application of the Framework  Page 81  sections by simply upgrading the curb-lane of the facility. In other sections, the facility was widened to accommodate an additional HOV lane.  The three distinct arterial and freeway sections of the Barnet/Hastings HOV corridor were previously categorized by MoTH as Hastings Street, Barnet Highway, and St. Johns Street (MoTH , 1997). The westbound HOV lane was constructed in all three 1  sections. Therefore, the length of the westbound HOV lane is 18 kilometers, the approximate length of the HOV facility. However, no HOV lane was constructed eastbound along the St. Johns Street section. Consequently, the length of the eastbound HOV lane is only 15 kilometers. Curtis Street (or Parker Street) was identified as a parallel route along the Hastings Street section of the HOV facility.  Table 6-1: Summary of the Data Collection Surveys and Records TYPE OF DATA  SOURCE OF DATA  "BEFORE" DATES  Vehicle Classification & Occupancy  Transtech Data Services  Mar.'96 Mar.'96  Travel Time  Transtech Data Services The University of British Columbia  Compliance  "AFTER" DATES Oct.'96 Mar.'97 Oct.'97 Oct.'96 Oct.'97 Oct.'96 Mar.'97 Oct.'97 SepU96 - Sept'97  Transtech Data Services ICBC  Accident  MoTH City of Burnaby City of Vancouver  Jan.'92 - Dec.'95 Jun.'92 - Dec.'95 Jan.'92 - Dec.'95  Sept.'96 - Jun.'97 Sept.'96 - Oct.'97 Sept.'96 - Dec'97  Parallel Routes Data  City of Burnaby  Feb.'96 - Mar.'96  Nov.'96  Chapter 6: Application of the Framework  Page 82  The HOV lane is in operation part-time (Monday to Friday) during the peak periods with an occupancy requirement of 2 or more people. The peak period in the morning is westbound between 6:00 and 8:30 AM (2.5 hours). Similarly, the eastbound afternoon peak period is between 3:30 and 6:00 PM (2.5 hours).  As mentioned in Chapter Four, the peak hour of operation is defined to be the hour that GP traffic experienced their longest travel times along the HOV facility. Accordingly, the Barnet/Hastings' westbound peak hour is between 7:30 and 8:30 AM, the last hour of the morning peak period. Likewise, the last hour of the afternoon peak period (5:00 to 6:00 PM) is the corridor's eastbound peak hour.  H*  Hastings St. (7.1km) NOTES:  M4  Barnet Highway (7.9 km)  1  1  St. Johns St. (3.0 km)  -Westbound HOV lane starts at loco Rd. in Port Moody and runs along St. John Street, turns onto Clarke St. at Moody St. and continues along Bamet Highway and Hastings St. and terminates at Renfrew St. in Vancouver. -Eastbound HOV lane starts at Renfrew St. on Hastings and terminates just north of Clarke Street on Barnet Highway -Curtis Street parallel route turns into Parker Street parallel route at Holdom Avenue  Figure 6-1: The Barnet/Hastings HOV Corridor (MoTH , 1997) 1  Chapter 6: Application of the Framework  Page 83  6.1.3 Data Analysis In accordance with framework, the effectiveness of the HOV facility on the Barnet/Hastings corridor was analyzed using the available "before-and-after" data. The data was used to calculate the respective MOEs of the primary, supporting, and operational evaluation objectives of the framework. The results of the evaluation of the effectiveness of the Barnet/Hastings HOV corridor are presented in the following sections of this chapter.  Firstly, the primary evaluation objective of 'Person Throughput' is examined. Secondly, the supporting evaluation objectives are evaluated. The results of the 'Travel Time Savings' and Travel Time Reliability' objectives are presented. However, as mentioned previously, no public opinion survey data was available for the application. Therefore, no 'Public Support' results are presented. Finally, the effectiveness of the operational objectives of 'Compliance', 'Safety' and 'Impact on GP Lanes' are assessed. In summary, the objectives are evaluated in the following order: • Person Throughput • Travel Time Savings • Travel Time Reliability • Compliance • Safety • Impact on GP Lanes  In addition, a multi-criteria evaluation to measure the cost-effectiveness of the HOV facility is performed and the results are presented in the last section of this chapter.  Chapter 6: Application of the Framework  Page 84  6.2 PERSON THROUGHPUT As specified in Chapter Three, 'Person Throughput' is the primary evaluation objective of the HOV  evaluation framework. Accordingly, the 'Person Throughput' effectiveness of  the Barnet/Hastings HOV  corridor is examined in this section.  As shown in Table 6.1, "before-and-after" vehicle classification and occupancy counts as well as travel time data were obtained for the application. Therefore, all three of the 'Person Throughput' objective measures of effectiveness (MOEs) were quantified and analyzed as described in the HOV  evaluation framework. The results and discussions of  each of the 'Person Throughput' MOEs are presented in the following order:  The  •  Per-Lane Efficiency  •  Average Vehicle Occupancy  •  HOV  (AVO)  Market Share  only data relating to the operation of the parallel routes of the HOV  corridor were  some "before-and-after" 24-hour traffic volume counts conducted by the City of Burnaby. The  results of the counts are summarized in Table 6.2. The decreases in the "before-  and-after" traffic volumes ranged from a low of 12% to a high of 44%.  Table 6-2: Parker/Curtis Street 24-Hour Traffic Volume Counts :  LOCATION  Carleton to Madison Delta to Springer Springer to Howard Holdom to Fell Kensington to Grove Sperling to Duncan  BEFORE  AFTER;  13956 20178 17703 17835 16312 17579  12298 15704 14740 13454 10873 9823  %  DECREASE 12% 22 % 17% 25% 33 % 44 %  NOTES: - "before" counts conducted on either Feb.2-'96 or Mar.3-'96 - "after" counts conducted on either Nov.14-'96 or Nov.26-'96  Chapter 6: Application of the Framework  Page 85  6.2.1 Per-Lane Efficiency Figure 6.2 shows the "before-and-after" peak hour per-lane efficiencies of the HOV facility on the Barnet/Hastings corridor. The peak hour per-lane efficiencies were calculated for each section of the corridor. The "after" results are an average of two surveys: October, 1996, and October, 1997. 120  WBSt.Johns NOTES:  WBBarnet  WBHastings  EBHastings  EBBarnet  - After results are an average of two "after" surveys • WB-.westbound and EB:eastbound  Figure 6-2: Peak Hour Per-Lane Efficiencies The westbound per-lane efficiency has more than doubled on the St. Johns St. section of the Barnet/Hastings HOV corridor. Meanwhile, the westbound per-lane efficiency has increased on the Barnet Highway section and slightly decreased on Hastings St. section. As well, the eastbound per-lane efficiencies of both the Barnet Highway and the Hastings St. sections have increased quite significantly.  Discussion of Results Overall, the HOV lane has been effective in increasing the per-lane efficiency of the corridor. With the exception of the westbound Hastings St. section, the Barnet/Hastings' peak hour per-lane efficiencies have increased since the introduction of the HOV lane.  Chapter 6: Application of the Framework  Page 86  6.2.2 Average Vehicle Occupancy Westbound Figures 6.3 and 6.4 show the Barnet/Hastings HOV corridor's westbound "before-andafter" AVOs for both the peak hour and peak period of operation.  Mar-96  Oct-96  Mar-97  Oct-97  Mar-96  Oct-96  Mar-97  Oct-97  Figure 6-3: Westbound Peak Hour AVO Figure 6-4: Westbound Peak Period AVO The westbound peak hour AVO has increased from 1.25 to an average of 1.31 over the three "after" periods of evaluation. However, the "after" peak hour AVO has decreased slightly from 1.32 to 1.30 since the first evaluation in October, 1996 to the most recent evaluation in October, 1997. Similarly, the westbound peak period AVO has increased from 1.23 to an average of 1.28 over the first full year of operation of the HOV lane. There has also been a slight decrease in the westbound "after" peak period AVO.  Table 6.3 shows the 95% confidence limits and margins of error of the westbound "before-and-after" AVOs for both the peak hour and peak period of HOV operation. The margins of error of the AVO results were calculated as a percentage of the mean. The margins of error of the westbound AVO results were determined to be adequately small at 2% or less.  Chapter 6: Application of the Framework  Page 87  Table 6-3: Westbound AVO 95% Confidence Limits and Margins of Error BEFORE (March,1996)  AFTER (October,1996)  (March,1997)  -  (October,1997)  Peak Hour AVO 1.25 1.32 95% Confidence Limits 1.22 to 1.28 1.30 to 1.34 % Margin of Error 2% 2%  1.31 1.29 to 1.34 2%  1.30 1.28 to 1.32 2%  Peak Period AVO 1.23 95% Confidence Limits 1.22 to 1.24 % Margin of Error 1%  1.29 1.27 to 1.31 1%  1.27 1.25 to 1.29 1%  1.29 1.27 to 1.31 1%  The increases in the westbound peak hour and peak period AVOs along the Barnet/Hastings HOV corridor were each determined to be statistically significant for all three "after" evaluations. The differences between each of the "after" AVOs and the "before" AVO were statistically tested at the 95 percentile for both the peak period and th  peak hour of HOV operation. A two sample f-test was used as explained in the HOV evaluation framework.  Figure 6.5 shows the westbound "before-and-after" AVOs as well as the estimates of the minimum acceptable "after" AVOs to achieve a statistically significant improvement. Both the average "after" peak hour AVO of 1.31 and the average "after" peak period AVO  of 1.28 are greater than the minimum AVOs of 1.28 and 1.25, respectively.  Therefore, the average increases in the westbound AVOs are also determined to be statistically significant at the 95% confidence level.  Chapter 6: Application of the Framework  Page 88  Peak Hour NOTES::  Peak Period  - After results are an average of three "after" surveys - Minimum based on an average AVO standard deviation and sample  size  Figure 6-5: Westbound "Before-and-After" and Minimum AVOs  Eastbound Figures 6.6 and 6.7 show the Barnet/Hastings HOV corridor's eastbound "before-andafter" AVOs for both the peak hour and peak period of operation.  Mar-96  Oct-96  Mar-97  Oct-97  Mar-96  Oct-96  Mar-97  Oct-97  Figure 6-6: Eastbound Peak Hour AVO Figure 6-7: Eastbound Peak Period AVO The eastbound "before" peak hour AVO has increased from 1.32 to a high of 1.37 in March of 1997. The average eastbound "after" peak hour AVO is 1.35 for the three "after" periods of evaluation. The eastbound peak period AVO has increased from 1.29 to an "after" average of 1.33 with no decrease in the AVO over time.  Chapter 6: Application of the Framework  Page 89  Table 6.4 shows the eastbound "before-and-after" AVO 95% confidence limits as well as the AVO margins of error as a percentage of the mean. Similar to the westbound results, the eastbound margins of error were adequately small at 3% or less. Table 6-4: Eastbound AVO 95% Confidence Limits and Margins of Error  WBEFORE (March,1996)  AFTER (October,1996)  (March,1997)  (October,1997)  Peak Hour AVO 1.32 95% Confidence Limits 1.29 to 1.35 % Margin of Error 2%  1.33 1.29 to 1.36 3%  1.37 1.33 to 1.40 3%  1.35 1.33 to 1.38 2%  Peak Period AVO 1.29 95% Confidence Limits 1.27 to 1.31 % Margin of Error 1%  1.32 1.30 to 1.34 2%  1.33 1.31 to 1.35 1%  1.33 1.32 to 1.34 1%  Figure 6.8 shows the eastbound "before-and-after" AVOs as well as the estimates of the minimum acceptable "after" AVOs. The average eastbound "after" peak hour AVO is exactly equal to the minimum acceptable AVO of 1.35. Therefore, the increase in the eastbound AVO was determined to be statistically significant. However, in October of 1996, the eastbound "after" peak hour AVO of 1.33 was less than the minimum acceptable AVO of 1.35. Therefore, it was not found to be statistically different than the "before" AVO of 1.32. The other two "after" peak hour AVOs of 1.37 and 1.35 in 1997 were both greater than (or equal to) the minimum and produced statistically significant AVO  differences.  The average eastbound "after" peak period AVO of 1.33 is greater than the minimum of 1.31 and is determined to be statistically significant at the 95% confidence level. All three "after" peak period AVOs produced statistically significant AVO differences from the "before" AVO of 1.29. In other words, each "after" peak period AVO was greater than the minimum acceptable AVO of 1.31.  Chapter 6: Application of the Framework  Page 90  1.45 1.40 1.35 -  1.32  1.35  1.35  • • •  Before Minimum After  1.29  > 1.30 --  <  1.25 1.20 1.15 Peak Hour NOTES::  Peak Period  • After results are an average of three "after" surveys - Minimum based on an average AVO standard deviation and sample  size  Figure 6-8: Eastbound "Before-and-After" and Minimum AVOs Discussion of Results Overall, the HOV facility on the Barnet/Hastings corridor has been effective at increasing the average number of persons per car on the facility. The HOV corridor has generally been able to maintain a statistically significant increase in its AVO for its full year of operation for both the peak hour and peak period. However, there seems a downward trend in the westbound AVO. If this trend continues it may lead to a statistically insignificant AVO difference.  6.2.3 HOV Market Share As described in the HOV evaluation framework, the HOV Market Share is a measure of the number of persons that commute by means of an HOV (carpool, vanpool or bus). Figures 6.9 through 6.12 show the "before-and-after" peak period HOV Market Share percentages as well the exact percentage breakdown of single occupancy vehicle (SOV), high occupancy vehicle (carpool or vanpool), and bus person-trips.  Chapter 6: Application of the Framework  Page 91  The westbound HOV Market Share has improved slightly from 48% to an average of 51%. This net 3% growth is attributable to,an 8% growth in the percentage of people commuting in a carpool or vanpool and a 5% drop in the percentage of people commuting by bus.  Westbound  HOV MARKET  SHARE = 48%  HOV MARKET  SHARE = 51%  Figure 6-9: WB - "Before" Market Share Figure 6-10: WB - "After" Market Share  Eastbound The eastbound "before" HOV Market Share has improved by one percent to 53%. Similar to the westbound results, there has been a 5% drop in the percentage of people commuting by bus. The percentage of people commuting eastbound in a carpool or vanpool has grown from 32 to 38%.  HOV MARKET  SHARE = 52%  HOV MARKET  SHARE = 53%  Figure 6-11: EB - "Before" Market Share Figure 6-12: EB - "After" Market Share  Chapter 6: Application of the Framework  Page 92  Discussion of Results The  HOV  Market Share has slightly increased since the introduction of the HOV  lane to  the Barnet/Hastings corridor. However, there has been an interesting shift in the distribution of person-trips. The percentage of bus commuters has decreased 5% in both directions while the percentage of carpooling and vanpooling traffic has increased 8% and 6% in the westbound and eastbound directions respectively.  6.2.4 The  Person Throughput Summary Barnet/Hastings HOV  lane has been effective at improving the 'Person Throughput'  of the corridor. With the exception of the westbound Hastings St. section, the  HOV  corridor's per-lane efficiencies have all increased and are continuing to grow. The AVOs were determined to have increased by statistically significant amounts and the  HOV  Market Share of the corridor has also increased in both directions of travel since the introduction of the HOV  lane.  6.3 TRAVEL TIME SAVINGS The  effectiveness of the HOV  facility on the Barnet/Hastings corridor in providing a  relative travel time savings benefit to the occupants of HOVs is evaluated in this section. Travel time savings data from October of 1996 and 1997 were obtained for the application. The travel time difference MOE and  is quantified according to the framework  the results are discussed.  Chapter 6: Application of the Framework  Page 93  6.3.1 Travel Time Difference Westbound Figure 6.13 shows the westbound peak hour and peak period travel time differences of the HOV facility of the Barnet/Hastings corridor. The westbound peak hour is between 7:30 and 8:30 AM and the westbound peak period is between 6:00 and 8:30 AM.  Oct-96  Oct-97  Figure 6-13: Westbound Travel Time Differences Table 6.5 presents the westbound travel time difference 95% confidence limits. The westbound travel time differences were all determined to be statistically significant at the 95% confidence level. Over one year, the peak hour travel time difference almost doubled from 4.3 minutes to 8.1 minutes. As well, the peak period travel time difference has more than doubled from 2.2 minutes in 1996 to 4.6 minutes in 1997.  Table 6-5: Westbound Travel Time Difference 95% Confidence Limits AFTER (October, 1996)  (October,1997)  Peak Hour Travel Time Difference 4.3 minutes 8.1 minutes 95 % Confidence Limits 3.3 to 5.3 minutes 6.2 to 9.9 minutes Peak Period Travel Time Difference 2.2 minutes 4.6 minutes 3.0 to 6.2 minutes 95 % Confidence Limits 0.9 to 3.5 minutes  Chapter 6: Application of the Framework  Page 94  Table 6.6 shows the individual section peak hour travel time differences. The St. Johns St. section is experiencing a negative benefit for the occupants of HOVs. However, this negative benefit is decreasing and less than or equal to half a minute. The Barnet Highway peak hour travel time difference has doubled from 2.6 to 5.2 minutes. Finally, the Hastings St. section experienced a slight increase in its peak hour travel time difference from 2.2 to 2.9 minutes.  Table 6-6: Westbound Peak Hour Travel Time Differences by Section  AFTER St. Johns Street Barnet HighwayHastings Street TOTAL  (October, 1996)  (October,1997)  -0.5 minutes 2.6 minutes 2.2 minutes  -0.1 minutes 5.2 minutes 2.9 minutes  4.3 minutes  8.1 minutes  Table 6.7 shows the travel speed margins of error for both the GP and the HOV travel time runs expressed in km/h as well as in percentages of the mean travel speed. The westbound travel speed margins of error ranged from a minimum of 1.5 km/h to a maximum of 4.4 km/h, all below the minimum specified error of 5 km/h. Therefore, the travel time data obtained is considered to be statistically reliable. Table 6-7 Westbound Travel Speed Margins of Error  AFTER v (October, 1996)  GP  HOV  (October,1997)  GP  HOV  Peak Hour Margin of Error (km/h) 1.5 km/h 4 % Margin of Error (% of the mean)  3.1 km/h 6%  1.8 km/h 6%  1.7 km/h 4 %  Peak Period Margin of Error (km/h) 3.8 km/h Margin of Error (% of the mean) 8%  3.2 km/h 6%  4.4 km/h 11 %  3.1 km/h 6%  Chapter 6: Application of the Framework  Page 95  Eastbound Figure 6.14 shows the eastbound peak hour and peak period travel time differences of the HOV facility of the Barnet/Hastings corridor. The eastbound peak hour is between 5:00 and 6:00 PM and the eastbound peak period is between 3:30 and 6:00 PM.  10 9 a c 8 7 £ "W 6 5 4 3 > 2  • Peak Hour  • Peak Period  <D  It CD  Oct-96  Oct-97  Figure 6-14: Eastbound Travel Time Differences Table 6.8 presents the eastbound travel time difference 95% confidence limits. Similar to the westbound direction, each eastbound travel time difference was determined to be statistically significant at the 95% confidence level. Both the peak hour and peak period eastbound travel time differences have slightly increased from 1996 to 1997. Table 6-6: Eastbound Travel Time Difference 95% Confidence Limits AFTER (October, 1996)  (October,1997)  Peak Hour Travel Time Difference 2.4 minutes 2.8 minutes 95% Confidence Limits 1.5 to 3.3 minutes 1.2 to 4.5 minutes Peak Period Travel Time Difference 1.7 minutes 2.5 minutes 1.0 to 2.4 minutes 1.7 to 3.4 minutes 95%> Confidence Limits  Chapter 6: Application of the Framework  Page 96  Table 6.9 shows the individual section eastbound peak hour travel time differences. Both the Hastings St. section and the Barnet Highway peak hour travel time differences have slightly increased by 0.3 minutes and 0.1 minutes respectively. Table 6-9: Eastbound Peak Hour Travel Time Differences by Section ~ Hastings Street Barnet Highway TOTAL  AFTER (October, 1996)  (October,1997)  1.7 minutes 0.7 minutes  2.0 minutes 0.8 minutes  2.4 minutes  2.8 minutes  Table 6.10 shows the eastbound travel speed margins of error for both the GP and HOV travel time runs expressed in km/h as well as in percentages of the mean travel speed. The eastbound travel speed margins of error are all well below the minimum specified error of 5 km/h. The maximum eastbound travel speed margin of error was 3.4 km/h (or 8% of the mean travel speed). Similar to the westbound direction, the eastbound travel time data is determined to be statistically reliable. Table 6-10: Eastbound Travel Speed Margins of Error -  AFTER  (October, 1996)  GP  :  "  (October,1997)  HOV  GP  HOV  Peak Hour Margin of Error (km/h) 2.1 km/h Margin of Error (% of the mean) 5 %  3.3 km/h 8%  3.4 km/h 8%  2.0 km/h 5 %  Peak Period Margin of Error (km/h) 1.6 km/h 4 % Margin of Error (% of the mean)  1.7 km/h 4 %  2.0 km/h 5%  1.8 km/h 4 %  6.3.2 Travel Time Savings Summary Overall, the HOV facility on the Barnet/Hastings corridor has been effective at providing a statistically significant travel time savings benefit (at the 95% level) to the occupants of  Chapter 6: Application of the Framework  Page 97  HOVs. The peak hour and peak period travel time differences have increased in both directions of travel from 1996 to 1997.  The westbound travel time differences are significantly higher than the eastbound's. For example, the 1997 westbound peak hour travel time difference was 8.1 minutes. Comparatively, the 1997 eastbound peak hour travel time difference was only 2.8 minutes. However, it should be noted that the westbound HOV lane is longer in length than the eastbound by an additional 2.9 kilometers.  6.4 TRAVEL TIME RELIABILITY The effectiveness of the HOV facility at providing a more reliable trip to HOV drivers and their occupants is evaluated using travel time data obtained in October, 1997.  6.4.1 Travel Speed Standard Deviation Westbound Figure 6.15 shows the westbound travel speed standard deviations of the HOV and GP lanes for the departure times of 7:00 AM and 8:00 AM respectively. Both departure times are within the westbound peak period (6:00 to 8:30 AM) and the latter is within the peak hour of HOV operation (7:30 to 8:30 AM).  Chapter 6: Application of the Framework  Page 98  o _ 20%  S "S  DGPlane  > Q) <D Q L  UHOV lane  13.3%  Q w 15%  11  § c 10% --+  8  ,  2  %  8.0%  6.5%  CO g  •o E a. £ CO ~  15  5% ^  0%  1  7:00 AM  8:00 AM  Figure 6-15: Westbound Travel Speed Standard Deviations Table 6.11 presents the westbound travel speed standard deviations expressed in kilometers per hour as well as the results of the statistical tests performed on the differences between the HOV and GP standard deviations using the F-test as explained earlier. Table 6-11: Westbound Travel Speed Standard Deviations ... DEPARTURE TIME 7:00 AM 8:00 AM  GP LANE  HOV LANE  SIGNIFICANT DIFFERENCE?  3.3 km/h 5.2 km/h  3.9 km/h 3.0 km/h  No YES  It was determined that there is no statistically significant difference between the HOV and GP travel speed standard deviations at either the 95% or the 90% confidence level for the 7:00 AM departure time. As an absolute value, the GP travel speed standard deviation of 3.3 km/h was actually less than the HOV lane (3.9 km/h). However, the HOV lane is providing a more reliable trip at 8:00 AM. The HOV lane's travel speed standard deviation is statistically significantly less than that of the GP lane at the 95% confidence level.  Chapter 6: Application of the Framework  Page 99  Table 6.12 shows the westbound travel speed margins of error for both the GP and HOV  travel time runs expressed in kilometers per hour as well as in percentages of the  mean travel speed. The margins of error are all very low (less than 3.3 km/h) indicating that the travel time data is statistically reliable. Table 6-12: Westbound Travel Speed Margins of Error DEPARTURE TIME 7:00 AM  GP Margin of Error (km/h) 2.1 km/h Margin of Error (% of the mean) 5 %  8:00 AM  HOV  GP  HOV  2.5 km/h 5%  3.3 km/h 8%  1.9 km/h 4%  Eastbound Figure 6.16 shows the eastbound travel speed standard deviations of the HOV and GP lanes for the departure times of 4:30 PM and 5:30 PM respectively. Both departure times are within the eastbound peak period (3:30 to 6:00 PM) and the latter is within the peak hour of HOV operation (5:00 to 6:00 PM).  2 -  0% 4:30 PM  5:30 PM  Figure 6-16: Eastbound Travel Speed Standard Deviations  Chapter 6: Application of the Framework  Page 100  Table 6.13 presents the eastbound travel speed standard deviations expressed in kilometers per hour as well as the results of the statistical tests performed on the differences between the HOV and GP standard deviations. Table 6-13: Eastbound Travel Speed Standard Deviations DEPARTURE TIME 4:30 PM 5:30 PM  GP LANE  HOV LANE  SIGNIFICANT DIFFERENCE ?  4.2 km/h 3.2 km/h  3.5 km/h 2.5 km/h  No No  For both departure times, the differences in the eastbound travel speed standard deviations were not determined to be statistically significant at either the 95% or 90% confidence levels. Therefore, even though the eastbound GP lane's standard deviations are greater than the HOV lane, the differences are not adequate enough to conclude that the HOV lane is providing a more reliable trip.  Table 6.14 shows the eastbound travel speed margins of error for both the GP and HOV travel time runs expressed in kilometers per hour as well as in percentages of the mean travel speed. The travel speed margins of error ranged from a minimum of 2.5 km/h to a maximum of 4.2 km/h, all below the minimum specified error of 5km/h. Therefore, the eastbound travel time data is also considered to be statistically reliable. Table 6-14: Eastbound Travel Speed Margins of Error DEPARTURE TIME 4:30 PM  GP Margin of Error (km/h) 4.2 km/h Margin of Error (% of the mean) 9%  Chapter 6: Application of the Framework  ^~  5:30 PM  HOV  GP  HOV  3.5 km/h 7%  3.2 km/h 7%  2.5 km/h 5%  Page 101  6.4.2 Travel Time Reliability Summary Overall, the HOV facility on the Barnet/Hastings corridor has not provided a statistically significant more reliable trip. It was determined that there is no statistically significant difference between the eastbound HOV and GP travel speed standard deviations. The HOV lane was determined to be more reliable than the adjacent GP lanes only during the morning peak hour of HOV operation (departure time of 8:00 AM).  6.5 COMPLIANCE The compliance rates of the HOV facility on the Barnet/Hastings corridor are determined according to the HOV evaluation framework in this section. The effectiveness of the HOV facility at achieving MoTH's desired compliance rate of 85% is evaluated and the results are discussed. The compliance rates were calculated based on vehicle classification and occupancy counts obtained in the following months: October, 1996; March, 1997; and October, 1997. In addition, the results of a search of HOV violation tickets are presented as well as a summary of the HOV enforcement activity of the Burnaby Royal Canadian Mounted Police (RCMP) Detachment.  6.5.1 Compliance Rates The compliance rates of each location of the vehicle classification and occupancy surveys were calculated for each 15-minute interval of the peak period. It was observed that the compliance rates of the first and last 15-minute intervals of the peak period were consistently  lower than the remaining 15-minute intervals. The following  compliance data obtained from a location along the St. Johns Street section of the Barnet/Hastings HOV corridor illustrates this point.  Chapter 6: Application of the Framework  Page 102  Example: Compliance data from Williams location of St. Johns St. section, March 1997 (Source of data: Appendix G - Compliance Data ) TIME  COMPLIANCE RATE  6:00 - 6:15 6:15-6:30 6:30 - 6:45 6:45 - 7:00 7:00 -7:15 7:15-7:30 7:30 - 7:45 7:45 - 8:00 8:00 - 8:15 8:15-8:30  80.6 % 86.6 % 84.7% 87.9 % 88.7% 88.3 % 89.4 % 88.2 % 70.3 %  Average = 83% (including  first & last interval)  Average = 87% (excluding  first & last interval)  The compliance rate of the location is only 83% when the first and last 15-minute intervals are included. If these intervals are excluded, the compliance rate is 87% and the desired rate of 85% is achieved.  Based on interviews with the local enforcement agencies, it was determined that the compliance rates of the first and last 15-minute intervals are consistently lower than the average because the start and end times of the HOV operation are subjective. The local enforcement agencies usually allow a ten-minute period of grace during these times. In this specific example, police officers would most likely not have been ticketing HOV violators between 6:00 to 6:10 AM and between 8:20 to 8:30 AM. The compliance rate excluding the first and last 15-minute intervals is more representative of the compliance during routine HOV operation. Therefore, all of the compliance rates of the HOV facility on the Barnet/Hastings corridor were determined in this manner.  Chapter 6: Application of the Framework  Page 103  Westbound Figure 6.17 shows the compliance rates for the three westbound sections of the  HOV  facility on the Barnet/Hastings corridor for each compliance survey conducted. The westbound compliance rates in October, 1996, were all above the desired rate of 85%. By March of 1997, the compliance rates on the Barnet/Hastings HOV  corridor were  decreasing. The Barnet Highway section was already experiencing a less than desired rate of 81%. Each of the section's compliance rates were determined to be below 85% by October, 1997. Overall, the average westbound compliance rates of each section are 84%, 80%, and 83% respectively.  100%  Oct-96  Mar-97  Oct-97  Average  Figure 6-17 : Westbound Compliance Rates by Section Table 6.15 summarizes the westbound compliance rates by location as well the westbound corridor averages for each compliance survey. The average compliance rates for the corridor in October, 1996, and in March, 1997, were equal to or greater than the desired rate of 85%. However, similar to the section averages shown in Figure  Chapter 6: Application of the Framework  Page 104  6.17, the average compliance rates for the corridor decreased below 85% in October, 1997. Table 6-15: Westbound Compliance Rates by Location SECTION  , .  LOCATION  St. Johns Street  Williams Clark Reed Bayview Willingdon Esmond Lillooet*  Barnet Highway  Hastings Street  AVERAGE*  MAR-97  OCT-97  88% 87%  87% 89 %  86% 87%  76 % 85 %  83 % 70% 78 % 69 %  89 % 86 % 77%  89 % 82% 74 %  78 % 74 % 33%  87%  85%  75 %  OCT-96 :  * Lillooet location not included in average  The  Lillooet compliance rates are not considered to be representative of the routine  compliance along the corridor because they were obtained at the end of the  HOV  facility. Therefore, the Lillooet compliance rates were not included in the calculation of either the westbound Hastings Street section or westbound corridor averages.  Eastbound Figure 6.18 shows the compliance rates for the two eastbound sections of the  HOV  facility on the Barnet/Hastings corridor for each compliance survey conducted. Eastbound section averages are also presented. Similar to the westbound direction, the eastbound compliance rates were all above the desired rate of 85% in October, 1996, decreasing by March, 1997, and below 85% by October, 1997. Overall, the average eastbound compliance rate on the Hastings Street section was determined to be 79%. However, the average eastbound compliance rate on the Barnet Highway section was determined to be equal to the desired rate of 85%.  Chapter 6: Application of the Framework  Page 105  100% • Hastings  Oct-96  Mar-97  • Barnet  Oct-97  Average  Figure 6-18: Eastbound Compliance Rates by Section Table 6.16 summarizes the eastbound compliance rates by location as well the eastbound corridor averages for each compliance survey. Table 6-16: Eastbound Compliance Rates by Location SECTION  Hastings Street  Barnet Highway  LOCATION  OCT-96  MAR-97  % % % % % %  OCT-97  Lillooet* Cassiar Kootenay Carleton Willingdon Howard  65 % 82% 86 % 87% 85 % 84 %  64 68 75 80 74 81  55 73 80 76 79 78  % % % % % %  Bayview Union  89% 87%  95% 80 %  89% 69 %  AVERAGE*  86%  80%  78 %  * Lillooet location not included in average  The HOV facility on the Barnet/Hastings corridor achieved its desired rate of compliance in the eastbound direction for the surveys conducted in October, 1996, and March, 1997. However, the eastbound corridor compliance rate of 78% in October of 1997 is  Chapter 6: Application of the Framework  Page 106  7% short of the desired rate of 85%. Once again, the Lillooet compliance rates were not included in the calculation of either the eastbound compliance averages.  6.5.2 Violation Data A search for traffic violation ticket information concerning HOV lanes was requested and performed by the Research Services of ICBC. Traffic violations related to the HOV lane were retrieved based on the section numbers of offences in the Motor Vehicle Act (MVA). For example, the offence related to HOV lanes is presently described in section number 152 of the MVA  as follows:  "If a laned roadway  has a high occupancy  vehicle lane, a person  motor vehicle or other device in that lane unless permitted  must not drive a  by the  regulations."  (Source: MVA, Section Number 152: High Occupancy Vehicle Lane, page 94)  The following violation data was retrieved from ICBC for the period of September 1, 1996 to October 24, 1997: •  Date of Violation  •  Location of Violation  •  Section of Motor Vehicle Act  •  Count of the number of HOV violations  Table 6.17 summarizes the number of violations identified from the ICBC search on a per month basis for the first full year of HOV operation (September, 1996 to August, 1997). The results of the ICBC search, however, are not conclusive on the total number of HOV violation tickets that have been issued for the following reasons: • The section number of the HOV offence was 155.1 prior to April 1, 1997 and section number 152 thereafter. Violation data may have not been reported  Chapter 6: Application of the Framework  Page 107  under the correct section number during the transition period. For example, the search produced no count of the number of violation tickets issued for the month of April, 1997. The  MVA section  numbers are not consistently  entered  into ICBC's  Contravention System. For example, some counts were retrieved from a search of section number "155.1" and others were retrieved from a search of "155/1". •  It was identified that some of the local enforcement agencies have the choice of issuing either a MVA ticket or a Bylaw ticket for an HOV offence. No statistics on Bylaw tickets by section or location were available. Many HOV violation warnings were issued in the first year of HOV operation. Warnings would not have appeared as HOV violation tickets in the search. Table 6-17: ICBC HOV Violation Ticket Information MONTH  September, 1996 October, 1996 November, 1996 December, 1996 January, 1997 February, 1997 March, 1997 April, 1997 May, 1997 June, 1997 July, 1997 August, 1997  NUMBER  OF HOV  TICKETS  38 19 11 14 91 16 76 -  62 20 96 44  6.5.3 Enforcement Data A search was also performed by the Burnaby RCMP Detachment on their level of enforcement activity on the HOV facility of the Barnet/Hastings corridor. The purpose of the search was to determine an order of magnitude estimate on how many man-hours  Chapter 6: Application of the Framework  Page 108  have been invested in enforcing the HOV lane in the city of Burnaby thus far. The search was conducted for the period between September, 1996, and December, 1997.  MoTH funded special enforcement on the HOV facility for the first six weeks of HOV operation. Table 6.18 shows an estimate of the number of man-hours invested in the special enforcement for the Burnaby RCMP Detachment. It should be noted that MoTH also funded special enforcement with the Vancouver Police and the Port Moody Police during this time. Table 6-18: Burnaby RCMP Special Enforcement Man-Hours MONTH  MAN-HOURS  September, 1996 October, 1996  95 45  Table 6.19 shows the routine HOV enforcement activity of the Burnaby RCMP Detachment in man-hours on a per month basis for the first full year of HOV operation. The estimate of the number of man-hours ranged from a low of 4 per month (December, 1996) to a high of 32 per month (August, 1997).  It is interesting to note that the months with the highest number of invested man-hours (Jan.=29, Mar.=29, and Aug.=32) correspond directly with the months with the highest number of HOV violation tickets (Jan.=91, Mar.=76, and Aug.=96) indicating that the violation and enforcement data are somewhat consistent with each other. Unfortunately, it is not possible to correlate the number of invested man-hours with the compliance rate due to the lack of extensive compliance rate data.  Chapter 6: Application of the Framework  Page 109  Table 6-19: Burnaby RCMP Routine Enforcement Man-Hours MONTH September, 1996 October, 1996 November, 1996 December, 1996 January, 1997 February, 1997 March, 1997 April, 1997 May, 1997 June, 1997 July, 1997 August, 1997  MAN-HOURS 20 10 25 4 29 14 29 27 21 11 32 17  6.5.4 Compliance Summary The average compliance rate for any given section or direction ranged from 79% to 84%. Similarly, the average corridor compliance rate for any given compliance survey ranged from 75% to 87% in the westbound direction and from 78% to 86% in the eastbound direction. Overall, the HOV facility on the Barnet/Hastings corridor has not achieved its desired rate of compliance of 85% but has maintained an effective rate of no less than 75% on any given section or direction for any given survey. However, it should be noted that the compliance rate has been decreasing since the opening of the HOV lane in the Fallot 1996.  In addition to the compliance rates, estimates of the total number of HOV violation tickets issued along the corridor and the level of enforcement activity along the Burnaby section of the HOV facility were presented. While not conclusive, the violation and enforcement data results are useful in developing an understanding of the level of enforcement that occurred during the HOV facility's first full year of operation.  Chapter 6: Application of the Framework  Page 110  6.6 SAFETY The  'Safety' objective is evaluated in this section for the HOV facility on the  Barnet/Hastings corridor. Four years of "before" accident data (Jan.'92 - Dec.'95) and ten months of "after" accident data (Sept.'96 - Jun.'97) from MoTH's accident database were analyzed according to the framework. In addition, raw accident data was also obtained from the City of Burnaby and the City of Vancouver. However, the accident data was determined to be inconsistent and, therefore, inaccurate. The results of the three sources contradict each other. Therefore, only MoTH's accident data was analyzed as part of the application. The accident data of the City of Burnaby and the City of Vancouver can be found in Appendix H - Safety Data.  The "before-and-after" accident data from MoTH did not include any accident statistics on the most western part of the Hastings St. section of the HOV facility. The safety of the last few kilometers of the HOV facility, west of Boundary Road to Renfrew Street, is the responsibility of the City of Vancouver (see Figure 6.1). Therefore, the safety analysis was performed from Boundary Road in Vancouver to loco Road in Port Moody. The corridor was divided up into major intersections and short road sections comprised of minor intersections and road segments. Table 6.20 summarizes the intersections as well as the sections and their respective lengths.  To calculate the accident rates, many assumptions had to be made. For example, no "after" AADT data was available so it was assumed to have remained constant. Therefore, no conclusions or interpretations were made from the analysis of the safety data.  Chapter 6: Application of the Framework  Page 111  Table 6-20: Summary of Intersections and Sections INTERSECTION  LKI  Boundary Gilmore Willingdon Springer Ellensmere Holdom Sperling  0.0 0.7 1.5 2.7 3.0 3.2 4.3  Gore St. Johns Queens Moody Hugh Dewdney loco  12.8 13.5 14.1 14.7 14.9 16.3 16.4  SECTION  LKI  LKI  LENGTH  (start)  (finish)  (kms)  0.0 0.7 1.5 2.7 3.0 3.2 4.3 5.5 7.3 10.2 11.4 12.8 13.5 14.1 14.7 14.9  0.7 1.5 2.7 3.0 3.2 4.3 5.5 7.3 10.2 11.4 12.8 13.5 14.1 14.7 14.9 16.3  0.7 0.8 1.2 0.3 0.2 1.1 1.2 1.8 2.9 1.2 1.4 0.7 0.6 0.6 0.2 1.4  Boundary to Gilmore Gilmore to Willingdon Willingdon to Springer Springer to Ellensmere Ellensmere to Holdom Holdom to Sperling Sperling to Malibu Malibu to Texaco Texaco to PetroCan PetroCan to Reed Reed to Gore Gore to St. Johns St. Johns to Queens Queens to Moody Moody to Hugh Hugh to Dewdney  However, the accident rates were still calculated as part of the application as an example for future HOV evaluations. Information concerning the calculation of the accident rates and any assumptions that were made are described in the following list: • The "before" AADT volume DATA for the road sections was obtained from a short count location at Holdom Ave. on Hastings St. in the city of Burnaby (see Table 6.21). The average AADT for the 4 years between 1992 and 1995 was determined to be 33,400 vehicles per day. It was assumed that the AADT was constant for the entire corridor. Table 6-21: Holdom Avenue Traffic Volume Data (1992 -1995) (Source: Traffic Information Management System, MoTH) YEAR  1992  1993  1994  1995  AADT  30,725  34,671  33,012  35,101  Chapter 6: Application of the Framework  AVERAGE 33,400  Page 112  • No "after" AADT volume data for the sections was available for the application. Therefore, an assumption was made that the "after" AADT volume data for sections had remained constant at 33,400 vehicles per day. • No "before-and-after" AADT volume data was available for the intersections. Therefore, intersections were assumed to have the same "before-and-after" AADT volume data as the road sections (33,400 veh/day). • Only ten months of "after" data were available at the time of the application. It was assumed that this was a representative amount of time post-construction of the HOV lane. • The lengths of the sections were assumed to be equal to the exact distance in between two major intersections. In other words, the length of the intersections themselves was assumed to be negligible. • Since the opening of the HOV facility, there have been changes made by the local enforcement agencies in the reporting procedure of accidents. Therefore, it is possible that the "after" accident data was not reported under the same criteria as the "before" data. It is generally conceived that less accidents are being reported since these changes were made.  6.6.1  Accident Rates Intersections  Figure 6.19 summarizes the "before-and-after" accident rates of the intersections of the HOV facility on the Barnet/Hastings corridor. Ten out of the fourteen intersections (71%) experienced a decrease in their "before-and-after" accident rates. The four intersections  Chapter 6: Application of the Framework  Page 113  whose accident rates have increased since the implementation of the HOV  lane are as  follows:  The  •  Ellensmere Avenue  (Hastings St.)  •  Gore Street  (Barnet Highway)  • Queens Street  (St. Johns St.)  •  (St. Johns St.)  Hugh Street  mean "before" intersection accident was calculated to be 0.74 acc/MEV as  compared to the mean "after" intersection accident rate of 0.58 acc/MEV. The difference between the "before-and-after" mean intersection accident rates was tested for statistical significance. It was determined that even though the mean accident rate for intersections along the Barnet/Hastings corridor has decreased since prior to the  HOV  lane, the difference between the two means was not statistically significant at the 95% or 90% confidence levels.  Sections Figure 6.20 summarizes the "before-and-after" accident rates of the sections of the  HOV  facility on the Barnet/Hastings corridor. Thirteen out of the sixteen road sections (81%) experienced an increase in their "before-and-after" accident rates. The intersections whose accident rates have decreased since the implementation of the HOV •  Reed Point to Gore Street  (Barnet Highway)  •  St. Johns St. to Queens Street  (St. Johns St.)  •  Hugh Street to Dewdney Trunk Road  (St. Johns St.)  Chapter 6: Application of the Framework  lane are:  Page 114  The difference between the mean "before-and-after" accident rates of the section was determined to be statistically significant at the 95% confidence level. The mean "before" section accident rate was calculated to be 0.83 acc/MVK and the mean "after" section accident rate was calculated to be 1.63 acc/MVK.  6.6.2 Safety Summary The accident and traffic volume data obtained for the application of the framework to evaluate the 'Safety' objective was incomplete and inconsistent. It was not possible to perform  a  meaningful  "before-and-after'  safety  analysis  where  representative  conclusions could be made. However, the application of the safety objective was still performed with the data obtained as an example for future HOV evaluations.  6.7 IMPACT ON GP LANES The effect of the HOV facility on the Barnet/Hastings corridor on the operation of the general purpose (GP) lanes was evaluated in this section. As noted previously, the HOV lane should not adversely affect the GP lane. Travel time survey and vehicle classification and occupancy data were obtained for the application.  "Before" data was obtained in March, 1996 and "after" was obtained in both October, 1997 and October, 1997. The GP travel speeds of the entire corridor were quantified as well as the GP vehicular volumes for each section of the HOV facility.  Chapter 6: Application of the Framework  Page 115  > co  oo  CO  • •  c O) q "D  o o o  o  00  Q  • •  CD  cu c CD  Q  &  *:  *  d  >  X CM  CD  •  •  •  •  o m O d d  CD  CO  o O d d • •  ci Ul  c CO CD CO CO CO d d  Que  CD  < 3 Ul -J  • •  CO  c CD o CD CM d  o CO  • •  o  c  o  00 CO CO d d Q CO  • •  E o if) o CO  T> O d I  co  0  ra rr  d  lensmere  • •  o c d 1  Q CO  o o d  • •  • •  LU  CO  •g  o o < c o o CO CO CD  -*—'  _c  c  o  T3 D CO CD C •= d d  • •  CO  T3 C CO I  CO  o M—  0  e°  aCO CM  TJ CO CO d d O CD  i in  CD  c  CM 00 O d  • •  CD  CD CO g  Chapter 6: Application of the Framework  Figure 6.20 : "Before-and-After" Section Accident Rates  LEGEND:  Page 117  m After  O Before  6.7.1 GP Travel Speeds Figures 6.21 and 6.22 show the mean peak hour "before-and-after" GP travel speeds of the corridor. The westbound GP travel speed of the corridor has increased from 30 km/h to an average of 36 km/h since the introduction of the HOV lane. In the eastbound direction, the GP travel speed has also increased from 28 km/h to 39 km/h.  Mar-96  Oct-96  Oct-97  Mar-96  Oct-96  Oct-97  Figure 6-21: Westbound GP Travel Speeds Figure 6-22: Eastbound GP Travel Speeds  6.7.2 GP Vehicular Volumes Westbound Table 6.25 summarizes the westbound peak hour "before" and most recent "after GP travel speeds and GP vehicular volumes of specific locations along the Barnet/Hastings corridor where the HOV lane was added to the existing facility. Along St. Johns St., the GP travel speed has more than doubled from 18 km/h to 42 km/h as the vehicular volume has remained the same. The GP travel speed along the Barnet Highway has slightly decreased from 44 km/h to 40 km/h. However, the GP volumes have increased at both the Reed and Bayview locations. Finally, the Hastings St. GP travel speed has slightly decreased as its GP vehicular volume has remained approximately the same.  Chapter 6: Application of the Framework  Page 118  Table 6-22: Westbound "Before-and-After" GP Vehicular Volumes LOCATION Williams (St. Johns St.)  Reed (Barnet Hwy)  Bayview (Barnet Hwy)  Willingdon (Hastings St.)  TYPE OF GP DATA; -  Mar-1996 ^  Oct-1997  travel speed (km/h) vehicular volume  18 2060  42 2060  travel speed (km/h) vehicular volume  44 1360  40 1500  travel speed (km/h) vehicular volume travel speed (km/h) vehicular volume  44  40 1400 24 1890  1390 28 1910  Eastbound Table 6.26 summarizes the eastbound peak hour "before" and most recent "after" GP travel speeds and GP vehicular volumes of specific locations along the HOV facility on the Barnet/Hastings corridor. Table 6-23: Eastbound "Before-and-After" GP Vehicular Volumes LOCATION Willingdon (Hastings St.)  Bayview (Barnet Hwy)  Union (Barnet Hwy)  TYPE OF GP DATA  Mar-1996 ,  Oct-1997  travel speed (km/h) vehicular volume  23 1830  28 2070  travel speed (km/h) vehicular volume  35 1380  59 1640  travel speed (km/h) vehicular volume  35 1430  59 1820  Along Hastings St., the GP travel speed has increased from 23 km/h to 28 km/h as the vehicular volume has increased from 1830 to 2070. The GP travel speed along the Barnet Highway has increased significantly from 35 km/h to 59 km/h. At the same time, the GP volumes have also increased significantly at both the Bayview and Union locations.  Chapter 6: Application of the Framework  Page 119  6.7.3  Impact on GP Lanes Summary  Overall, the HOV  lane has not adversely affected the operation of the GP lanes. There  has been an increase in the peak hour corridor GP travel speed in both directions of travel: 6 km/h westbound and 11 km/h eastbound. On an aggregate level, the GP travel speed has increased on one out of the three westbound sections and on all of the eastbound sections while the GP vehicular volume has increased or remained the same.  6.8 MULTI-CRITERIA EVALUATION A multi-criteria evaluation was performed on the HOV HOV  facility of the Barnet/Hastings  corridor. This section summarizes the results of the evaluation and sensitivity  analysis. It was not possible to calculate the present value of the travel time saved because the Greater Vancouver Regional District (GVRD) EMME/2 model was under calibration at the time of the application. Therefore, it was not possible to quantify the cost-effectiveness MOEs as described in the framework. However, for illustration purposes only, the EMME/2 model was used to obtain an understanding of the magnitude of the travel time savings.  No enforcement or marketing costs were accrued or anticipated in the future for the HOV  facility. Therefore, the present values of these cost were not quantified. However,  the present value of the following benefits and costs were determined: • Vehicle operating cost savings • Air quality savings •  Construction costs  •  Maintenance costs  Chapter 6: Application of the Framework  Page 120  6.8.1 Project Parameters It was assumed that the most likely value of the life of the HOV facility on the Barnet/Hastings corridor is 20 years, as recommended in MoTH's Economic Analysis Guidebook (MoTH, 1993). A typical year is comprised of 250 working days. As previously mentioned, the range of discount rates to be used in the multi-criteria evaluation is between 6% to 12% with a most likely value of 8%. However, it is common practice to assume a discount rate of 0% to 4% for environmental benefits and costs (Bein et al., 1994). Therefore, a conservative fixed discount rate of 4% was selected to calculate the air quality savings of the facility. Actual person-volumes were obtained from vehicle classification and occupancy counts for 1996 and 1997. It is assumed that the traffic volume on the HOV facility will increase over time. Therefore, the personvolumes of future years were based on the 1997 figures and a most likely growth rate of 2%.  6.8.2 Travel Time Savings The westbound travel time savings was quantified from the base year of 1996 to 2006, half of the life of the project. Peak hour 2006 travel time differences were obtained from the EMME/2 model. Peak hour travel time differences for 1996 were obtained from actual travel time survey data. The peak hour travel time differences of the years between 1996 and 2006 were obtained through interpolation. Table 6.28 summarizes the 1996 and 2006 peak hour travel time differences. Table 6-24: Peak Hour Travel Time Differences LANE  GP HOV  YEAR  1996 2006 1996 2006  JSt.Jbhns St.  5.8 6.4 5.4 7.5  BarnetHwM  Hastings St.  1.9 6.9 4.5 7.6  0.8 19.6 3.0 20.2  TOTAL  8.5 32.9 12.9 35.3  NOTES: - 1996 values based on "before" travel times (Mar-96) and "after" travel times (Oct-96): APPENDIX E - 2006 values based on EMME/2 model output: APPENDIX J  Chapter 6: Application of the Framework  Page 121  The most likely present value of the westbound peak hour travel time savings was estimated to be $22,530,000 for half of the life of the project. The total amount of travel time savings in the westbound direction was calculated by multiplying the present value of the westbound peak hour travel time savings by a factor of 1.97. The factor was determined based on the travel time savings values reported in the 'Benefit/Cost Evaluation of Barnet/Hastings Project' (MoTH, 1993). The determination of the westbound and eastbound factors is found in Appendix J - Multi-Criteria Evaluation. Therefore, the most likely present value of the total westbound travel time savings was estimated to be $44,380,000 (for 1996 to 2006). The results of the sensitivity analysis performed on the total westbound travel time savings are summarized in Table 6.29. Table 6-25: Westbound Travel Time Savings Sensitivity Analysis CRITERIA Discount Rate Average Wage Rate Growth Rate  WORST CASE 12% $14/hr 0%  m/IOSTMlKELY^  TOTAL PV  $26,090,000  $44,380,000  NOTES:  8% $18/hr 2%  BESJ^ASEM 6% $23/hr 5% $72,710,000  - calculated for the years 1996 to 2006 only - Total PV = output of sensitivity x 1.97 (APPENDIX J)  6.8.3 Vehicle Operating Cost Savings The most likely present value of the vehicle operating cost savings was determined to be $1,590,000. The vehicle operating cost savings was calculated for the westbound and eastbound peak periods based on the mean and 95% confidence limits of the 1997 data (see Appendix C - AVO Data). Table 6.30 summarizes the results of the sensitivity analysis on the vehicle operating cost savings.  Chapter 6: Application of the Framework  Page 122  Table 6-26: Vehicle Operating Cost Savings Sensitivity Analysis CRITERIA  WORST CASE  Discount Rate AVO - Westbound AVO - Eastbound Unit Cost Growth Rate  12% 1.25 1.32 $0.08/km 0% $660,000  TOTAL PV  6.8.4  MOST LIKELY 8% 1.27 1.33 $0.09/km 2%  BEST CASE 6% 1.29 1.34 $0.12/km 5%  $1,590,000  $4,150,000  Air Quality Savings  The most likely present value of the air quality savings was determined to be $960,000. Similar to the vehicle operating cost savings, the air quality savings was calculated from 1997 AVO  data for the peak periods of HOV  operation only. Table 6.31 summarizes the  results of the sensitivity analysis on the air quality savings. Table 6-27: Air Quality Savings Sensitivity Analysis CRITERIA AVO - Westbound AVO - Eastbound Unit Cost Growth Rate  WORST CASE 1.25 1.32 $0.02/km 0%  MOST LIKELY 1.27 1.33 $0.04/km 2%  TOTAL PV  $270,000  $960,000  .BESTCASE 1.29 1.34 $0.08/km 5% $3,320,000  NOTES: - a fixed discount rate of 4% was used to calculate each air quality savings PV  6.8.5  Construction Costs  The actual construction costs of the HOV facility on the Barnet/Hastings HOV corridor were obtained from the 'Barnet/Hastings People Moving Project Completion Report'  (MoTH , 1997). The most likely total present value of the construction costs was 2  calculated to be $127,330,000. The worst case construction cost (discount rate of 12%) was determined to be $140,340,000 and the best case construction cost (discount rate of 6%) was determined to be $121,340,000.  Chapter 6: Application of the Framework  Page 123  6.8.6 Maintenance Costs The total number of lane-kilometers of the HOV facility was calculated to be 82.75 kilometers. The total maintenance cost per year for the HOV facility was obtained from MoTH's 'Service Area #6 Year 2 Maintenance Services Contract' (MoTH , 1997). A 3  yearly rate of $6326.83/lane-km was reported for Class 1A highways such as the Barnet/Hastings corridor. The yearly rate was increased and decreased by 10% for the sensitivity analysis. A total present value of $5,550,000 was determined for the maintenance costs. Table 6.32 summarizes the results of the sensitivity analysis on the maintenance costs.  Table 6-28: Maintenance Costs Sensitivity Analysis CRITERIA  WORST CASE  MOST LIKELY  BEST CASE  Maintenance Cost Discount Rate  $5694/lane-km 6%  $6327/lane-km 8%  $6960/lane-km 12%  TOTAL PV  $7,000,000  $5,551,000  $3,940,000  6.8.7 Summary The following benefits and costs of the HOV facility on the Barnet/Hastings corridor were quantified as part of the multi-criteria evaluation: • Construction costs  $127,330,000 (PV)  • Maintenance costs  $5,550,000 (PV)  • Vehicle operating cost savings  $1,590,000 (PV)  • Air quality savings  $960,000 (PV)  In addition, the present value of the westbound travel time savings for half of the life of the project was determined to be $44,380,000. Based on these estimates, it is clear that the HOV travel time savings alone will greatly exceed the construction and maintenance costs of the HOV facility. Therefore, it is estimated that the Barnet/Hastings HOV facility is cost-effective.  Chapter 6: Application of the Framework  Page 124  7. CONCLUSION An HOV monitoring and evaluation framework was developed to evaluate the effectiveness of either arterial or freeway HOV facility. An extensive literature review on HOV evaluation practices in North America was performed and the framework was designed based on these HOV evaluation procedures that currently exist.  The HOV monitoring and evaluation framework is a procedure to evaluate an HOV facility based on a set of quantifiable objectives. As a contribution to existing HOV evaluation procedures, the evaluation objectives are selected and categorized into three groups by their relationship to the goals of the HOV network: primary, supporting, and operational. Measures of effectiveness (MOEs) that directly relate to the set of objectives are selected to determine the effectiveness of each objective. The data requirements to evaluate each of the MOEs is summarized. In addition, a data collection methodology and comprehensive procedures for analyzing and presenting the MOEs are described in the framework. To enhance current practices of evaluating HOV facilities, the framework describes procedures to calculate the statistical reliability and relative uncertainty of the data collected as well the testing of the difference between two MOEs for statistical significance. A multi-criteria HOV evaluation and sensitivity analysis methodology is presented for evaluating the cost-effectiveness of an HOV facility. An application of the proposed HOV monitoring and evaluation framework was performed to evaluate the effectiveness of the HOV facility on the Barnet/Hastings corridor. "Before-and-after" data was retrieved through numerous agencies and used to calculate the respective MOEs of the primary, supporting, and operational objectives of  Chapter 7: Conclusion  Page 125  the framework. The objectives of the framework that are considered to be effective include: 'Person Throughput', 'Travel Time Savings', and 'Impact on GP Lanes'. The travel speed standard deviations of the HOV lane are less than the GP lane. However, it was determined the HOV lane is not providing a statistically significant more reliable trip at the 95% confidence level. The Ministry of Transportation and Highways desired compliance rate of 85% is not presently being attained on the Barnet/Hastings corridor. No public opinion survey data relating to the operation of the HOV facility is available for the application and the safety results are not conclusive. Finally, a preliminary analysis indicated that the HOV facility on the Barnet/Hastings corridor is cost-effective.  In applying the framework, deficiencies in the availability of applicable data were noted as well as suggestions for further HOV evaluation research. The following is a list of the deficiencies and/or recommendations presented in the thesis: • Extensive baseline information should always be collected on a corridor prior to the construction of an HOV facility. The HOV facility should be regularly evaluated in its first few years of operation. • Parallel routes adjacent to an HOV facility should be identified in advance and appropriate "before-and-after" data collected accordingly. • AADT volume data should be collected yearly at a minimum of one representative location along an HOV facility. • Travel time and vehicle classification data should be collected at approximately the same time of year for HOV evaluation purposes, preferably the same week of a calendar year.  Chapter 7: Conclusion  Page 126  • A public opinion survey should be conducted on the HOV facility to establish the public's perception of the operation and effectiveness of the HOV lane. • The level of enforcement activity of local enforcement agencies should be kept in records and correlated to the compliance rate to enhance the evaluation of the compliance objective. • The safety of the HOV lane should be compared to that of the GP lanes. However, current accident reporting practice does not include referencing to either the HOV or GP lane specifically. In addition, the current accident reporting procedures should be improved so that meaningful "before-and-after" safety evaluations can be performed on the facility. • A "before-and-after" safety comparison method such as the Empirical Bayes approach which makes use of a reference group to account for the randomness inherent in the accident occurrence process should be used to determine the reduction in accidents on an HOV facility. • Only the quantifiable benefits and costs should be considered in the benefitcost analysis. However, criteria that is difficult to quantify should still be considered in the economic justification of an HOV facility. • A probabilistic economic evaluation which accounts for the uncertainty of the cost-effectiveness of an HOV facility should be performed, if possible.  Chapter 7: Conclusion  Page 127  BIBLIOGRAPHY Bacquie, Ft., and Bahar, G., (1996). "Toronto Arterial High Occupancy Vehicle (HOV) Lanes: Effectiveness, Operational Safety and Enforcement", Institute of Transportation Engineers (ITE) 1996 Compendium of Technical Papers. Bein, P., Litman, T., and Johnson, C , (1994). "Unit Costs of Environmental Impact Report", Ministry of Transportation and Highways, Planning Services Branch, Research and Development Section. Brown, W., and Jacobson, E., (1996). "HOV Evaluation and Monitoring Phase III", Washington State Transportation Center (TRAC), University of Washington, Seattle, Washington. CAA - Canadian Automobile Association, (1998). "1996 - 1997 Driving Costs", Canadian Automobile Association, Ottawa, Ontario. Fuhs, C , (1993). "Preferential Treatment for High-Occupancy Vehicles", National Cooperative Highway Research Program (NCHRP) Synthesis of Highway Practice 185,  Transportation Research Board, National Research Council, Washington, DC. Golob, T. and Recker, W. (1988). "Safety Impacts Associated with Installation of HOV (High Occupancy Vehicle) Lanes", Institute of Transportation Studies, University of California, Irvine, California. Henk, R., Christiansen, D., and Lomax, T., (1991). "Simplified Approach for Estimating the Cost-Effectiveness of HOV Facilities", Transportation Research Record 1299,  Transportation Research Board, National Research Council, Washington, DC. Henk, R., Morris, D. and Christiansen, D., (1995). "An Evaluation of High-Occupancy Vehicle Lanes in Texas, 1994", Texas Transportation Institute, College Station, Texas. MoTH, (1993). "Benefit/Cost Evaluation of Barnet-Hastings Transportation and Highways, South Coast Region.  Project", Ministry of  MoTH, (1995). "Value of Life Study: Draft Report and Recommendations", Ministry of Transportation and Highways, Highway Safety Branch. MoTH , (1997). "Barnet/Hastings People Moving Project Monitoring and Evaluation Program After-Implementation Interim Report", Ministry of Transportation and Highways, South Coast Region. 1  MoTH , (1997). "Barnet/Hastings People Moving Project Completion Report", Ministry of Transportation and Highways, South Coast Region. 2  MoTH , (1997). "Round IV Service Area #6 - Year 2 Maintenance Services Contract", Ministry of Transportation and Highways, South Coast Region. ITE - Institute of Transportation Engineers, (1982). "Traffic Studies", Transportation and 3  Traffic Engineering Handbook: Second Edition, Prentice-Hall Inc., New Jersey.  Page 128  Pint, A., Zimmer, C , Kern, J. and Palek, L , (1995). "Evaluation of Minnesota I-394 HighOccupancy-Vehicle Transportation System", Transportation Research Record 1494, Transportation Research Board, National Research Council, Washington, DC. RSMI - Resource Systems Management International, Objectives" (Paper PLN-1), HOV Implementation Project. 1  (1997). "Planning HOV  RSMI - Resource Systems Management International, (1997). "Monitoring and Evaluation Benchmarking" (Paper MEV-1), HOV Implementation Project. 2  RSMI - Resource Systems Management International, (1997). "Monitoring and Evaluation Program" (Paper MEV-2), HOV Implementation Project. 3  Sayed ,T., (1996) "How to Do Before and After Safety Studies", Course Notes 582, The University of British Columbia, Civil Engineering Department, Vancouver, BC. 1  Sayed ,T., (1996) "How to Measure Safety", Course Notes 582, The University of British Columbia, Civil Engineering Department, Vancouver, BC. 2  SRF - Strgar-Roscoe-Fausch, Inc., (1993). "I-394 Phase III Evaluation Interim Report", Minnesota Department of Transportation (Mn/DOT), Plymouth, Minnesota. Turnbull , K., (1992). "An Assessment of High-Occupancy Vehicle Facilities in North America: Executive Report", Texas Transportation Institute, College Station, Texas. 1  Turnbull , K., (1992). "High-Occupancy Vehicle Project Case Studies: Historical Trends and Project Experiences", Texas Transportation Institute, College Station, Texas. 2  Turnbull, K., Henk, R., and Christiansen, D., (1991). "Suggested Procedures for Evaluating the Effectiveness of Freeway HOV Facilities", Texas Transportation Institute, College Station, Texas. Turner, S., Naples, C. Ill, and Henk, R., (1994). 'Travel Time Reliability of HOV Facilities", Institute of Transportation Engineers (ITE) 1994 Compendium of Technical Papers. Ulberg, C , and Jacobson, E., (1993). "HOV Lane Enforcement Evaluation", Washington State Transportation Center (TRAC), University of Washington, Seattle, Washington. Ulberg, C , and Jacobson, K. (1988). "Evaluation of the Cost-Effectiveness of HOV Lanes", Transportation Research Record 1181, Transportation Research Board, National Research Council, Washington, DC. Waters, W., (1992). 'The Value of Time Savings for the Economic Evaluation of Highway Investments in British Columbia", Ministry of Transportation and Highways, Planning Services Branch.  Page 129  APPENDIX A - STATISTICAL ANALYSIS  INTRODUCTION This appendix outlines the statistical procedures for collecting and analysing the data as described in the HOV monitoring and evaluation framework. The following statistical tests or methodologies are presented: • • • • • •  Sample size estimation Statistical reliability 95% confidence limits Two sample Mest Paired sample Mest F-test  SAMPLE SIZE ESTIMATION Travel time runs should be performed using the 'average' car technique as described in the 'ITE Transportation and Traffic Engineering Handbook' (ITE, 1982). The travel times  of test vehicles as they pass through the middle of designated intersections and at the beginning and terminal points of the HOV facility should be recorded. In addition, traffic conditions and critical events that should also be recorded include: weather conditions; locations and time of incidents; stalled or parked vehicles; etc.. To determine an estimate of the sample size required for a travel time survey, the level of confidence is firstly selected. As presented in the framework, a 95% level of confidence is recommended for all HOV monitoring and evaluation purposes. The desired level of accuracy of the travel time runs is then selected. For before-and-after studies, the  'ITE Transportation and Traffic Engineering Handbook' (ITE,  1982)  suggests a range of specified permitted errors between ± 2 km/h to ± 5km/h. Finally, an estimate of the standard deviation of the travel time runs to be performed is selected. The standard deviation should be estimated based on the most recent available travel time data. When all of the assumptions have been made, the following equation is used to determine an estimate of the required sample size: estimate of the sample size required =  Zot 1  2  lO  d  where  d o  standard normal statistic corresponding to the 1-a level of confidence 1.96 for a 95% level of confidence acceptable margin of error of mean value (km/h) 2.0 to 5.0 km/h = estimate of the standard deviation (km/h)  APPENDIX A - Statistical Analysis  Page 130  EXAMPLE: d a  = 1.96 = 2.0 to 5.0 km/h = 7.7km/h  Acceptable Margin of Error 2.0 km/h 3.5 km/h 5.0 km/h  Estimate of Required Sample Size 57 19 10  It should be noted that the number of travel time runs required is a function of the estimate of standard deviation estimate (7.7 km/h in the example). To understand the importance of selecting a meaningful and representative standard deviation, a sensitivity analysis on the estimate of the required sample size is performed by varying the standard deviation. Standard Deviation km/h  Acceptable Margin of Error 2.0 km/h  5.0 6.0 7.0 8.0 9.0 10.0  3.5 km/h  5.0 km/h  8 12 16 21 26 32  4 6 8 10 13 16  25 36 48 62 78 97  The sensitivity analysis shows that the estimate of the required sample size increases as: • the estimate of the standard deviation increases • the acceptable margin of error decreases Therefore, a conservative acceptable margin of error of 5km/h per hour is recommended to estimate the required sample size of a travel time survey.  STATISTICAL RELIABILITY The statistical reliability of a sample of data is determined by quantifying the sample's margin of error as follows: margin of error = — where tj2  = test-statistic at the 95% confidence level (to.025 or  n s  = number of travel times runs performed = the standard deviation of the travel speeds (km/h)  APPENDIX A - Statistical Analysis  Zn.025)  Page 131  The test-statistic is obtained from either the z-distribution (n>30) or the t-distribution (n<30), depending on the sample size. In addition, it is recommended that the margin of error be presented as a percentage of the mean as follows: s*t margin of error = ( — ^  ) x 100%  mean  Travel Time Data For travel time data, the margin of error should be less than the specified permitted error of 5 km/h as  recommended in the 'ITE Transportation and Traffic Engineering  Handbook' (ITE, 1982). Travel time data with margins of error less than 5km/h are considered to be statistically reliable. EXAMPLE: t. n  margin of error  = 2.110 =18 s = 7.7 km/h mean = 47.2 km/h  = 3.8 km/h = 8% of the mean  Therefore, because the margin of error is less than 5.0 km/h, the sample of travel time data is considered to be statistically reliable.  AVO Data For AVO data, only the margin of error expressed as a percentage of the mean is calculated and discussed. EXAMPLE: Zjz n  margin of error  =1.96 =50 s = 0.049 mean = 1.230  =1% of the mean  95% CONFIDENCE LIMITS 95% confidence limits are determined with the following formula: —  s-  Vn where  x  = sample mean  t^2 = test-statistic at the 95% confidence level (Ws or  s-  Z0.025)  = standard deviation of the sample  A P P E N D I X A - Statistical Analysis  Page 132  n = sample size The test-statistic is obtained from either the z-distribution (n>30) or the t-distribution (n<30), depending on the sample size.  Large Samples (n>30) Samples sizes greater than 30 are assumed to have a distribution that is approximately normal and a population standard deviation that is known and fixed. Therefore, the zdistribution is used to establish the 95% confidence limits for sample sizes greater than 30. EXAMPLE:  The mean peak period AVO was calculated to be 1.230 with a standard deviation of 0.049. The sample size was 50. x Zjz  s-  = 1.230 = 1.96 = 0.049  X  n  =50  Therefore, the 95% confidence interval of the mean peak period AVO is: 1.23 +/- 0.01 1.22 to 1.24  Small Samples (n<30) To determine the 95% confidence limits for samples less than 30 in size, it is necessary to assume that the sample is approximately normally distributed. However, the population mean is unknown and the sampling distribution is, therefore, based on the sample's standard deviation as opposed to the population's. Therefore, the f-distribution with n-1 degrees of freedom is used to establish the 95% confidence limits for sample sizes of less than 30. EXAMPLE:  The mean peak hour travel time savings was calculated to be 4.31 minutes with a standard deviation of 1.19 minutes. The sample size was 8. x tj2  5-  = 4.31 minutes = to.025 for a 95% confidence level and 7 degrees of freedom = 2.365 =1.19 minutes  x  n  =8  Therefore, the 95% confidence interval of the mean peak hour travel time savings is: 4.31 minutes +/-1.0 minutes 3.31 to 5.31 minutes APPENDIX A - Statistical Analysis  Page 133  STATISTICAL SIGNIFICANCE Statistical hypothesis testing is used to determine the statistical significance of a measure of effectiveness (MOE). The construction of a hypothesis test requires the formation of a null hypothesis, H , and an alternative hypothesis, H . The null hypothesis should always be defined as the hypothesis which, it is hoped, will be rejected. The alternative hypothesis can be defined in a number of ways. The following examples illustrate the use of hypothesis testing to determine the statistical significance of MOEs. For HOV evaluations, all of the hypothesis testing is performed at the 95% confidence level. 0  A  Two Sample t-test (large samples)  z-statistic -  ,—1-  Ik where X X OA OB n n A  B  A  B  + —  n  B  = mean of sample A = mean of sample B = standard deviation of sample A = standard deviation of sample B = sample size of A = sample size of B Null Hypothesis H : u -u =0 0  A  B  Alternative Hypothesis H :u -u 0 H :u -u > 0 H :u -u < 0 A  A  B  A  A  B  A  A  Rejection Region | z-statistic | > z z-statistic > z z-statistic < z J2  a  B  n  EXAMPLE:  The difference between two mean AVOs is tested using the two-sample t-test and the hypothesis that the "after" AVO is greater than the "before" AVO (H : u - u > 0). The "before" AVO was 1.230 with a standard deviation of 0.049 and a sample size of 50. The "after" AVO was 1.293 with a standard deviation of 0.065 and a sample size of 50. A  A  B  The z-statistic is calculated to be 5.5 and the z is 1.645. Therefore, the difference in the "before-and-after" AVOs is considered to be statistically significant. o  A P P E N D I X A - Statistical Analysis  Page 134  Two Sample t-test (small samples) Before we can compute the t-statistic for the difference in means for small samples, a pooled estimate of the population standard deviation (s ) is required: p  c  2  (n -\)s +(n -\)s 2  =  A  2  A  B  A  where s s n n A B A B  B  n +n -2  P  B  = standard deviation of sample A = standard deviation of sample B = sample size of A = sample size of B t-statistic =  (n - n + 2) degrees of freedom A  where X X s n n A  B  p  A B  B  = mean of sample A = mean of sample B = pooled standard deviation = sample size of A = sample size of B Null Hypothesis H : u -u = 0 0  Alternative Hypothesis H :u -u *0 H :u -u > 0 H :u -u < 0 A  A  B  A  A  B  A  A  B  A  B  Rejection Region | t-statistic | > t t-statistic > t t-statistic < t a/2  0  EXAMPLE:  The difference between two mean AVOs is tested using the two-sample t-test and the hypothesis that the "after" AVO is greater than the "before" AVO (H : u - u > 0). The "before" AVO was 1.248 with a standard deviation of 0.060 and a sample size of 20. The "after" AVO was 1.323 with a standard deviation of 0.046 and a sample size of 20. A  A  B  The pooled variance is calculated to be 0.053. The t-statistic is calculated to be 4.4 and the t is 1.645 for 38 degrees of freedom. Therefore, the difference in the "before-andafter" AVOs is considered to be statistically significant. a  A P P E N D I X A - Statistical Analysis  Page 135  Paired Sample t-test A paired sample Mest is performed when sample data has been collected in pairs. The mean difference between two samples is determined to be statistically significant if the tstatistic obtained is greater than a t-critical for a 95% level of confidence and n-1 degrees of freedom.  (n-1) degrees of freedom where D = mean of the differences ( X - X ) S = standard deviation of the differences n = sample size (number of paired samples or differences) A  B  D  Null Hypothesis H: u =0 0  d  Alternative Hypothesis H :u *0 H :u >0 H :u <0 A  d  A  d  A  d  Rejection Region | t-statistic | > \j t-statistic > t t-statistic < t 2  a  a  EXAMPLE (Travel Time Difference Data):  The mean of the peak hour travel time differences was calculated to be 4.31 minutes with a standard deviation of 1.19 minutes. The sample size was 8. D S n  =4.31 minutes =1.19 minutes =8  D  H  A  t-statistic  :u >0  t-statistic  d  =10.244  . >t > to.os @ 95% with 7 d.o.f. > 1.895 a  Therefore, because the t-statistic is greater than the t-critical, the travel time difference between travelling in the GP and HOV lanes is significant at the 95% confidence level. EXAMPLE ("Before-and-after" Safety Data):  The difference between the mean "before" accident rate and mean "after" accident rate is also determined with a paired sample test. However, a pooled standard deviation is used as follows: Sn = 2  SB + S A " 2 2  APPENDIX A - Statistical Analysis  2  [ X (XR- XRrmfianM XA~ XAImfianO ] n-1  Page 136  F-test An F-test is performed to compare two sample standard deviations. The difference in the two sample standard deviations is determined by examining the ratio of sample variances (the F-statistic). If the F-statistic is greater than an F-critical for a 95 % level of confidence and respective degrees of freedom, then the difference in the standard deviation is considered to be significant. F-statistics in statistical tables are always greater than one in value, therefore, the sample standard deviation of A should be greater than sample B's for performing such a test. F-statistic = —4(n -1) degrees of freedom in the numerator (n -1) degrees of freedom in the denominator A  B  where  c o n n  = standard deviation of sample A standard deviation of sample B sample size of A sample size of B  A B A B  Null Hypothesis H : a\ = a £ 0  Alternative Hypothesis H : a\ >(ji A  Rejection Region F-statistic > F (n -1, n o  A  B  1)  EXAMPLE: The standard deviation of the HOV lane was calculated to be 2.95 km/h while that of the adjacent GP lane was calculated to be 5.17 km/h. The sample size of the travel time reliability survey was 12 for both lanes.  on n  B  A  B  = 5.17 km/h = 2.95 km/h = 12 = 12 A  >  (  J  B  F-statistic  = 3.06  F-statistic  > F (n -1, n -1) >F„.o5 (11,11) >2.82 a  A  B  Therefore, because the F-statistic is greater than the F-critical, the difference in the standard deviations is considered to be statistically significant at the 95% confidence level. In other words, the HOV lane is more reliable than the GP lane.  A P P E N D I X A - Statistical Analysis  Page 137  PER-LANE EFFICIENCY MARCH, 1996 WESTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  St. Johns Street  Location  Williams  Total # of lanes Total Person Volume time (hrs)  Barnet Highway  Hastings Street  Reed  Bayview  Willingdon  Lilloet  1 1814 1  1  2  1939 1  3278 1  3 4592 1  GP # of lanes person volume speed (km/h)  3 3294 18  1 1814 44  1 1939 44  2  3  3278 28  4592 28  Per-Lane Efficiency  20  80  85  46  43  80  85  46  43  HOV # of lanes person volume speed (km/h) Per-Lane Efficiency PEAK HOUR (7:30 TO 8:30 AM) Location Per-Lane Efficiency  20  Section Per-Ijine Efficiency  20  APPENDIX B - Efficiency Data  83  44  Page 1  PER-LANE EFFICIENCY OCTOBER, 1996 WESTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  St. Johns Street  Location  Williams  Reed  Bayview  Willingdon  3 3519 1  2 3091 1  2 2826 1  3 3693 1  1  Total # of lanes Total Person Volume time (firs)  Barnet Highway  Hastings Street  GP •# of lanes person volume speed (km/h)  1941 53  1 1826 53  1505 30  Per-Lane Efficiency  103  97  23  Lillooet*  2  HOV # of lanes person volume speed (km/h)  1 1012 40  1 1150 75  1 1000 75  1 2188 35  Per-Lane Efficiency  40  86  75  77  95  86  41  PEAK HOUR (7:30 TO 8:30 AM) Location Per-Lane Efficiency  51  Section Per-Lane Efficiency  51  90  41  NOTES: 1. Volumes at Lillooet not included due to an accident on day of survey  A P P E N D I X B - Efficiency Data  Page 1  PER-LANE EFFICIENCY OCTOBER, 1997 WESTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  St. Johns Street  Location  Williams  Total # of lanes Total Person Volume time (hrs)  Barnet Highway  Hastings Street Willingdon  Reed  Bayview  2 3117 1  2  3  3  3312 1  4028 1  4397 1  1  2  2  2086 24  2266 24  Lilloet  GP # of lanes person volume speed (km/h)  1 1527 40  1468 40  Per-Lane Efficiency  61  59  25  27  1 1942 29  1 2131 29  HOV # of lanes person volume speed (km/h)  1 1220 41  1 1590 72  1 1844 72  Per-Lane Efficiency  50  114  133  56  62  88  96  35  39  PEAK HOUR (7:30 TO 8:30 AM) Location Per-Lane Efficiency  54  Section Per-Lane Efficiency  54  A P P E N D I X B - Efficiency Data  92  37  Page 1  PER-LANE EFFICIENCY MARCH, 1996 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location Total # of lanes Total Person Volume time (hrs)  Hastings Street Lillooet  Carleton*  3 4140 1  Barnet Highway  Willingdon  Bayview  Union  2  1  2889 1  2112 1  1 1995 1  2  1 2112 35  1 1995 35  GP # of lanes person volume speed (km/h)  3 4140 23  2889 23  Per-Lane Efficiency  32  33  74  70  33  74  70  HOV # of lanes person volume speed (km/h) Per-Lane Efficiency PEAK HOUR (5:00 TO 6:00 PM) Location Per-Lane Efficiency Section Per-Lane Efficiency  32  32  72  NOTES: 1. no data collected at Carleton in March, 1996  A P P E N D I X B - Efficiency Data  Page 141  PER-LANE EFFICIENCY OCTOBER, 1996 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location Total # of lanes Total Person Volume time (hrs) GP # of lanes person volume speed (km/h) Per-Lane Efficiency HOV # of lanes person volume speed (km/h) Per-Lane Efficiency  Hastings Street Lillooet  Carleton*  Barnet Highway  Willingdon  Bayview  Union  3  3  2  2  4306  3910  2739  2868  1  1  1  1  2  2  1  1  2508  2392  1648  1932  27  27  63  63  34  32  104  122  1  1  1  1  1798  1518  1091  936  31  31  70  70  56  47  76  66  37  90  94  PEAK HOUR (5:00 TO 6:00 PM) Location Per-Lane Efficiency Section Per-Lane Efficiency  41  39  92  NOTES: 1. no data collected at Carleton in October, 1996  APPENDIX B - Efficiency Data  Page 142  PER-LANE EFFICIENCY OCTOBER, 1997 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location  Barnet Highway  Hastings Street Lillooet  Carleton  Willingdon  Bayview  Union  3  3  3  2  2  3989 1  4602 1  4504 1  3196 1  3356 1  GP # of lanes person volume speed (km/h)  2 2025 28  2 2448 28  2 2475 28  1 1984 59  1 2084 59  Per-Lane Efficiency  28  34  35  117  123  HOV # of lanes person volume speed (km/h)  1 1964 33  1 2154 33  1 2029 33  1 1212 67  1 1272 67  Per-Lane Efficiency  65  71  67  81  85  47  45  99  104  Total # of lanes Total Person Volume time (hrs)  PEAK HOUR (5:00 TO 6:00 PM) Location Per-Lane Efficiency Section Per-Lane Efficiency  A P P E N D I X B - Efficiency Data  41  44  102  Page 143  A VERAGE VEHICLE OCCUPANCY (A VP) MARCH, 1996 WESTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location  St. Johns Street Williams  6:00-6:15 6:15-6:30 6:30-6:45 6:45-7:00 7:00-7:15 7:15-7:30 7:30-7:45 7:45-8:00 8:00-8:15 8:15-8:30  Clark  1.229 1.134 1.223 1.229 1.233 1.246 1.237 1.297 1.375 1.367  Barnet Highway  Hastings Street  Reed  Bayview  Willingdon  Lilloet  1.185 1.185 1.205 1.240 1.261 1.236 1.176 1.178 1.217 1.171  1.188 1.202 1.181 1.216 1.275 1.231 1.241 1.190 1.183 1.175  1.195 1.183 1.171 1.198 1.288 1.234 1.279 1.236 1.291 1.256  1.203 1.208 1.233 1.197 1.293 1.253 1.267 1.284 1.280 1.260  1.205 0.032  1.208 0.032  1.233 0.045  1.248 0.035  PEAK PERIOD (6:00 TO 8:30 AM) Locution Mean SD  1.257 0.072  Section A lean SD  1.257 0.072  Corridor A/can SD  1.207 0.031  1.240 0.040  1.230 0.049  PEAK HOUR (7:30 TO 8:30 AM) Location Mean SD Section Mean SD  1.319 0.065 1.319 0.065  Corridor Mean SD  1.186 0.021  1.197 0.030  1.191 0.025  1.266 0.024  1.273 0.011  1.269 0.018  1.248 0.060  NOTES: 1. no data collected at Carleton in March, 1996 2. SD = Standard Deviation  APPENDIX C - AVO Data  Page 144  A VERAGE VEHICLE OCCUPANCY (A VP) OCTOBER, 1996 WESTBPUND Corridor Section Location 6:00-6:15 6:15-6:30 6:30-6:45 6:45-7:00 7:00-7:15 7:15-7:30 7:30-7:45 7:45-8:00 8:00-8:15 8:15-8:30  Barnet/Hastings H O V Corridor St. Johns Street  Barnet Highway  Hastings Street  Williams  Clark  Reed  Bayview  Willingdon  Lilloet  1.220 1.145 1.198 1.194 1.206 1.233 1.258 1.284 1.367 1.329  1.476 1.394 1.388 1.371 1.455 1.541 1.457 1.399 1.380 1.311  1.368 1.348 1.336 1.413 1.386 1.401 1.383 1.372 1.335 1.385  1.190 1.288 1.288 1.270 1.351 1.340 1.341 1.283 1.298 1.256  1.211 1.290 1.229 1.266 1.265 1.259 1.377 1.314 1.370 1.382  1.245 1.229 1.240 1.244 1.281 1.255 1.279 1.303 1.260 1.289  1.373 0.026  1.291 0.048  1.296 0.062  1.263 0.024  PEAK PERIPD (6:00 TP 8:30 AM) Location Mean SD Section Mean SD  1.243 0.067  1.417 0.065  1.243 0.067  1.332 0.056  Corridor Mean SI)  1.279 0.049  1.293 0.065  PEAKHPUR (7:30 TP 8:30 AM) Location Mean SD Section Mean SD Corridor Mean SD  1.310 0.048  1.387 0.060  1.310 0.048  1.369 0.023  1.295 0.036  1.361 0.032  1.332 0.048  1.283 0.018  1.322 0.048  1.323 0.046  NOTES: 1. Clarke St. not included in averages because separate from most of GP traffic  A P P E N D I X C - A V O Data  Page 145  A VERAGE VEHICLE OCCUPANCY (A VP) MARCH, 1997 WESTBPUND Corridor Section Location 6:00-6:15 6:15-6:30 6:30-6:45 6:45-7:00 7:00-7:15 7:15-7:30 7:30-7:45 7:45-8:00 8:00-8:15 8:15-8:30  Barnet/Hastings H O V Corridor St. Johns Street  Barnet Highway  Hastings Street  Williams  Clark  Reed  Bayview  Willingdon  Lilloet  1.224 1.209 1.202 1.233 1.228 1.274 1.259 1.287 1.336 1.380  1.373 1.425 1.426 1.454 1.482 1.447 1.473 1.466 1.486 1.473  1.171 1.201 1.238 1.263 1.223 1.299 1.225 1.243 1.235 1.231  1.312 1.356 1.270 1.295 1.343 1.356 1.371 1.339 1.323 1.301  1.304 1.344 1.287 1.326 1.363 1.376 1.377 1.362 1.359 1.352  1.164 1.281 1.274 1.237 1.281 1.372 1.341 1.311 1.298 1.314  1.233 0.034  1.327 0.032  1.345 0.030  1.287 0.057  PEAK PERIPD (6:00 TP 8:30 AM) Location Mean SD Section Mean SD  1.263 0.058  1.451 0.035  1.263 0.058  1.280 0.058  Corridor Mean SD  1.316 0.053  1.291 0.059  PEAKHPUR (7:30 TP 8:30 AM) Location Mean SD Section Mean SD Corridor Mean SD  1.316 0.053  1.475 0.008  1.316 0.053  1.234 0.008  1.334 0.029  1.284 0.057  1.363 0.011  1.316 0.018 1.339 0.028  1.312 0.051  NOTES: 1. Clarke St. not included in averages because separate from most of GP traffic  A P P E N D I X C - A V O Data  Page 1  A VERAGE VEHICLE OCCUPANCY (A VP) OCTOBER, 1997 WESTBOUND Corridor Section Location 6:00-6:15 6:15-6:30 6:30-6:45 6:45-7:00 7:00-7:15 7:15-7:30 7:30-7:45 7:45-8:00 8:00-8:15 8:15-8:30  Barnet/Hastings H O V Corridor St. Johns Street  Barnet Highway  Hastings Street  Williams  Clark  Reed  Bayview  Willingdon  Lilloet  1.267 1.176 1.169 1.216 1.249 1.286 1.313 1.283 1.411 1.319  1.362 1.296 1.326 1.393 1.508 1.426 1.493 1.520 1.498 1.352  1.271 1.194 1.262 1.252 1.271 1.323 1.379 1.296 1.255 1.219  1.365 1.361 1.283 1.290 1.270 1.390 1.384 1.355 1.335 1.267  1.173 1.237 1.197 1.174 1.193 1.289 1.312 1.265 1.257 1.285  1.290 1.190 1.184 1.197 1.233 1.207 1.257 1.251 1.261 1.288  1.272 0.052  1.330 0.048  1.238 0.051  1.236 0.040  PEAK PERIOD (6:00 TO 8:30 AM) lMcation Mean SD Section Mean SO  1.269 0.072  1.417 0.083  1.269 0.072  Corridor Mean SI)  1.301 0.057  1.237 0.044  1.269 0.062  PEAK HOUR (7:30 TO 8:30 AM) Location Mean SD Section Mean SD Corridor Mean SD  1.332 0.055  1.466 0.077  1.332 0.055  1.287 0.069  1.335 0.050  1.311 0.061  1.280 0.025  1.264 0.016  1.272 0.021  1.300 0.051  NOTES: 1. Clarke St. not included in averages because separate from most of GP traffic  APPENDIX C - AVO Data  Page 147  A VERAGE VEHICLE OCCUPANCY (A VP) MARCH, 1996 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location 3:30-3:45 3:45-4:00 4:00-4:15 4:15-4:30 4:30-4:45 4:45-5:00 5:00-5:15 5:15-5:30 5:30-5:45 5:45-6:00  Hastings Street Lillooet  Carleton  1.360 1.293 1.316 1.306 1.344 1.386 1.365 1.380 1.385 1.340  Barnet Highway  Willingdon  Bayview  Union  1.260 1.232 1.273 1.215 1.226 1.288 1.309 1.327 1.398 1.345  1.199 1.292 1.235 1.248 1.210 1.254 1.308 1.285 1.299 1.341  1.257 1.284 1.239 1.280 1.225 1.268 1.203 1.309 1.209 1.264  1.287 0.058  1.267 0.045  1.254 0.034  PEAK PERIPD (3:30 TP 6:00 PM) Location Mean SD  1.348 0.034  Sect ion Mean SD  1.317 0.056  Corridor Mean SD  1.260 0.040 1.289 0.056  PEAK HOUR (5:00 TO 6:00 PM) Location Mean SD Section Mean SD  1.368 0.020  1.345 0.038 1.356 0.031  Corridor Mean SD  1.308 0.024  1.246 0.050 1.277 0.049  1.317 0.057  NOTES: 1. no data collected at Carleton in March, 1996 2. SD = Standard Deviation  A P P E N D I X C - A V O Data  Page 148  A VERAGE VEHICLE OCCUPANCY (A VP) OCTOBER, 1996 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  Hastings Street  Location  Lillooet  3:30-3:45  1.447 1.399 1.445 1.337 1.431 1.441 1.423 1.434 1.413 1.438  3:45-4:00 4:00-4:15 4:15-4:30 4:30-4:45 4:45-5:00 5:00-5:15 5:15-5:30 5:30-5:45 5:45-6:00  Carleton  Barnet Highway  Willingdon  Bayview  Union  1.310 1.320 1.301 1.311 1.315 1.307 1.353 1.321 1.375 1.290  1.260 1.222 1.259 1.266 1.324 1.291 1.298 1.268 1.239 1.271  1.249 1.240 1.237 1.300 1.287 1.268 1.271 1.302 1.283 1.274  1.320 0.025  1.270 0.029  1.271 0.023  PEAK PERIOD (3:30 TO 6:00 PM) location Mean SD  1.421 0.033  Section Mean SD  1.371 0.059  Corridor Mean SD  1.270 0.026 1.321 0.068  PEAK HOUR (5:00 TO 6:00 PM) Location Mean SD Section Mean SD  1.427 0.011  1.335 0.037 1.381 0.055  Corridor Mean SD  1.269 0.024  1.283 0.014 1.276 0.020  1.328 0.068  NOTES: 1. no data collected at Carleton in October, 1996 2. SD = Standard Deviation  A P P E N D I X C - A V O Data  Page 149  A VERAGE VEHICLE OCCUPANCY (A VP) MARCH, 1997 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  Hastings Street  Location  Lillooet  3:30-3:45  1.299 1.356 1.289 1.318 1.283 1.301 1.388 1.417 1.367 1.404  3:45-4:00 4:00-4:15 4:15-4:30 4:30-4:45 4:45-5:00 5:00-5:15 5:15-5:30 5:30-5:45 5:45-6:00  Carleton  Barnet Highway  Willingdon  Bayview  Union  1.283 1.300 1.266 1.231 1.324 1.299 1.285 1.317 1.280 1.333  1.343 1.329 1.324 1.356 1.389 1.373 1.430 1.434 1.477 1.507  1.285 1.272 1.277 1.284 1.306 1.293 1.297 1.301 1.327 1.272  1.292 0.030  1.396 0.063  1.291 0.017  PEAK PERIOD (3:30 TO 6:00 PM) Location Mean SD  1.342 0.050  Section Mean SD  1.317 0.048  Corridor Mean SD  1.344 0.070 1.330 0.061  PEAK HOUR (5:00 TO 6:00 PM) Location Mean SD Section Mean SD  1.394 0.022  1.304 0.025 1.349 0.053  Corridor Mean SD  1.462 0.037  1.299 0.023 1.381 0.091  1.365 0.074  NOTES: 1. no data collected at Carleton in March, 1997 2. SD = Standard Deviation  A P P E N D I X C - A V O Data  Page 150  A VERA GE VEHICLE OCCUPANCY (A VP) OCTOBER, 1997 EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  Hastings Street  Location  Lillooet  Carleton  3:30-3:45 3:45-4:00 4:00-4:15 4:15-4:30 4:30-4:45 4:45-5:00 5:00-5:15 5:15-5:30 5:30-5:45 5:45-6:00  1.311 1.314 1.293 1.311 1.338 1.374 1.364 1.383 1.390 1.367  1.352 1.346 1.347 1.360 1.383 1.425 1.474 1.430 1.450 1.380  1.395 0.047  Barnet Highway  Willingdon  Bayview  Union  1.317 1.307 1.299 1.328 1.329 1.361 1.368 . 1.338 1.379 1.362  1.260 1.283 1.277 1.328 1.273 1.283 1.300 1.282 1.290 1.323  1.295 1.257 1.255 1.266 1.302 1.256 1.343 L287 1.264 1.290  1.339 0.027  1.290 0.021  1.282 0.028  PEAK PERIOD (3:30 TO 6:00 PM) location Mean SD  1.345 0.035  Section Mean SD  1.359 0.044  ( \yrndor Mean SD  1.286 0.025 1.330 0.052  PEAK HOUR (5:00 TO 6:00 PM) Location Mean SD Section Mean SD Corridor Mean SD  1.376 0.013  1.434 0.040  1.362 0.017  1.390 0.040  1.299 0.018  1.296 0.033 1.297 0.025  1.353 0.058  NOTES: 1. SD = Standard Deviation  A P P E N D I X C - A V O Data  Page 151  AVERAGE VEHICLE OCCUPANCY (AVO) CONFIDENCE INTERVALS MARCH, 1996 TO OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) AVO Standard Sample t-critical (Mean) Deviation Size (t.025) March 1996 October, 1996 March, 1997 October, 1997  1.230 1.293 1.291 1.269  0.049 0.065 0.059 0.062  50 50 50 50  1.960 1.960 1.960 1.960  CONFIDENCE INTERVALS 1.23 1.29 1.29 1.27  +/+/+/+/-  0.01 0.02 0.02 0.02  1.22 1.27 1.27 1.25  to to to to .  1.24 1.31 1.31 1.29  to to to to  1.28 1.34 1.34 1.32  to to to to  1.31 1.34 1.35 1.34  to to to to  1.35 1.36 1.40 1.38  PEAK HOUR (7:30 TO 8:30 AM) AVO Standard Sample t-critical (Mean) Deviation Size (t.ozs) March, 1996 October, 1996 March, 1997 October, 1997  1.248 1.323 1.312 1.300  0.060 0.046 0.051 0.051  20 20 20 20  2.093 2.093 2.093 2.093  CONFIDENCE INTERVALS 1.25 1.32 1.31 1.30  +/+/+/+/-  0.03 0.02 0.02 0.02  1.22 1.30 1.29 1.28  EASTBOUND PEAK PERIOD (3:30 to 6:00 PM) CONFIDENCE INTERVALS  AVO Standard Sample t-critical (Mean) Deviation Size (t.025) March, 1996 October, 1996 March 1997 October, 1997  1.289 1.321 1.330 1.330  0.056 0.068 0.061 0.052  40 40 40 50  1.960 1.960 1.960 1.960  1.29 1.32 1.33 1.33  0.02 0.02 +/- 0.02 +/- 0.01  +/+/-  1.27 1.30 1.31 1.32  PEAK HOUR (5:00 to 6:00 PM) AVO Standard Sample t-critical (Mean) Deviation Size (t.025) March, 1996 October, 1996 March, 1997 October, 1997  A P P E N D I X C - A V O Data  1.317 1.328 1.365 1.353  0.057 0.068 0.074 0.058  16 16 16 20  2.131 2.131 2.131 2.093  CONFIDENCE INTERVALS 1.32 1.33 1.37 1.35  +/+/+/+/-  0.03 0.04 0.04 0.03  1.29 1.29 1.33 1.33  Page 152  A VERAGE VEHICLE OCCUPANCY (A VP) MARGIN OF ERRORS MARCH, 1996 TO OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) AVO Standard Sample t-critical (Mean) Deviation Size (t.02S> March, 1996 October, 1996 March, 1997 October, 1997  1.230 1.293 1.291 1.269  0.049 0.065 0.059 0.062  50 50 50 50  1.960 1.960 1.960 1.960  MARGIN OF ERROR  MARGIN OF ERROR AS A % OF THE MEAN  0.01 0.02 0.02 0.02  1% 1% 1% 1%  MARGIN OF ERROR  MARGIN OF ERROR AS A % OF THE MEAN  o~ol 0.02 0.02 0.02  2% 2% 2% 2%  MARGIN OF ERROR  MARGIN OF ERROR AS A % OF THE MEAN  0.02 0.02 0.02 0.01  1% 2% 1% 1%  MARGIN OF ERROR  MARGIN OF ERROR AS A % OF THE MEAN  0.03 0.04 0.04 0.03  2% 3% 3% 2%  PEAK HOUR (7:30 TO 8:30 AM) AVO Standard Sample t-critical (Mean) Deviation Size (tins) March, 1996 October, 1996 March, 1997 October, 1997  1.248 1.323 1.312 1.300  0.060 0.046 0.051 0.051  20 20 20 20  2.093 2.093 2.093 2.093  EASTBOUND PEAK PERIOD (3:30 to 6:00 PM) AVO Standard Sample t-critical (Mean) Deviation Size (t.025> March. 1996 October. 1906 March. 1997 October. 1997  1.289 1.321 1.330 1.330  0.056 0.068 0.061 0.052  40 40 40 50  1.960 1.960 1.960 1.960  PEAK HOUR (5:00 to 6:00 PM) AVO Standard Sample t-critical (Mean) Deviation Size (t.025) March, 1996 October, 1996 March, 1997 October, 1997  APPENDIX C - AVO Data  1.317 1.328 1.365 1.353  0.057 0.068 0.074 0.058  16 16 16 20  2.131 2.131 2.131 2.093  Page 153  AVERAGE VEHICLE OCCUPANCY (AVO) STATISTICAL SIGNIFICANCE MARCH, 1996 TO OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) AVO Standard Sample (Mean) Deviation Size March, 1996 October, 1996 March, 1997 October, 1997 MINIMUM*  test t-critical statistic (t.os) A B  1.230 1.293 1.291 1.269  0.049 0.065 0.059 0.062  50 50 50 50  5.5 5.6 3.5  1.645 1.645 1.645  1.25  0.06  50  1.7  1.645  STATISTICALLY SIGNIFICANT A>B ? yes yes yes  PEAK HOUR (7:30 TO 8:30 AM) AVO Standard Sample Pooled test t-critical (Mean) Deviation Size Variance statistic (t.os) A B 1.248 0.060 20 0.053 1.323 0.046 20 4.4 1.645 0.056 1.312 0.051 20 3.6 1.645 1.300 0.056 0.051 20 3.0 1.645  March, 1996 October, 1996 March, 1997 October, 1997 MINIMUM*  H"2TT  0.05  |  20  |[T056 ||  17  STATISTICALLY SIGNIFICANT A>B ? yes yes yes  | 1.645 |  NOTES: 1. MINIMUM based on an average standard deviation & sample size  APPENDIX C - AVO Data  Page 154  A VERAGE VEHICLE OCCUPANCY (A VP) STA TISTICAL SIGNIFICANCE MARCH, 1996 TO OCTOBER, 1997 EASTBOUND PEAK PERIOD (3:30 to 6:00 PM) AVO (Mean)  Standard Sample Deviation Size  test t-critical statistic (t.os) A B  March 1996 October, 1996 March, 1997 October, 1997  1.289 1.321 1.330 1.330  0.056 0.068 0.061 0.052  40 40 40 50  2.3 3.1 3.6  1.645 1.645 1.645  MINIMUM*  1.31  0.06  43  1.7  1.645  STATISTICALLY SIGNIFICANT A>B ? yes yes yes  PEAK HOUR (5:00 to 6:00 PM) AVO (Mean) 1.317 1.328 1.365 1.353  March, 1996 October, 1996 March, 1997 October, 1997 MINIMUM*  I  1-35 |  Standard Sample Pooled test t-critical Deviation Size Variance statistic (t.os) A B 0.057 16 1.645 0.068 16 0.063 0.5 1.645 0.074 16 0.066 2.1 1.645 0.058 20 0.058 1.9 0.06  I 17 |H)061 || ~  STATISTICALLY SIGNIFICANT A>B ? NO yes yes  J 1 645 |  M?7ESV 1. MINIMUM based on an average standard deviation & sample size  A P P E N D I X C - A V O Data  Page 155  HOV SHARE OF TOTAL TRIPS MADE MARCH, 1996 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) Location  Williams  SO Vs HOVs BUS TOTAL SOVs % HOVs % BUS % HOVSHARE  Clarke*  Reed  Bayview  Willingdon  Lillooet  3868 2365 1047  2933 1370 490  2877 1381 585  3546 1976 1809  3685 2189 3469  7280  4793  4843  7331  9343  53% 32% 14%  61% 29% 10%  59% 29% 12%  48% 27% 25%  39% 23% 37%  47%  39%  41%  52%  61%  Section  St. Johns Street  Barnet Highway  Hastings Street  SOVs % HOVs% HUS%  53% 32% 14%  60% 29% 11%  44% 25% 31%  47%  40%  56%  HOVSHARE  Corridor  Barnet/Hastings H O V Corridor  SOVs % HOVs % BUS%  52% 28% 20%  HOVSHARE  48%  NOTES: 1. no data collected at Clarke in March, 1996 2. HOVSHARE =% of HOVs + % of Bus  A P P E N D I X D - H O V Market Share Data  Page 156  HOVSHARE OF TOTAL TRIPS MADE OCTOBER, 1996 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) Location  Williams  Clarke*  Reed  Bayview  Willingdon  Lillooet*  SOVs HOVs BUS TOTAL  4212 2451 1053  1350 1761 585  3514 3689 665  3718 2824 541  3650 2739 2362  3296 2128 1971  7716  3696  7868  7083  8751  7395  SOVs % HOVs % BUS%  55% 32% 14%  37% 48% 16%  45% 47% 8%  52% 40% 8%  42% 31% 27%  45% 29% 27%  HOVSHARE  45%  63%  55%  48%  58%  55%  Section  St. Johns Street  Barnet Highway  Hastings Street  SOVs% HOVs % BUS%  55% 32% 14%  49% 43% 8%  43% 30% 27%  HOVSHARE  45%  51%  57%  Corridor  Barnet/Hastings H O V Corridor  SOVs % HOVs % BUS%  48% 36% 16%  HOVSHARE  52%  NOTES: 1. Volumes at Lilloet are low due to an accident 2. HOVSHARE =% of HOVs + % ofBus 3. Clarke location separate from most of GP traffic, therefore, not included as part of averages  A P P E N D I X D - H O V Market Share Data  Page 157  HOV SHARE OF TOTAL TRIPS MADE MARCH, 1997 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) Location  Williams  Clarke*  Reed  Bayview  Willingdon  Lillooet  SOVs HOVs BUS TOTAL  4311 2796 727  1503 2162 760  3962 2299 660  4145 3506 555  4125 3980 1140  3837 2840 2979  7834  4425  6921  8206  9245  9656  SOVs % HOVs % BVS%  55% 36% 9%  34% 49% 17%  57% 33% 10%  51% 43% 7%  45% 43% 12%  40% 29% 31%  45%  66%  43%  49%  55%  60%  HOVSHARE  Section  St. Johns Street  Barnet Highway  Hastings Street  SOVs % HOVs % BUS%  55% 36% 9% Vo  54% 38% 8%  42% 36% 22%  HOV SHARE  45%  46%  58%  Corridor  Barnet/Hastings H O V Corridor  SOVs % H()Vs% BVS%  49% 37% 14%  HOVSHARE  51%  NOTES: 1. Clarke location separate from most of GP traffic, therefore, not included as part of averages 2. HOV SHARE =% of HOVs + % of Bus  A P P E N D I X D - H O V Market Share Data  Page 158  HOV SHARE OF TOTAL TRIPS MADE OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 TO 8:30 AM) Location  Williams  Clarke*  Reed  Bayview  Willingdon  Lillooet  SOVs HOVs BUS  4630 3012 1053  1628 2059 475  4430 3040 742  3932 3365 615  4590 2660 1729  3993 2260 2678  TOTAL  8695  4162  8212  7912  8979  8931  SOVs % HOVs % BUS%  53% 35% 12%  39% 49% 11%  54% 37% 9%  50% 43% 8%  51% 30% 19%  45% 25% 30%  47%  61%  46%  50%  49%  55%  HOVSHARE  Section  St. Johns Street  Barnet Highway  Hastings Street  SOVs % HOVs % BUS%  53% 35% 12% Vo  52% 40% 8%  48% 27% 25%  HOV SHAME  47%  48%  52%  Corridor  Barnet/Hastings H O V Corridor  SOVs % HOVs % BVS%  50% 34% 16%  HOVSHARE  49%  NOTES: 1. Clarke location separate from most of GP traffic, therefore, not included as part of averages 2. HOVSHARE =% of HOVs + % of Bus  APPENDIX D - HOV Market Share Data  Page 159  HOVSHARE OF TOTAL TRIPS MADE MARCH, 1996 EASTBOUND PEAK PERIOD (3:30 TO 6:00 PM) Location  Lillooet  SOVs HOVs BUS  Willingdon  Bayview  Union  3327 2924 3491  3475 2381 1502  2829 1725 655  2778 1592 575  TOTAL  9742  7358  5209  4945  SOVs % HOVs % BUS%  34% 30% 36%  47% 32% 20%  54% 33% 13%  56% 32% 12%  66%  53%  46%  44%  HOVSHARE  Carleton*  Section  St. Johns Street  Barnet Highway  SOVs % HOVs % BUS%  41% 31% 28%  55% 33% 12%  59%  45%  HOVSHARE  Corridor  Barnet/Hastings H O V Corridor  SOVs% HOVs % BUS %  48% 32% 20%  HOVSHARE  52%  NOTES: 1. no data collected at Carleton in March, 1996 2. HOVSHARE =% of HOVs + % of Bus  A P P E N D I X D - H O V Market Share Data  Page 160  HOVSHARE OF TOTAL TRIPS MADE OCTOBER, 1996 EASTBOUND PEAK PERIOD (3:30 TO 6:00 PM) Location  Lillooet  SOVs HOVs BUS TOTAL  Carleton*  Willingdon  Bayview  Union  3272 3885 3003  4597 3792 1252  3640 2442 691  3622 2395 545  10160  9641  6773  6562  SOVs % HOVs % BUS%  32% 38% 30%  48% 39% 13%  54% 36% 10%  55% 36% 8%  HOVSHARE  68%  52%  46%  45%  Section  St. Johns Streeet  Barnet Highway  SOVs % HOVs% BUS %  40% 39% 21%  54% 36% 9%  60%  46%  HOVSHARE  Corridor  Barnet/Hastings H O V Corridor  SOVs% HOVs % BUS%  47% 38% 15%  HOVSHARE  53%  NOTES: 1. no data collected at Carleton in October, 1996 2. HOVSHARE =% of HOVs + % of Bus  APPENDIX D - HOV Market Share Data  Page 161  HOVSHARE OF TOTAL TRIPS MADE MARCH, 1997 EASTBOUND PEAK PERIOD (3:30 TO 6:00 PM) Location  Lillooet  SOVs HOVs BUS TOTAL SOVs % HOVs % BVS% HOVSHARE  Carleton*  Willingdon  Bayview  Union  3244 2983 2461  4373 3247 755  3737 3858 661  3683 2579 420  8688  8375  8256  6682  37% 34% 28%  52% 39% 9%  45% 47% 8%  55% 39% 6%  63%  48%  55%  45%  Section  St. Johns Street  Barnet Highway  SOJs % HOVs% BUS %  45% 37% 19%  50% 43% 7%  55%  50%>  HOVSHARE  Corridor  Barnet/Hastings H O V Corridor  SOVs% H()Vs% BVS%  47% 40% 13%  HOVSHARE  53%  NOTES: 1. no data collected at Carleton in March, 1997 2. HOVSHARE =% of HOVs + % ofBus  A P P E N D I X D - H O V Market Share Data  Page 162  HOVSHARE OF TOTAL TRIPS MADE OCTOBER, 1997 EASTBOUND PEAK PERIOD (3:30 TO 6:00 PM) Location  Lillooet  Carleton*  Willingdon  Bayview  Union  SOVs HOVs BUS  3643 3203 2431  3800 4149 2533  4718 4077 2056  3998 2872 532  4192 2825 459  TOTAL  9277  10482  10851  7402  7476 .  SOVs% HOVs % BliS%  39% 35% 26%  36% 40% 24%  43% 38% 19%  54% 39% 7%  56% 38% 6%  61%  64%  57%  46%  44%  HOVSHARE  Section  St. Johns Street  Barnet Highway  SOVsVo HOVs % BUSH  40% 37% 23%  55% 38% 7%  60%  45%  HOVSHARE  Corridor  Barnet/Hastings H O V Corridor  SOVs % HOVs % BUS %  46% 38% 17%  HOVSHARE  54%  NOTES: 1. HOVSHARE =% of HOVs + % of Bus  APPENDIX D - HOV Market Share Data  Page 163  TRA VEL TIME AND SPEEDS MARCH, 1996 WESTBOUND PEAK HOUR (7:30 AM TO 8:30 AM) Section  St. Johns St.  Barnet Hwy  Hastings St.  H O V Corridor  Distance (kms)  3.02  7.81  7.02  17.85  9.9 18  10.7 44  14.9 28  35.5 30  GP  Travel Time (minutes) Travel Speed (km/h) HOV Travel Time (minutes) Travel Speed (km/h)  A P P E N D I X E - Travel Time Savings Data  Page 164  TRA VEL TIME AND SPEEDS OCTOBER, 1996 WESTBOUND PEAK HOUR (7:30 AM TO 8:30 AM) H O V Corridor  Section  St. Johns St.  Barnet Hwy  Hastings St.  Distance (kms)  3.02  7.81  7.02  17.85 "  4.1 45  8.8 53  14.1 30  27.0 40  4.5 40  6.2 75  11.9 35  22.7 47  -0.5  2.6  2.2  4.3  GP Travel Time (minutes) Travel Speed (km/h) HOV Travel Time (minutes) Travel Speed (km/h) |  TRAVEL  TIME DIFFERENCE  | TRA VEL TIME DIFFERENCE  |  0.24  PER KM  PEAK PERIOD (6:00 AM TO 8:30 AM) Section  St. Johns St.  Barnet Hwy  Hastings St.  H O V Corridor  Distance (kms)  3.02  7.81  7.02  17.85  3.8 48  7.6 61  11.8 36  23.3 46  6.2  Travel Speed (km/h)  4.3 42  75  10.5 40  21.1 51  TRA VEL TIME SA VINGS  -0.5  1.4  1.3  2.2  GP Travel Time (minutes) Travel Speed (km/h) HOV Travel Time (minutes)  TRA J El. TIME DIFFERENCE  PER KM  A P P E N D I X E - Travel Time Savings Data  0.12  Page 165  TRA VEL TIME AND SPEEDS OCTOBER, 1997 WESTBOUND PEAK HOUR (7:30 AM TO 8:30 AM) Section  St. Johns St.  Barnet Hwy  Hastings St.  H O V Corridor  Distance (kms)  3.02  7.81  7.02  17.85  Travel Time (minutes)  4.3  11.7  17.6  Travel Speed (km/h)  42  40  24  33.7 32  Travel Time (minutes)  4.4  6.5  14.7  Travel Speed (km/h)  41  72  29  25.7 42  -0.1  5.2  2.9  8.0  GP  HOV  TRAVEL  TIME  DIFFERENCE  TRA VEL TIME DIFFERENCE  0.45  PER KM  PEAK PERIOD (6:00 AM TO 8:30 AM) Section  St. Johns St.  Barnet Hwy  Hastings St.  H O V Corridor  Distance (kms)  3.02  7.81  7.02  17.85  Travel Time (minutes)  4.0  9.3  13.9  Travel Speed (km/h)  46  51  30  27.1 39  Travel Time (minutes)  4.3  6.4  12.0  Travel Speed (km/h)  43  74  35  22.6 47  TRA VEL TIME SA VINGS  -0.3  2.9  1.9  4.6  GP  HOV  TRAVEL TIME DIFFERENCE  PER KM  A P P E N D I X E - Travel Time Savings Data  0.25  Page 166  TRA VEL TIME AND SPEEDS MARCH, 1996 EASTBOUND PEAK HOUR (5:00 PM TO 6:00 PM) Section Distance (kms)  Hastings St.  Barnet Hwy  H O V Corridor  7.05  7.93  14.97  18.7  13.5  23  35  32.2 28  GP  Travel Time (minutes) Travel Speed (km/h) HOV Travel Time (minutes) Travel Speed (km/h)  APPENDIX E - Travel Time Savings Data  Page 167  TRA VEL TIME AND SPEEDS OCTOBER, 1996 EASTBOUND PEAK HOUR (5:00 PM TO 6:00 PM) Section  Hastings St.  Barnet Hwy  H O V Corridor  Distance (kms)  7.05  7.93  14.97  Travel Time (minutes)  15.6  7.6  Travel Speed (km/h)  27  63  23.2 39  Travel Time (minutes)  13.9  6.9  Travel Speed (km/h)  30  69  20.8 43  1.7  0.7  2.4  GP  HOV  TRAVEL  TIME  DIFFERENCE  TRA VEL TIME DIFFERENCE  0.16  PER KM  PEAK PERIOD (3:30 PM TO 6:00 PM) Section  Hastings St.  Barnet Hwy  H O V Corridor  Distance (kms)  7.05  7.93  14.97  Travel Time (minutes)  14.5  7.4  Travel Speed (km/h)  29  65  21.8 41  Travel Time (minutes)  13.2  6.9  Travel Speed (km/h)  32  69  20.1 45  1.2  0.5  1.7  GP  HOV  |  TRA VEL TIME  DIFFERENCE  \TR4 VEL TIME DIFFERENCE  A P P E N D I X E - Travel Time Savings Data  PER KM  0.11  Page 1  TRA VEL TIME AND SPEEDS OCTOBER, 1997 EASTBOUND PEAK HOUR (5:00 PM TO 6:00 PM) Section  Hastings St.  Barnet Hwy  Distance (kms)  7.05  7.93  14.97  15.0  8.0  28  59  23.0 39  13.0  7.2  33  67  20.1 45  2.0  0.8  2.8  H O V Corridor  GP  Travel Time (minutes) Travel Speed (km/h) HOV  Travel Time (minutes) Travel Speed (km/h) TRAVEL  TIME  DIFFERENCE  TRA VEL TIME DIFFERENCE  PER  0.19  KM  PEAK PERIOD (3:30 PM TO 6:00 PM) Section  Hastings St.  Barnet Hwy  Distance (kms)  7.05  7.93  14.97  GP Travel Time (minutes) Travel Speed (km/h)  14.1  7.5  30  63  21.7 41  HOV Travel Time (minutes) Travel Speed (km/h)  12.2  7.0  35  68  2.0  0.6  TRA VEL TIME DIFFERENCE  \  [TRAVEL TIME DIFFERENCE PER KM \  H O V Corridor  19.1 47  1  2.5  |  0.17  A P P E N D I X E - Travel Time Savings Data Page 1  TRA VEL TIME DIFFERENCE CONFIDENCE INTERVALS OCTOBER, 1996 TO OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 AM TO 8:30 AM) Travel Time Difference Standard Sample t-critical (Mean) Deviation Size (t.025> October, 1996 October, 1997  2.16 4.59  2.64 3.86  18 25  2.110 2.064  CONFIDENCE INTERVALS 2.16 4.59  +/+/-  1.31 1.59  0.85 3.00  to to  3.47 6.18  to to  5.31 9.90  to to  2.37 3.37  to to  3.32 4.46  PEAK HOUR (7:30 AM TO 8:30 AM) Travel Time Difference Standard Sample t-critical (Mean) Deviation Size (t.025) October, 1996 October, 1997  4.31 8.05  1.19 2.76  8 11  2.365 2.228  CONFIDENCE INTERVALS 4.31 8.05  +/+/-  1.00 1.85  3.31 6.20  EASTBOUND PEAK PERIOD (3:30 PM TO 6:00 PM) Travel Time Difference Standard Sample t-critical (Mean) Deviation Size (t.025) October, 1996 October, 1997  1.71 2.54  1.66 2.10  27 27  2.056 2.056  CONFIDENCE INTERVALS 1.71 2.54  +/+/-  0.66 0.83  1.05 1.71  PEAK HOUR (5:00 PM TO 6:00 PM) Travel Time Difference Standard Sample t-critical (Mean) Deviation Size (t.025) October, 1996 October, 1997  2.41 2.83  1.44 2.57  A P P E N D I X E - Travel Time Savings Data  12 12  2.201 2.201  CONFIDENCE INTERVALS 2.41 2.83  +/+/-  0.91 1.63  1.50 1.20  Page 170  TRA VEL SPEED MARGIN OF ERRORS OCTOBER, 1996 TO OCTOBER, 1997 WESTBOUND PEAK PERIOD (6:00 AM TO 8:30 AM) Travel Speed Standard Sample t-critical  MARGIN OF ERROR  MARGIN OF ERROR  (Mean)  Deviation  Size  (t.02s)  (km/h)  47.2 41.9  7.72 10.74  18 25  2.110 2.064  3.8 4.4  8% 11%  October, 1996  51.5  6.34  18  6%  48.5  7.39  25  2.110 2.064  3.2  October, 1997  3.1  6%  MARGIN OF ERROR (km/h)  MARGIN OF ERROR (as a % of the mean)  GP October, 1996 October, 1997  (as a % of the mean) ,  HOV  PEAK HOUR (7:30 AM TO 8:30 AM) Travel Speed Standard Sample t-critical (Mean) Deviation Size (W  GP October, 1996 October, 1997  39.8 32.0  1.75  8  2.365  2.73  11  2.228  1.5 1.8  4% 6%  47.5 41.9  3.66 2.49  8 11  2.365 2.228  3.1 1.7  6% 4%  MARGIN OF ERROR  MARGIN OF ERROR  HOV October, 1996 October, 1997  ABOUND PEAK PERIOD (3:30 PM TO 6:00 PM) Travel Speed Standard Sample t-critical (Mean)  Deviation  Size  (t.025)  (km/h)  (as a °/o of the mean)  October, 1996  41.6  4.08  27  2.056  42.1  4.95  27  2.056  1.6 2.0  4%  October, 1997 October, 1996  45.1  4.41  27  2.056  1.7  4%  October, 1997  47.4  4.48  27  2.056  1.8  4%  MARGIN OF ERROR (km/h)  MARGIN OF ERROR (as a % of the mean)  3_4  5% 8%  2.0  8% 5%  GP  5%  HOV  PEAK HOUR (5:00 PM TO 6:00 PM) Travel Speed Standard Sample t-critical (Mean) Deviation Size (t.02j)  GP October, 1996 October, 1997  39.0 39.8  3.34  12  2.201  5.31  12  2.201  43.8 44.9  5.23  12  3.19  12  2.201 2.201  HOV October, 1996 October, 1997  A P P E N D I X E - Travel Time Savings Data  Page 171  TRA VEL TIME DIFFERENCE OCTOBER,  STA TISTICAL  1996 TO OCTOBER,  SIGNIFICANCE  1997  WESTBOUND  Travel Time Difference Standard Sample (Mean) Deviation Size  test t-critical statistic (t.<») B  A>B ?  yes yes  October, 1997  2.16 4.59  2.64 3.86  18 25  3.47 5.95  1.740 1.711  MINIMUM*  1.21  3.25  22  1.73  1.721  October, 1996  PEAK HOUR  SIGNIFICANT  A  (7:30 AM TO 8:30 AM)  Travel Time Difference Standard Sample (Mean) Deviation Size 4.31 8.05  October, 1996 October, 1997  |  MINIMUM*  1.18  8 11  1.19 2.76 |  1.98  |  test t-critical statistic (tos)  STATISTICALLY SIGNIFICANT  A  B  A>B ?  10.24 9.67  1.895 1.812  yes yes  10 | | 1 84 | 1.833 \  EASTBOUND PEAK PERIOD  (3:30 PM TO 6:00 PM)  Travel Time Difference Standard Sample (Mean) Deviation Size October, 1996 October, 1997 MINIMUM* PEAK HOUR  test t-critical statistic (t.os)  STATISTICALLY SIGNIFICANT  A  B  A>B ?  yes yes  1.71 2.54  1.66 2.10  27 27  5.35 6.28  1.706 1.706  0.62  1.88  27  1.71  1.706  (5:00 PM TO 6:00 PM)  Travel Time Difference Standard Sample (Mean) Deviation Size October, 1996 October, 1997  2.41 2.83  MINIMUM*  1.04  1.44 2.57 I  201  12 12 I  test t-critical statistic (t.05)  STATISTICALLY SIGNIFICANT  A  B  A>B?  5.80 3.81  1.796 1.796  yes yes  12 I J 1.80 I 1.796"|  NOTES: 1. MINIMUM based on an average standard deviation & sample size  A P P E N D I X E - Travel Time Savings Data  Page 172  TRA VEL TIME AND SPEEDS OCTOBER, 1997 WESTBOUND 7:00 A M  16-Oct 17-Oct 20-Oct 21-Oct 22-Oct 23-Oct 24-Oct 27-Oct 28-Oct 29-Oct 30-Oct 31-Oct  8:00 A M  GP LANE  HOV LANE  GP LANE  HOV LANE  42.00  47.75  32.59  47.22  42.99  52.21  43.02  45.32  42.11  47.85  40.02  46.34  39.21  47.14  35.82  44.21  42.65  50.84  31.06  43.28  34.74  39.72  35.62  39.92  42.84  51.09  47.71  51.33  45.35  50.45  42.17  46.14  39.40  49.67  37.72  45.71  35.53  47.26  36.15  44.29  37.21  50.41  38.32  45.48  38.53  56.23  46.40  49.90  Mean Travel Speed  40.21  49.22  38.88  45.76  Standard Deviation  3.29  3.93  5.17  2.95  Standard Deviation as a % of the mean  8.2%  8.0%  13.3%  6.5%  F-statistic F-critical (F 0.05)  A B  Statistically Significant? | A>B Sample Size  0.70 2.82  3.06 2.82  NO  yes  t-critical (t 0.025)  12 2.201  12 2.201  12 2.201  12 2.201  | Margin of Error (km/h)  2.09  2.50  3.28  1.88  Margin of Error as a % of the mean  5.2%  5.1%  8.4%  4.1%  A P P E N D I X F - Travel Time Reliability Data  I  Page 173  TRA VEL TIME AND SPEEDS OCTOBER, 1997 EASTBOUND 4:30 P M  16-Oct 17-Oct 20-Oct 21-Oct 22-Oct 23-Oct 24-Oct 27-Oct 28-Oct 29-Oct 30-Oct 31-Oct  5:30 P M  GP LANE  HOV LANE  GP LANE  HOV LANE  55.12  59.34  39.97  50.84  48.91  54.93  51.13  55.88  48.87  54.88  47.26  53.78  52.89  57.69  49.7  51.66  49.17  58.96  46.71  55.69  45.55  53.2  38.25  49.51  50.41  56.37  40.83  43.42  56.37  56.87  51.58  47.08  42.70  43.99  40.29  50.29  45.45  56.62  47.46  48.32  42.96  51.25  52.33  54.23  31.53  42.48  42.45  46.17  Mean Travel Speed  47.49  53.88  45.66  50.57  Standard Deviation  6.66  5.48  5.07  3.91  Standard Deviation as a % of the mean  14.0%  10.2%  11.1%  7.7%  F-statistic F-critical (F .os) 0  A B  Statistically Significant? A>B  1.48 2.82  1.68 2.82  NO  NO  t-critical (t 0.025)  12 2.201  12 2.201  12 2.201  Margin of Error (km/h)  4.23  3.48  3.22  Margin of Error as a % of the mean  8.9%  6.5%  Sample Size  APPENDIX F - Travel Time Reliability Data  I  7.1%  12 2.201 |  2.49 4.9%  Page 174  COMPLIANCE RA TES WESTBOUND Corridor  Barnet/Hastings HOV Corridor  Section  St. Johns Street  Location  Williams Clark  Barnet Highway  Hastings Street  Reed  Bayview  Willingdon  Esmond  Lillooet  86%  87%  89%  86%  77%  OCTOBER, 1996 PEAK PERIOD  (6:15 TO 8:15 AM)  Location Mem  88%  87% 88%  Section Sfeun  87%  Corridor Mean  87% 87%  PEAK HOUR (7:30 TO 8:30 AM) |  Location Mean  \ |  85%  |  Section Mean  |  |  Corridor Mean  { £  [~  | 77% |  |  82%  ] [  81%  }  87%  { {  88%  |  3C  85%  83%  |  70%  |  80%  82%  MARCH, 1997 PEAK PERIOD  (6:15 TO 8:15 AM)  Location Mean  87%  Section Mean  89%  76%  88%  85%  89%  82% 85%  81%  ('orridor Mean  74%  85%  PEAK HOUR (7:30 TO 8:30 AM) |  Location Mean  \ |  |  Section Mean  \ |  |  Corridor Mean  |  84%  | 9"o%""| \ 87%  |  73% f~~  |  82%  ) \  89%  |  ][  78%  83%  76%  83%  85%  |  OCTOBER, 1997 PEAK PERIOD  (6:15 TO 8:15 AM) 83%  Location Mean  70%  78%  76%  Section Mean  69%  78%  74%  73%  33%  76% 75%  Corridor Mean PEAK HOUR (7:30 TO 8:30 AM) \  Location Mean  \ \  |  Section Mean  | J  |  Corridor Mean  78%  | 74%~|  76%  |  ] [  79%  | 76%  73%  \ \  ][  79%  |  79%  |  44%  \  67%  72%  NOTES: Lillooet rates not included in averages  APPENDIX G - Compliance Data  Page 175  COMPLIANCE RA TES EASTBOUND Corridor  Barnet/Hastings HOV Corridor  Section  Hastings Street  Location  Bamet Highway  Lillooet Cassiar Kootenay Carleton Willingdon Howard  Bayview  Union  89%  87%  OCTOBER, 1996 PEAK PERIOD  (3:45 TO 5:45 PM)  Location Mean  65%  82%  86%  Section Mean  87%  85%  84%  85%  88%  Corridor Mean  86%  PEAK HOUR (5:00 TO 6:00 PM) |  Location Mean  | | 63% | 80% |  |  Section Mean  | |  |  Corridor Mean  { £  82%  | 86%  |  82%  79%  j 81% II II  1 82%  88%  85%  80%  |  1  MARCH, 1997 PEAK PERIOD  (3:45 TO 5:45 PM)  Location Mean Section Mean  64%  68%  75%  |  80%  74%  81%  95%  76%  84% 89%  Corridor Mean  80%  PEAK HOUR (5:00 TO 6:00 PM) |  Location Mean  \ { 64% | 75% |  |  Section Mean  \ |  Corridor Mean  J[  [  80%  |  83% |  76%  \ 84% | |  96%  ][  77% 80%  |  86% |  91%  OCTOBER, 1997 PEAK PERIOD  (3:45 TO 5:45 PM)  Location Mean  55%  73%  80%  Section Mean  76%  79%  78%  11%  Corridor Mean  89%  69% 79%  78%  PEAK HOUR (5:00 TO 6:00 PM) |  Location Mean  \ \ 45% | 72% |  |  Section Mean  { |  |  Corridor Mean  { £  84%  |  73% |  76%  72%  | 80% \ \  86%  |  66% |  76% 73%  NOTES: Lillooet rates not included in averages  A P P E N D I X G - Compliance Data  Page 176  13.5  Hugh to Dewdney  Moody to Hugh  141  134  674 885 968  476 484  8  16.4 16.27  789  381 400 428 449  7  647  635  590  491 524  312 354  235 282 267 316  7 7  8  222 262  7 13  0  35  48  476 484  49  46  968  83  96  115  12 27  2  66 45  33  97 87  39  20 2  17 14  242  21  24 4  29 4  7  3  11  8 17  22  24  10  5 1  4 4  1 2  5  4  2 6  0  0  2  4  9  5  3  5  8  4  4  4  4  4 4  4  4  4  4  4 4  4  4  8  4 4 4  17 18 4  4 4  4  4  4  4  22  1 7  0  7  16  4  4  4  10 19 37  4  4  33400  33400  33400  33400  33400  33400  33400  33400  33400  33400 33400  33400  33400  33400  33400  33400 33400  33400 33400  33400  33400  33400  33400  33400  33400  33400  33400  33400  33400  33400  ACCIDENT YEARS FREQUENCY (acc/yr) AADT (yr)  19  30  TOT (acc)  14  22  8  69 47  10 12  23  20 34  0  0  0  1  0  0  32  48  0 0  39 39  16 58  23  0 0  8 0  0  268  12 2  0  266  10 7  7  3 2  1 4  11  0 6  307 404  5 13 12  7  9 5  1  0  3  9  21  11  8  0  0 0  0  290 338 300 340  174 223  7  0  0  232 246  7  135 165  7  0  0 2  3  0  6  1  2 1  215  203 207  185  168  139 146  138  138  131  1  8 16  0  0  115  78  1  9  10  1  16  13  59  1 0  30  I PDO  49  8  1.4  98  7  122 124  119  96 108  83  78  77  110 149 112 149  85 91  7 7  81  6 7  80  68 73  4  4  4 4  59  57  4  57  77  4  74  57  65  44  33  54  4  3 4  16 25  8  16.3  0.2  0.6  0.6  32 48  2 2  24  2  13 23  1 1  I  F  CUMULATIVE PDO TOT F  16.3  14.9  14.9  14.7 14.7  Maaty  14.9  14.7  14.1  Queens to Moody  '  ,  14.1  3tidbs|7A)  0.8  1.3  12.8  11.4  11.4  ; 12.8 12.8  7.3 10.2  ' " ' 13.5 St Johns to Queens 13.5 14.1  Gore to St Johns  Gore-  1.1  1.2 •  0.1  0.4  1.8 3.0 1.2  5.5  4.3  3.2  3.0  1.2  0.8  0.7  7.3 10.2  5.5  4.3 4.3  3.2  3.0 3.2  APPENDIX H - Safety Data  '  ^peMn£  Sperling to Malibu Malibu to Texaco Texaco to PetroCan PetroCan to Reed Reed to Gore  '  J  ?.  Holdom to Sperling  muz  Ellnsmere to Holdom  2.7  Spnnger to Ellensmer  3.0  1.5 2.7  Willingdon to Springe  2.7  1.5  1.5  '!  £3ttaere  ' 0.7 Gilmore to Willingdon 0.7  \  0.7  0.0  0.0  Boundary to Gilmore  LKI LKI Start Finish Length  BEFORE ACCIDENT RATES  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  48.76  MEV  67.78  7.31  29.75  29.26  36.57  64.86  59.98  144.34  55.59 86.31  58.03  5.85  18.53  56.57  39.01  33.65  MVK  1.70  1.97  0.55  0.92  0.68  1.99  0.04  0.37  0.45  0.02  0.14  0.76  0.21  0.62  KATE  IM'FKSKCIION  Page 177  1.70  1.64  2.22  2.97  1.07  0.31  0.12 0.23  0.09  0.07  0.29  1.20  0.00  0.28  0.49  0.56  SEC HON RAIE  16.4  APPENDIX H - Safety Data  5  Boundary to Gilmore ''', OiBKtf* " Gilmore to Wilhngdo  0.0 0.0 0.7 0.7 :mk&& ?\ 1.5 Willingdon to Springe 1.5 2.7 Springer to hllensmer 2.7 3.0 Ellnsmcre to Holdom 3.0 '", /,WM&to ' \\ 3.2 3.2 Holdom to Sperling 4.3 4.3 Sperling to Malibu 5.5 Malibu to Texaco 7.3 Texaco to PetroCan 10.2 PetroCan to Reed 11.4 Reed to Gore 12.8 12.8 Gore to St Johns 13.5 St Johns to Queens 13.5 14.1 14.1 Queens to Moody 14.7 14.7 Moody to Hugh 14.9 Hugh to Dewdney 14.9 16.3 1.2 0.4 0.1 1.2 1.1 1.8 3.0 1.2 1.3 0.8 0.6 0.6 0.2 1.4  2.7  3.0  3.2  4.3  5.5 7.3 10.2 11.4 12.8  13.5  14.1  14.7  14.9  16.3  16.27  0.8  0.7  1.5  0.7  LKI LKI Start Finish Length  AFTER ACCIDENT RATES  0 3  274  4 5 1 8 2 5 0 2 7 1 3 19 3 12 7 10 9 1 4 4 0 5 6 11 4 3 4 10 8  PDO  8 7 106 165  0 5 1 6 3 4 0 1 4 1 0 5 0 5 4 7 4 0 1 5 2 3 1 9 2 6 4 12 3  0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0  4 14 16 30 35 45 45 48 60 62 65 89 92 109 120 137 150 151 156 165 167 175 182 202 208 218 226 248 259  4 9 10 18 20 25 25 27 34 35 38 57 60 72 79 89 98 99 103 107 107 112 118 129 133 136 140 150 158  I  106 165  0 5 6 12 15 19 19 20 24 25 25 30 30 35 39 46 50 50 51 56 58 61 62 71 73 79 83 95 98  0 0 0 0 0 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3  I  F  CUMULATIVE PDO TOT F  15 274  4 10 2 14 5 10 0 3 12 2 3 24 3 17 11 17 13 1 5 9 2 8 7 20 6 10 8 22 11  TOT (acc)  0.83  0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 18 328.8  5 12 2 17 6 12 0 4 14 2 4 29 4 20 13 20 16 1 6 11 2 10 8 24 7 12 10 26 13 33400  33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400 33400  \ f f I D F VI Y F \ R S KKlQl.h>CY (acc/yr) AADT (yr)  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  10.16  MEV  14.12  1.52  6.20  6.10  7.62  11.58 17.98 30.07 12.50 13.51  12.09  1.22  3.86  11.78  8.13  7.01  MVK  1.48  1.08  0.79  0.59  0.69  0.20  0.49  0.30  0.30  1.18  0.00  0.49  0.20  0.39  INTFKShOIO.N RATE  i:  Page 178  1.56  6.56  3.23  1.31  1.18  1.47 0.61 0.57 1.04 0.07  1.99  1.64  0.78  0.85  1.72  1.43  RA'I  SI'CUON  0.49 0.20 0.69 0.59 0.79 1.08 1.48  1.99 0.68 0.92 0.55 1.97 1.70  14  0.30  0.37 0.04  ooS  A P P E N D I X H - Safety Data  STATISTICALLY SIGNIFICANT ? [  0  t-statistic t @ 95% with 13 d.o.f. t .os @ 90% with 13 d.o.f.  ST.DEVDilTertnce  e  NO  1.350  1.771  0.914  0.655  0.429  SUM Of (XBi-XBavgXXAi-XAavg) DifrerenC  1.346  VARIANCE  13  N N-l  0.18  0.30  0.45  0.46  1.18  0.02  0.42  0.00  0.14  0.58  0.49  0.76  0.68  0.20  0.21  Boundary Gilmore Willingdon Springer Ellensmere Holdom Sperling Gore St Johns (7 A) Queens Moody Hugh Dewdney loco  0.74  0.39  0.62  MEAN ST.DEV. VARIANCE  "After"  "Before"  INTERSECTION  ACCIDENT RATES BEFORE & AFTER  DilT(ren  g  Ai  Yes  STATISTICALLY SIGNIFICANT ?|  1.753  2.326  1.372  1.883  8.626  15  1.50  |  Page 179  2.24  1.63  1.70  0.89 0.79  1.56  1.64  0.83  3.23 6.56  2.22  1.18 1.31  1.07 2.97  1.04 0.07 0.31  0.57 0.23  0.12  0.29  1.47  1.99  1.20  0.61  1.64  0.00  0.09  0.78  0.28  0.07  1.72 0.85  0.49  1.43  0.56  1.341  Aavg  )  16  X i  Xbj  A  "After"  "Before"  t-statistic t„. @ 95% with 15 d.o.f. t„.„s @ 90% with 15 d.o.f. oS  B  ST.DEVDifTertnce  B  N N-l SUM of (X rX ,v )(X -X VARIANCE «  MEAN ST.DEV. VARIANCE  Boundary to Gilmore Gilmore to Willingdon Willingdon to Springer Springer to Ellensmere Ellensmere to Holdom Holdom to Sperling Sperling to Malibu Malibu to Texaco Texaco to PetroCan PetroCan to Reed Reed to Gore Gore to St. Johns St Johns (7A) to Queens Queens to Moody Moody To Hugh Hugh to Dewdney  SECTION  BURNABY ACCIDENT REPORT Before (3 years, 7 months)  After (1 year, 2 months)  Intersection  TOTAL  OTAL/YR  TOTAL  TOTAL/YR  Hastings and Boundary Hastings and Esmond Hastings and Ingleton Hastings and Macdonald Hastings and Gilmore Hastings and Carleton Hastings and Madison Hastings and Rosser Hastings and Willingdon Hastings and Alpha Hastings and Beta Hastings and Gamma Hastings and Hythe Hastings and Springer Hastings and Howard Hastings and Ellesmere Hastings and Warwick Hastings and Holdom Hastings and Fell Hastings and Kensington Hastings and Grove Hastings and Sperling Hastings and Clare Hastings and Duncan Hastings and Ellerslie Hastings and Cliff Hastings and Inlet  66 8 14 17 42 15 19 29 82 34 10 38 5 12 10 6 7 69 17 31 7 46 7 12 2 4 14 623  18 2 4 5 12 4 5 8 23 9 3 11 1 3 3 2 2 19 5 9 2 13 2 3 1 1 4 174  9 9 5 2 17 4 7 11 12 8 4 3 4 3 8 3 1 5 7 10 1 9 1 1 0 2 2 148  8 8 4 2 15 3 6 9 10 7 3 3 3 3 7 3 1 4 6 9 1 8 1 1 0 2 2  APPENDIX H - Safety Data  127  Page 180  VANCOUVER ACCIDENT REPORT  Intersection Hastings and Renfrew Hastings and Lilloet Hastings and Windermere Hastings and Rupert Hastings and Cassiar Hastings and Skeeena Hastings and Kootenay Hastings and Boundary T O T A L @ INTERSECTION TOTAL @ MID-BLOCK TOTAL TOTAL/YR  APPENDIX H - Safety Data  Before (4 years)  After (4 months)  Fatalities Injuries PDOs 1 37 86 0 16 23 0 6 8 0 7 11 0 10 14 0 10 19 0 0 1 0 3 14 0 20 37 0 12 29 0 17 42 0 8 40 0 11 19 0 9 34 1 17 47  Fatalities Injuries PDOs 0 1 6 0 0 3 0 0 1 1 0 1 0 1 1 0 0 0 0 0 0 0 0 0 0 1 3 0 0 2 0 0 5 0 0 0 0 0 0 0 0 0 0 2 4  2 0 2 llllllll  118 65 183 46  254 170 424 106  0 1 1 3  5 0 5 15  20 6 26 78  Page 181  IMP A CT ON GP LANES WESTBOUND Corridor  Barnet/Hastings H O V Corridor  Section  St. Johns Street  Location  Williams  Reed  Bayview  Willingdon  Lilloet  2060  1363  1392  1909  2191  Barnet Highway  Hastings Street  MARCHJ996 GP Volume Section GP Volume Section Travel Speed  2060  1378  2050  18  44  28  | Corridor Travel Speed  30  OCTOBER, 199 6 GP Volume Section GP Volume Section Travel Speed  1898  1574  1606  1352  1898  1590  1352  45  53  30  Corridor Travel Speed  40  OCTOBER,1997 GP Volume Section GP Volume Section Travel Speed Corridor Travel Speed \ £  A P P E N D I X I - Impact on G P Lanes Data  2063  1499  1401  1889  1818  2063  1450  1854  42  40  24  32  Page 182  IMP A CT ON GP LANES EASTBOUND Corridor  Barnet/Hastings H O V Corridor  Section Location  Hastings Street Lillooet  Carleton  Barnet Highway Willingdon  Bayview  Union  1832  1380  1431  MARCH,1996 GP Volume  1927  Section GP Volume Section Travel Speed  1880  1406  23  35  Corridor Travel Speed  28  OCTOBER,1996 GP Volume  1478  Section GP Volume Section Travel Speed  2029  1523  1678  1754  1601  27  63  Corridor Travel Speed  39  OCTOBER, 199 7 GP Volume Section GP Volume Section Travel Speed | Corridor Travel Speed  A P P E N D I X I - Impact on G P Lanes Data  1493  1837  2071  1800  1644  1819 1732 .  28  59  39  Page 183  EMME/2 2006 MODELING OUTPUT WESTBOUND Travel Times "Do nothing" "Add HOV GP St. Johns Street 2.32 1.30 4.93 2.36 1.62 0.95 0.95 0.67 0.58 0.55 3.45 1.70 0.22 0.12 Section Total 14.07 7.65 Barnet Highway  Section Total Hastings Street  Section Total Corridor Total |  Travel Time Differences lane " HOV 1.11 1.78 0.91 0.48 1.19 0.44 0.62 6.53  1.58 0.80 1.00 0.56 0.92 1.08 1.68 5.12 1.12 0.61 0.93 15.40  0.77 0.39 0.49 0.27 0.45 0.52 0.82 2.48 1.08 0.59 0.69 8.55  0.59 0.36 0.51 0.23 0.39 0.46 0.72 2.19 1.08 0.62 0.67 7.82  2.96 1.35 0.31 3.30 0.61 3.89 1.58 2.89 0.48 2.82 3.08 2.91 3.37 1.56 31.11  1.09 0.79 0.30 1.04 0.59 0.79 0.57 1.06 0.47 0.87 1.13 1.07 0.83 0.91 11.51  0.80 0.77 0.31 0.88 0.71 0.60 0.58 1.01 0.58 0.80 1.10 1.04 0.93 0.81 10.92  60.58  APPENDIX J - Multi-Criteria Evaluation  ||  27.71  |  25.27  |  GP 6.4  HOV 7.5  GP  6.9  HOV 7.6  GP 19.6  HOV 20.2  32.9  |  35.3  "1  Page 184  PERSON- VOLUMES 1996 & 1997 OCTOBER, 1996 WESTBOUND  EASTBOUND  St. Johns Street  GP HOV Total  Drivers 1886 265 2151  Occupants 424 247 671  Bus 197 500 697  Total 2507 1012 3519  Barnet Highway  GP HOV Total  1574 447 2021  367 378 745  0 325 325  1941 1150 3091  Hastings Street  GP HOV Total  1353 539 1892  152 528 680  0 1121 1121  1505 2188 3693  Hastings Street  GP HOV Total  Drivers 2027 560 2587  Occupants 365 523 888  Bus 0 435 435  Total 2392 1518 3910  Barnet Highway  GP HOV Total  1523 433 1956  125 402 527  0 256 256  1648 1091 2739  St. Johns Street  GP HOV Total  Drivers 2057 411 2468  Occupants 436 384 820  Bus 149 425 574  Total 2642 1220 3862  Barnet Highway  GP HOV Total  1499 665 2164  28 604 632  0 321 321  1527 1590 3117  Hastings Street  GP HOV Total  1888 565 2453  188 505 693  10 872 882  2086 1942 4028  Hastings Street  GP HOV Total  Drivers 2070 662 2732  Occupants 395 593 988  Bus 10 774 784  Total 2475 2029 4504  Barnet Highway  GP HOV Total  1644 656 2300  340 346 686  0 210 210  1984 1212 3196  OCTOBER, 1997 WESTBOUND  EASTBOUND  NOTES:  - Westbound locations: St. Johns @ Williams, Barnet @ Reed, Hastings @ Willingdon - Eastbound locations: Hastings @ Willingdon, Barnet @ Bayview  APPENDIX J - Multi-Criteria Evaluation  Page 1  0 10 149  28 188 436 28 188 436 28 188 436 28 188 436 28 188 436 28  1499  1888 2057  1499 1888 2057  1499  1888 2057  1499  1888 2057  1499  1888 2057  1499  1888 2057  1499  1888  Bamet  Hastings St. Johns  Hastings St. Johns  Bamet  Hastings St. Johns  Bamet  Hastings St. Johns  Bamet  Hastings St. Johns  Bamet  Hastings St. Johns  Bamet  Hastings  0  505 384  565 411  872 425 321  604  604 505  665 565 411 665 565  6.4 17.72 6.4 6.9 19.6  0 10 149 0 10  872 425  505 384  872  321  321  604  505 384  565 411  665  5.9 15.84 6.34  0 10 149  872 425  321  505 384  565 411  604  665  5.4 13.96 6.28  872 425  321  4.9 505 384  872 425  505 384  565 411 604  321  604  665  4.4 10.2 6.16 565 411  872 425  505 384  565 411  665  321  604  665  3.9 8.32 6.1  12.08 6.22  872 425  505 384  565 411  321  604  665  872 425  3.4 6.44 6.04  0 10 149  10 149  604  665  321  872 425  505 384  76  807  257  7.6 20.2  732 271  231  80  211 95  241 18.5 7.5  7.29  73 191 93  225 656 267  170 91  70  150 89  66  130 87  63  109 86  60  89 84  56  68 82  53  48 80  50  40 28 78  16.8 7.29  209 581 264  506 260  193  430 257  177  355 253  161  279 250  145  129 204 246  129 243  113  83 45 239  Occupants 63  6.98  15.0 7.08  6.67  13.3 6.87  6.36  11.6 6.66  6.05  9.9 6.45  5.74  8.2 6.24  5.43  5.12 6.4 6.03  4.7 5.82  4.81  4.5 3.0 5.61  Drivers 206  297  41  272 69  39  246 67  37  221 66  36  196 64  34  170 62  32  145 61  31  120 59  29  94 58  27  69 56  26  24 56 54  Bus 64  TRAVEL TIME SAVED PER DAY (hours)  107931  34323  97853 36218  32192  87774 35750  30062  77696 35283  27932  67617 34815  25801  57538 34348  23671  47460 33880  21541  37381 33413  19411  27303 32945  17280  15150 17224 32478  11151 6017 32010  Drivers 27574  20250  6976  18468 8269  6683  16686 8114  6389  14904 7958  6096  13122 7802  5802  11339 7646  5509  5215 9557 7491  4922 7775 7335  5993 7179  4628  4211 7023  4335  3498 2487 6868  Occupants 5531  3. assumed 250 days per year  25974  3558  23759 6039  3413  21544 5896  3268  19329 5753  3122  17115 5610  2977  14900 5466  2832  12685 5323  2687  2542 10471 5180  8256 5037  2397  2252 6041 4894  2133 4904 4750  Bus 5604  154155  44857  140080 50526  42288  39719 126004 49760  111929 48993  37150  97853 48227  34581  83778 47461  32012  69702 46694  29443  55627 45928  26874  41552 45161  24305  27476 44395  21736  16781 13409 43628  TOTAL 38710  WAGE HOURS SAVED PER YEAR  NOTES: 1. travel time differences between 1996 and 2006 interpolated 2. travel rime saved converted to wage hours saved as follows: drivers • 0.535 (5% commercial drivers and 95% regular drivers), occupants * 0.35, bus *0.35  188  0 10 149  28 188 436  1888 2057  Bamet  0 10 149  28 188 436  1499  Barnet  Hastings St. Johns  10 149  2.9 4.56 5.98  1888 2057  0  28 188 436  1499  Barnet  Hastings St. Johns  APPENDIX J - Multi-Criteria Evaluation  2006  2005  2004  2003  2002  2001  2000  1999  1998  321  604  565 411  665  2.4 2.68 5.92  0  1888 2057  10 149  28 188 436  1499  Barnet  Hastings St. Johns  1997  325 1121 425  378 528 384  447 539 411  (minutes) 5.4  Occupants 247  Drivers 265  Bus 500  Travel Time Difference  HOV Person Volumes  1.9 0.8 5.86  (minutes) 5.8  Travel Time Dinerence  0 0 149  367 152 436  Bamct Hastings St. Johns  Bus 197  OP  Drivers Occupants 424 1886  Person Volumes  1574 1353 2057  St. Johns  1996  WESTBOUND  WA GE HOURS SA VED PER YEAR  wmmm  Page 186  | 249538 |  |  214716 |  197306 |  179*95 |  1624«4 |  145073 |  127662 |  110252 |  92841 |  68900 |  PEAK HOUR FACTORS WESTBOUND  Time from to 0:00 1:00 1:00 2:00 2:00 3:00 3:00 4:00 4:00 5:00 5:00 6:00 6:00 7:00 7:00 8:00 8:00 9:00 9:00 10:00 10:00 11:00 11:00 12:00 12:00 13:00 13:00 14:00 14:00 15:00 15:00 16:00 16:00 17:00 17:00 18:00 18:00 19:00 19:00 20:00 20:00 21:00 21:00 22:00 22:00 23:00 23:00 0:00  Vehicle Travel Time Savings (vehicle-hours)  EASTBOUND Percentage of Peak Hour Travel Time Savings  0.15 215.4 483 0 244.0 2.60 0.48 0.19 0.18 0.28 0.44 1.13 0.91 0.31 0.30 0.04 0.01 0.01 0.01  0.03% 44.6% 100.0% 50.5% 0.54% 0.10% 0.04% 0.04% 0.06% 0.09% 0.23% 0.19% 0.06% 0.06% 0.01% 0.00% 0.00% 0.00%  TOTAL  197%  FACTOR  1.97  |  Vehicle Travel Time Savings (vehicle-hours)  Percentage of Peak Hour Travel Time Savings (%)  0.05 0.13 0.08 0.04 0.05 0.12 0.23 0.55 1.14 100.0 473 0 200.4 13.29 0.53 0.14 0.12 0.04 0.04  0.01% 0.03% 0.02% 0.01% 0.01% 0.03% 0.05% 0.12% 0.24% 21.1% 100.0% 42.4% 2.81% 0.11% 0.03% 0.03% 0.01% 0.01%  TOTAL  167%  FACTOR  1.67  NOTES: vehicle-hours obtained from the "Benefit/Cost Evaluation of Barnet/Hastings Project" (MoTH,1993)  APPENDIX J - Multi-Criteria Evaluation  Page 187  PEAK HOUR TRAVEL TIME SAVINGS WESTBOUND INPUT:  Worst Case DISCOUNT RATE  12%  Most Likely 8%  AVERAGE WAGE RATE  14.00  18.00  23.00  18 00  GROWTH RATE  0%  2%  5%  2%  Best Case 6%  CHOICE 8%  OUTPUT: PRESENT VALUE  FROM 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015  TO 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016  S  13,245.558  S  22,532,918  $  36,908,539  PROJECT WAGE HOURS WAGE HOURS TRAVEL TIME SAVED/YEAR WITH GROWTH SAVED/YEAR YEAR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 68,900 68,900 1,240,200 2 92,841 92,841 1,671,138 3 110,252 2,024,227 112,457 4 127,662 2,390,752 132,820 5 145,073 2,771,147 153,953 6 162,484 3,165,802 175,878 7 179,895 3,575,135 198,619 8 197,306 3,999,575 222,199 9 214,716 4,439,541 246,641 10 232,127 4,895,528 271,974 11 5,367,978 249,538 298,221 12 0 13 0 14 0 15 0 16 0 17 0 18 0 19 0 20 0 1,780,794  APPENDIX J - Multi-Criteria Evaluation  1,974,501  $ 35,541,024  DISCOUNTED TOTALS 0 0 0 0 0 0 1,240,200 1,547,350 1,735,448 1,897,856 2,036,876 2,154,592 2,252,942 2,333,714 2,398,546 2,448,983 2,486,413 0 0 0 0 0 0 0 0 0 $ 22,532,918  Page 188  APPENDIX J - Multi-Criteria Evaluation  OCTOBER, 1996  252 272 247 260  5165 5584 5071 5328 4953 4953  5417 5856 5319 5587 5194 5194  6663 7203 6542 6873 6389 6389  St. Johns ReedPt. Bayview Barnet Willingdon Lilloet Hastings  5548 6503 6026  7157 8389 7773 6082 6017 6050  Lilloet Willingdon Hastings Bayview Union Hastings  NOTES: - based on 250 working days per year  4715 4664 4690  1.29 # of vehicles  Person Volume  Location  1810  14.97 TOTAL KMS/YEAR  TOTAL KMS  844 7.92 |  107 106 107 4608 4558 4583  965  7.05  KMS SAVED  4487 17.85  KMS  1696  7.02 |  2030  761  KMS SAVED  126 148 137  Difference  |  7.81  3.02  KMS  5422 6355 5889  1.32 # of vehicles  "BEFORE" AVO "AFTER" AVO  TOTAL KMS TOTAL KMS/YEAR  242  242  Difference  1.29 # of vehicles  1.23 # of vehicles  'BEFORE" AVO "AFTER" AVO Person Volume  Location  NUMBER OF KILOMETRES SA VED  |  |  Page 189  1,574,036  452,395  1,121,642  1997  OF KILOMETRES  A P P E N D I X J - Multi-Criteria Evaluation  EASTBOUND PEAK PERIOD  WESTBOUND PEAK PERIOD  OCTOBER,  NUMBER  6749 7836 8741 7775 6836 6981 6909  Lilloet Carleton Willingdon Hastings Bayview Union Hastings  5299 5412 5355  5232 6074 6776 6027  1.29 # of vehicles  |  274 236 255  282 276 279  289  198 202 200  195 227 253 225  TOTAL KMS TOTAL KMS/YEAR  5101 5210 5156  5037 5848 6523 5802  1.34 # of vehicles  {Difference  TOTAL KMS TOTAL KMS/YEAR "BEFORE" AVO "AFTER" AVO  NOTES: - based on 250 working days per year  Person | Volume  5620 4847 5234  5894 5084 5489  7250 6253 6752  Willingdon Lilloet Hastings  Location  5791 5657 5724  6073 5933 6003  7470 7297 7384  Reed Pt. Bayview Barnet  5924  6213  7642  1,29 """^Difference # of vehicles  St. Johns  |  1.23 # of vehicles  Person | Volume  "BEFORE" AVO " A F T E R " AVO Location  SA VED  |  |  14.97  7.92  7.05  KMS  17.85  7.02 |  7.81  3.02  KMS  3168  1583  1586  KMS SAVED  4846  1792  2181  873  KMS SAVED  Page 190  | 2,003,43(71  792,053  1,211,377  VEHICLE OPERA TING COST SA VINGS WESTBOUND & EASTBOUND  INPUT:  Worst Case 12%  Most Likely 8%  Best Case 6%  CHOICE  OPERATING COST RATE  0.08  0.09  0.12  0 09  GROWTH RATE  0%  2%  5%  2%  KMS SAVED AH  899,146  1,458,709  2,003,430  1.458.709  DISCOUNT RATE  8%  OUTPUT: PRESENT VALUE  FROM 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015  TO 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016  PROJECT YEAR  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  S  655,756  S  1,591,122  KMS SAVED/YEAR 0 0 0 0 0 0 1,574,036 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709  KMS SAVED WITH GROWTH 0 0 0 0 0 0 1,574,036  29,289,507  34,891,764  1,458,709  1,487,883 1,517,641 1,547,994 1,578,954 1,610,533 1,642,743 1,675,598 1,709,110 1,743,292 1,778,158 1,813,721 1,849,996 1,886,996 1,924,736 1,963,230 2,002,495 2,042,545 2,083,396  S  4,151,110 |  VEHICLE OPERATING $ SAVED/YEAR 0 0 0 0 0 0 141,663 131,284 133,909 136,588 139,319 142,106 144,948 147,847 150,804 153,820 156,896 160,034 163,235 166,500 169,830 173,226 176,691 180,225 183,829 187,506 $  3,140,259  DISCOUNTED TOTALS 0 0 0 0 0 0 141,663 121,559 114,806 108,428 102,404 96,715 91,342 86,267 81,475 76,948 72,673 68,636 64,823 61,222 57,820 54,608 51,574 48,709 46,003 43,447 $  1,591,122  NOTES: - calculated for the peak periods only  APPENDIX J - Multi-Criteria Evaluation  Page 191  AIR QUALITY SA VINGS WESTBOUND & EASTBOUND  INPUT:  Worst Case  Most Likely  DISCOUNT RATE  Best Case  CHOICE  4%  4%  POLLUTANT COST RATE  0.02  0.04  0.08  0.04  GROWTH RATE  0%  2%  5%  2%  KMS SAVED/YR  899,146  1,458,709  2,003,430  1.458.709  OUTPUT: PRESENT VALUE  FROM 1990 1991 1992 1993 1994  TO 1991 1992 1993 1994 1995  1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015  1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016  PROJECT YEAR  S  267,667  S  963,089  S  3.321,759  KMS SAVED/YEAR 0 0 0 0 0  KMS SAVED WITH GROWTH 0 0 0 0 0  POLLUTANTS SAVED/YEAR 0 0 0 0 0  DISCOUNTED TOTALS 0 0 0 0 0  0 1,574,036 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709 1,458,709  0 1,574,036 1,487,883 1,517,641 1,547,994 1,578,954 1,610,533 1,642,743 1,675,598 1,709,110 1,743,292 1,778,158 1,813,721 1,849,996 1,886,996 1,924,736 1,963,230 2,002,495 2,042,545 2,083,396  0 62,961 58,348 59,515 60,706 61,920 63,158 64,421 65,710 67,024 68,364 69,732 71,126 72,549 74,000 75,480 76,989 78,529 80,100 81,702 83,336  0 62,961 56,104 55,025 53,967 52,929 51,911 50,913 49,934 48,974 48,032 47,108 46,202 45,314 44,442 43,588 42,749 41,927 41,121 40,330 39,555  29,289,507  34,891,764  $ 1,395,671  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  1,458,709  $  963,089  NOTES: - calculated for the peak periods only  APPENDIX J - Multi-Criteria Evaluation  Page 192  CONSTRUCTION COSTS INPUT: DISCOUNT RATE  Worst Case  Most Likely  Best Case  CHOICE  12%  1%  6%  fllllllllll  OUTPUT: PRESENT VALUE | S  FROM 1990 1991 1992 1993 1994  TO 1991 1992 1993 1994 1995  1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020  1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021  PROJECT YEAR  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25  CONSTRUCTION COSTS 7,718,000 11,062,000 8,590,000 13,920,000 20,811,400 27,940,500 14,825,400 354,800  $  127,327,070  S 121^35^3^} DISCOUNTED TOTALS 15,233,963 19,495,024 13,516,531 19,556,598 26,105,820 31,293,360 14,825,400 316,786 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  $  APPENDIX J - Multi-Criteria Evaluation  140,343,482  105,222,100  $140,343,482  Page 193  MAINTENANCE COSTS INPUT: DISCOUNT RATE MAINTENANCE ATI  Worst Case  Most Likely  Best Case  CHOICE  6%  8%  12%  8%  575,900  523,545  471,191  52\545  OUTPUT: PRESENT VALUE  FROM 1990 1991 1992 1993 1994  TO 1991 1992 1993 1994 1995  1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015  1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016  S  7,001,853  S 5,551,461  PROJECT MAINTENANCE YEAR COSTS 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  NUMBER OF LANE-KILOMETRES # oflanes St.Johns W/B 3 Barnet W/B 2 Hastings 3 W/B Hastings E/B 3 Barnet 2 E/B  APPENDIX J - Multi-Criteria Evaluation  S  3,941,875 |  DISCOUNTED TOTALS 0 0 0 0 0 0  523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545 523,545  523,545 484,764 448,855 415,607 384,821 356,316 329,922 305,483 282,855 261,903 242,503 224,539 207,907 192,506 178,247 165,043 152,818 141,498 131,017 121,312  $ 10,470,900  $ 5,551,461  kms 3.02 7.81 7.02 7.05 7.93 TOTAL COST/KM COST/YR  lane-kms 9.06 15.62 21.06 21.15 15.86 82.75 6,326.83 $ 523,545 $  Page 194  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0063791/manifest

Comment

Related Items