|
Transportation Planning for Your Community - System Planning
Click HERE for graphic. PREFACE This publication is part of a series entitled Transportation Planning for Your Community and is designed to acquaint officials and planners with transportation planning for communities of from 25,000 to 200,000 population. The series consists of two guides that explain the concepts of transportation planning and five technical manuals that describe techniques for carrying out transportation planning programs. The guides are: A Guide for the Decisionmaker and The Manager's Guide for Developing a Planning Program. The five technical manuals are titled: Traffic Planning Transit Planning System Planning Monitoring and Forecasting Programming Projects A Guide for the Decisionmaker describes the importance of urban transportation and the benefits of transportation planning. It includes a review of how transportation planning works, and the role of city, county and town officials in transportation planning. The Manager's Guide for Developing a Planning Program describes the principles of transportation planning and is directed to those engineers, planners and administrators who are charged with the responsibility of organizing and administering the transportation planning program. The individual technical manuals describe transportation planning techniques appropriate for small communities. The manuals also include references to other publications that describe appropriate planning techniques. The Traffic Planning manual is a reference of basic traffic engineering techniques,and their potential for improving traffic flow and traffic safety of urban arterial streets and highways. The manual identifies the traffic engineering measures appropriate for consideration in development of transportation improvement plans and programs. The Transit Planning manual includes techniques for estimating transit patronage, service options, and operating requirements. Also included are procedures for evaluating the need for specialized services for the elderly and handicapped. i The System Planning manual details the steps required for the functional classification of streets and highways, the estimation of future traffic, the estimation of the impacts of future traffic, and the estimation of street and highway system requirements. An Appendix includes alternative methods for forecasting traffic. The Monitoring and Forecasting manual provides instructions for assembling inventories of transportation and land activity. It describes methods for monitoring the performance of the transportation system and general community development and methods for forecasting information needed in urban transportation planning. The Programming Projects manual contains procedures for development of the transportation improvement program. Included are procedures for identification of candidate improvement projects, determination of the plan to fund candidate improvement projects, assignment of priorities to candidate improvement projects, budget allocation and project scheduling, and monitoring, adjusting and evaluating the programs. This series was prepared by the COMSIS Corporation and the Highway Users Federation for Safety and Mobility under a grant from the Federal Highway Administration with the aid of a "steering committee" made up of the following officials: Dan C. Dees Illinois Department of Transportation Springfield, Illinois James Echols Tidewater Transportation Commission Norfolk, Virginia David D. Grayson Automobile Club of Southern California Los Angeles, California John J. Holland Cumberland County Planning Board Bridgeton, New Jersey F.W. Landers Department of Public Works Worcester, Massachusetts Marion R. Poole North Carolina Department of Transportation Raleigh, North Carolina ii The principal investigators were: Arthur B. Sosslau COMSIS Corporation Wheaton, Maryland Marshall F. Reed, Jr. Highway Users Federation for Safety and Mobility Washington, D.C. other principal authors were Maurice M. Carter of COMSIS Corporation and Woodrow W. Rankin of the Highway Users Federation. iii TABLE OF CONTENTS Page PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . 1 CHAPTER ONE: SYSTEM FUNCTIONAL CLASSIFICATION . . . . . . . . . . 2 Basic Concept. . . . . . . . . . . . . . . . . . . . . . . . 2 Functional Classes Defined . . . . . . . . . . . . . . . . . 3 Information Required . . . . . . . . . . . . . . . . . . . . 4 Urban Boundary . . . . . . . . . . . . . . . . . . . . . . . 5 Major and Minor Centers. . . . . . . . . . . . . . . . . . . 5 Major and Minor Rural Roads. . . . . . . . . . . . . . . . . 5 Miles of Street and Highway. . . . . . . . . . . . . . . . . 6 Motor Vehicle Travel . . . . . . . . . . . . . . . . . . . . 6 Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . 6 CHAPTER TWO: ESTIMATING TRANSPORTATION DEMAND . . . . . . . . . . 9 Information Required . . . . . . . . . . . . . . . . . . . . 9 Transportation Inventory . . . . . . . . . . . . . . . . . .10 Operating Characteristics. . . . . . . . . . . . . . . . . .10 Area Dynamics. . . . . . . . . . . . . . . . . . . . . . . .13 Travel Characteristics . . . . . . . . . . . . . . . . . . .15 Traffic Estimation Procedures. . . . . . . . . . . . . . . .24 Basic Considerations . . . . . . . . . . . . . . . . . . . .25 Non-Computer Techniques. . . . . . . . . . . . . . . . . . .30 Computer Techniques. . . . . . . . . . . . . . . . . . . . .42 Validating Non-Computer and Computer Technique Results . . .45 Ground Count Projection Technique. . . . . . . . . . . . . .49 Redistribution of Assigned Volumes Among Available Facilities . . . . . . . . . . . . . . . . . . .51 Partial Matrix Technique (PMT) . . . . . . . . . . . . . . .51 Land-Use/Facility Spacing Relationships. . . . . . . . . . .52 Corridor and Site Analysis . . . . . . . . . . . . . . . . .53 CHAPTER THREE: EVALUATING THE TRANSPORTATION SYSTEM . . . . . . .55 Impact Analyses. . . . . . . . . . . . . . . . . . . . . . .55 Social Impacts . . . . . . . . . . . . . . . . . . . . . . .55 Economic Costs and Benefits. . . . . . . . . . . . . . . . .59 Energy Impacts . . . . . . . . . . . . . . . . . . . . . . .66 Environmental Impacts. . . . . . . . . . . . . . . . . . . .66 Evaluating Transportation Service. . . . . . . . . . . . . .66 Evaluating Intersection Capacity . . . . . . . . . . . . . .74 The "Critical Movement Summation" Technique. . . . . . . . .74 Evaluating Corridor Capacity . . . . . . . . . . . . . . . .76 Determining Levels of Service. . . . . . . . . . . . . . . .88 Determining Solutions. . . . . . . . . . . . . . . . . . . .88 REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . .92 iv LIST OF FIGURES Figure Number Page 1 On-Board Auto Survey--Dwelling Unit Summary . . . . . .22 2 On-Board Auto Survey Trip Log . . . . . . . . . . . . .23 3 Relationship of Cordon Count to External Trips. . . . .33 4 Typical Study Area Map. . . . . . . . . . . . . . . . .38 5 Simplified Chain of UTPS Programs for Four-Step Transportation Planning . . . . . . . . . . . . . . . .44 6 District Accessibility Index Contour Plots for 1970 Home-Based Work Trips (Index X 108). . . . . . . . . .60 7 Accessibility to Employment By Automobile During Peak Hours . . . . . . . . . . . . . . . . . . . . . . . . .61 8 Isochronal Representation of Travel Times to Employment Opportunities . . . . . . . . . . . . . . . . . . . . .62 9 Cost of Owning and Operating an Automobile 1976 . . . .65 10 Traffic Data for Intersection Capacity Analysis . . . .77 11 Transportation Corridor Showing Cutlines for Corridor Capacity Analysis. . . . . . . . . . . . . . .79 12 Facilities Stress Diagram . . . . . . . . . . . . . . .80 13 Facilities Stress Diagram with Varying Service Levels .89 v LIST OF TABLES Table Number Page 1 Street and Highway Speed Classification System. . . . .12 2 Suggested Sample for On-Board Auto Survey . . . . . . .19 3 Number of Analysis Zones for Areawide Analyses. . . . .26 4 Number of Assignment Links for Areawide Analyses. . . .27 5 Work Effort Required for Non-Computer Traffic Assignment. . . . . . . . . . . . . . . . . . .43 6 Classification of Impact Areas. . . . . . . . . . . . .56 7 Impact Sub-Elements . . . . . . . . . . . . . . . . . .57 8 Estimated Impacts of Various Alternatives: Harrington, ME. . . . . . . . . . . . . . . . . . . . .58 9 Vehicle Operating Cost on Freeways. . . . . . . . . . .63 10 Vehicle Operating Cost on Arterial Streets. . . . . . .64 11 Urban Accident Frequency on Surface Streets Per Million Vehicle Miles . . . . . . . . . . . . . . .67 12 Vehicle Gasoline Consumption on Freeways. . . . . . . .68 13 Vehicle Gasoline Consumption on Arterial Streets. . . .69 14 Pollutant Emission Factor Components (1977) . . . . . .70 15 Pollutant Emission Factor Components (1995) . . . . . .71 16 Intersection Capacity by Level of Service . . . . . . .75 17 Freeway/Expressway Capacity Measures (Total Vehicles Per Hour) . . . . . . . . . . . . . . .81 18 Possible Intersection Capacity for Various Operating Conditions--Approach Width 20 Feet. . . . . . . . . . .83 19 Possible Intersection Capacity for Various Operating Conditions--Approach Width 30 Feet. . . . . . . . . . .84 20 Possible Intersection Capacity for Various Operating Conditions--Approach Width 40 Feet. . . . . . . . . . .85 vi LIST OF TABLES (Continued) Table Number Page 21 Possible Intersection Capacity for Various Operating Conditions--Approach Width 50 Feet. . . . . . . . . . .86 22 Possible Intersection Capacity for Various Operating Conditions--Approach Width 60 Feet. . . . . . . . . . .87 vii INTRODUCTION This manual includes transportation system planning techniques that are applicable to small urban areas. This System Planning manual is complemented by a separate System Planning Appendix that describes, in more detail, transportation demand estimation techniques. To carry on effective transportation system planning this manual should be used with the Appendix and indicated reference material. System planning is an evaluation of the need for transportation facilities and services based on travel demand. System planning may be accomplished on an areawide basis, within a jurisdiction, in a transportation corridor or in any other geographic unit. System planning includes an evaluation of how urban development and human behavioral characteristics change the demand for transportation. System planning techniques are used to evaluate transportation alternatives. The starting point for the analysis of transportation alternatives should be the present system and committed transportation improvements. To simplify system planning the evaluation should be limited to a single land-use plan. Both problem identification and problem solution are integral parts of system planning. Solutions may include public transportation service improvements; transportation system management (TSM) actions, such as traffic engineering improvements, a ridesharing program or parking restrictions; or capital investments related to widening street and highway facilities and other construction. Other manuals of this series on Transportation Planning for Your Community complement this manual. For detailed evaluation of existing system operations, refer to the Traffic Planning and Transit Planning manuals. The Programming Projects manual describes how improvements can be programmed for implementation. Chapter One of this manual describes system functional classification. Included is information on the basic concept of functional classification, definitions of functional classes, information required, and a step-by-step procedure. Chapter Two describes techniques for estimating transportation demand. Both non-computer and computer techniques are covered. Information required for system planning and simplifications for data collection are provided. Chapter Three addresses techniques for estimating system requirements and determining improvements. Impact analyses covering social, physical, economic and environmental effects are presented. 1 CHAPTER ONE SYSTEM FUNCTIONAL CLASSIFICATION Functional classification is the grouping of roads, streets and highways into integrated systems, each ranked by its importance to the general welfare, the motorist, and the land-use structure. Concepts of "importance" are based on economic and social values, measured in a variety of ways. The functional classification process is a practical technique for determining the travel corridors that should best serve through and local traffic in an urban area. The process determines the importance of all urban streets and highways in relation to one another and to urban development. In general, the segregation of the urban street and highway system into classification provides the legislator, highway administrator, engineer and transportation planner with a foundation for: (a) establishing administrative highway systems: (b) designating minimum design standards for highway systems; (c) evaluating critical highway deficiencies; and (d) apportioning highway funds. For a detailed explanation of the uses of the functional classification of streets and highways, see the Programming Projects manual. BASIC CONCEPT In developing a functional classification plan, all streets and highways are examined and grouped into functional classes, which are ranked by their importance. The objective is to define appropriate relative purposes of highways and streets in providing traffic service and influencing urban development, and to establish the most economic yet beneficial systems to meet both present and future transportation needs. The concentration of traffic on a limited arterial system is a basic goal of functional classification. Fiscal resources should be similarly concentrated on the limited arterial system to improve the level of service to attract a maximum portion of through traffic. Functional classification also is based on the concept that most streets and highways have a predominant process - either to provide the motorist with access to abutting land or to allow movement through an area. Traffic that gains access to abutting land is termed "local," whereas all other traffic is termed "through." In any designated neighborhood or community area, through traffic is that which neither originates nor is destined there, but is simply passing through. Local traffic has origin or destination within the designated area. Thus, the same traffic that may be considered "local" at its starting or ending place may be called "through" in communities in between. 2 With the exception of limited-access highways where local traffic is prohibited, or "cul-de-sacs" where through traffic is impossible, all other streets and highways carry varying amounts of both through and local traffic. However, a facility that is used predominantly by through traffic is termed an arterial street or highway, and one that is used predominantly by local traffic is termed a local-access street. This split between through and local traffic is different for each travel corridor and each link in the transportation network. Experience has shown that the location of centers of traffic attraction are key factors in the designation of arterial streets and highways. Since centers attract traffic from throughout an urban area, a graphic display of travel "desire lines" would reveal a "star pattern" emanating from each center. Because of the overlapping of travel desires, the corridors between the centers are usually the heaviest traveled in the urban area. Therefore, the connection of these centers by the best available and most logical regional routes forms the arterial classes of streets and highways. Collector streets collect traffic from local-access streets and distribute it to the system formed by arterial streets and highways and vice-versa. FUNCTIONAL CLASSES DEFINED There are three street and highway classes - Major Arterials, Minor Arterials and Collectors - where through traffic predominates, and a fourth class - Local Access - where local traffic is predominant. Major Arterial streets and highways contain the greatest proportion of through or long-distance travel. Such facilities serve the high-volume travel corridors that connect the major generators of traffic. The selected routes should provide an integrated system for complete circulation of traffic, including ties to the major rural highways entering the urban area. Experience has shown that this class normally accommodates 30-40 percent of a region's travel on 5-10 percent of the street and highway network. Generally, Major Arterials include all the higher traffic volume streets, except those serving short trips or those serving as alternatives to more direct facilities (i.e., interstate, freeways and expressways, and other principal arterials). Minor Arterial streets and highways connect with all remaining rural arterial and Collector roads that extend into the urban area. Minor Arterial streets and highways serve less concentrated traffic-generating areas such as neighborhood shopping centers and schools. This class distributes medium traffic volumes. Minor Arterial streets serve as boundaries to neighborhoods and collect traffic from Collector streets. Although the predominate function of Minor Arterial streets is the movement of through traffic, they also provide for considerable local traffic that originates or is destined to points along the corridor. 3 The integrated system formed by Major and Minor Arterial streets and highways should include 10-20 percent of the total street and highway network and serve 40-60 percent of motor vehicle travel. Collector streets provide direct service to residential areas, local parks, churches, etc. To preserve the amenities of neighborhoods, they are usually spaced at about half-mile intervals to collect traffic from Local-Access streets and convey it to Major and Minor Arterial streets and highways. Collector streets serve as local bus routes. Direct access to abutting land is essential; parking and traffic controls are usually necessary to insure safe and efficient through movement of moderate to low traffic volumes. The integrated arterial and collector system usually includes 20-25 percent of the street and highway system and serves 70-80 percent of all urban area motor vehicle travel. Local-Access streets are those not selected for inclusion in the arterial or collector classes. They allow access to individual homes, shops, and similar traffic destinations. Direct access to abutting land is essential, for all traffic originates from or is destined for abutting land. Through traffic should be discouraged by using appropriate geometric designs and traffic control devices. INFORMATION REQUIRED The "basic resources" required are: 1. A map, scaled between 1" = 400' and 1" = 1/2 mile, showing existing streets and highways; 2. Existing average daily motor vehicle traffic counts on all urban streets and highways as well as the rural roads that converge on the urban area; and, 3. Personal knowledge of both the land development pattern and the general characteristics of the street and highway system. To supplement personal knowledge aerial photographs and/or a map of existing land use is helpful. The basic resources will enable the determination of the following "other" essential information: 1. The approximate boundary between urban and rural land development 2. Major and minor urban centers 3. Major and minor rural roads entering the area 4. Total vehicle miles of urban motor vehicle travel 4 Urban Boundary Using information about the land-use pattern, the approximate limit of urban development can be determined. Included within the urban boundary: 1. Residential areas of over 1,000 persons per square mile; 2. Strip commercial development; and 3. Other development of an urban character such as shopping centers, industrial parks and recreation areas. Committed land development that meets the above criteria should be included within the urban area boundary. Major and Minor Centers A key factor in determining the magnitude of through traffic is the location of the ma or and minor urban centers. These are located through personal knowledge of the area or as revealed in the study of land-use maps and aerial photography. The term "major" is given to centers attracting heavy volumes of traffic. They include city core areas, regional shopping centers, industrial parks, port facilities, colleges and universities, large individual industrial plants and military establishments. Major centers are best served by the Major Arterial class. Neighborhood shopping centers, small industrial plants, high schools, etc., are considered "minor" urban centers, and are serviced by Minor Arterial streets. Although traffic generation is the primary criterion in ranking these centers, other factors must be considered. For example, not all trips have the same economic impact. An airport might be ranked as a "minor" center in terms of relative trip generation, but its economic importance may justify a higher designation. Major and Minor Rural Roads The rural arterial and collector roads entering the urban area must be.identified and segregated into major and minor classifications based on existing statewide functional highway classifications as designated by State and county officials, or on estimates of statewide function or on traffic volumes. The "major" classification should be reserved for all routes of principal statewide significance that serve heavy volumes of long-distance travel, such as Interstate Highways. Typically, major rural routes entering an urban area serve an average of 5,000 or more motor vehicles per day. 5 The "minor" designation is given to all other rural arterial and collector routes that enter the urban area. Rural roads that are chiefly serving as land-access routes and enter the urban area are classified as urban Local-Access streets. Miles of Street and Highway The total miles of urban street and highway are determined through the use of a properly calibrated map-measure wheel and the map previously described as Resource 1, with the urban boundary affixed thereon. It can also be obtained by driving all streets and highways within the urban boundary with a vehicle equipped with a calibrated odometer of the type used in developing road inventories. Total urban area street and highway mileage can also be determined by adding the section lengths of streets and highway identified in a road inventory. If the urban boundary was established subsequent to the inventory, measurement will have to be made of the length of the urban segment of each street and highway inventory section cut by the boundary and @his mileage added to the total length of the inventory sections wholly contained within the boundary. For further information on the availability of a map-measure wheel, specially equipped vehicles or road inventories, check with State transportation officials. Motor Vehicle Travel Calculate existing average daily motor vehicle travel on individual segments of urban streets and highway by multiplying the average daily traffic counts, or count estimates, on each urban route segment by the length of the segment. Add the calculated segment travel to determine total urban motor vehicle travel. PROCEDURE The functional classification process is sequential and the following steps are suggested: STEP 1 - Obtain or develop the "basic resource" information described previously. STEP 2 - Determine "other essential information"'described previously. STEP 3 - Locate on the street and highway base map the major and minor urban centers. STEP 4 - Locate on the street and highway base map the major and minor rural arterial roads converging on the urban area and crossing the urban boundary. 6 STEP 5 - Select as the Major Arterial streets and highways those that connect the major urban centers and that connect the major urban centers with the major arterial roads converging on the urban area via the best and most direct routes. STEP 6 - Select as the Minor Arterial streets and highways those that connect the minor and major urban centers to each other and to the rural minor and major arterial roads that enter the urban area. The selected Major and Minor Arterial streets and highways should form a continuous network. Major and Minor Arterials should be designated at approximately one-mile intervals in fully developed areas. It may be desirable to use a transparent overlay of the street and highway base map with a grease pencil to depict, study, review, and revise the Major Arterial and Minor Arterial and Collector classes. STEP 7 - Locate on the street and highway base map the neighborhood residential areas, many of which are bound by the system on Major Arterial streets and highways. STEP 8 - Select the Collector streets that collect and distribute traffic originating along or destined to homes along the Local Access streets of the neighborhood. To avoid circuitous routing of travel, it may be necessary to select two Collector streets in some neighborhoods. To prevent Collector streets from being used as arterial routes avoid selecting Collector streets in contiguous neighborhoods that form a continuous route. Collector streets should be spaced as nearly as possible to the midway point between arterials to be convenient to motorists. STEP 9 - Determine by measurement or from road inventory files the mileage of the Major Arterial, Minor Arterial and Collector classes. Determine Local-Access street mileage by subtracting the total Arterial and Collector mileage from the total street and highway class. STEP 10 - Calculate the proportion of street and highway mileage in each functional class. STEP 11 - Determine the average daily motor vehicle travel on the Major Arterial, Minor Arterial and Collector classes. Determine the Local-Access travel by subtracting the total Arterial and Collector travel from total urban area motor vehicle travel. STEP 12 - Calculate the proportion of urban area motor vehicle travel on each functional class. 7 STEP 13 - Calculate the average traffic density on each functional class by dividing the motor vehicle travel on a class by the class mileage. STEP 14 - Review and adjust functionally classified urban street and highway network. The importance of using a series of overlays with colored grease pencil tracings of the Arterial and Collector classes is seen in the review and adjustment process. Experience shows that several iterations are required to develop a functional classification plan that meets the criteria outlined and meets with the approval of citizen and official advisors. 8 CHAPTER TWO ESTIMATING TRANSPORTATION DEMAND The basic concept underlying material provided herein is that on an urban areawide scale, system analysis should be directed to the identification of current and potential problems, Areawide analysis should, generally, not be used for a testing of alternative land-use plans or alternative systems. Some of the methods can be used for such purposes, but based upon current concepts concerning transportation planning in smaller areas such analysis does not appear warranted or desirable in most cases. The approaches provided in this chapter for demand estimation consist of both non-computer and computerized methods, and methods which vary in both scope and scale of analysis. The selection of an appropriate approach should be based on the concepts described in Chapter One of the Managers Guide for Developing a Planning Program, "Determining the Planning Scope." Each of the methods described will include considerations with regard to the scale and scope of analysis most appropriately addressed with the technique. In addition to describing transportation estimation procedures that can be used to simplify demand estimation, this chapter first provides discussion relative to information required and simplifications that can be realized in data collection. INFORMATION REQUIRED Socioeconomic and transportation information is required for demand estimation, both in the development of relationships for analysis purposes and to assess current and potential problems. The collection of data for demand estimation should be closely tied to the data required for monitoring, both of which should be keyed to the level of effort determined to be appropriate for the area under study. Hence, there is no single design relative to information requirements for all urban areas. The reader is referred to the Monitoring and Forecasting manual for more detailed information concerning monitoring requirements, forecasting socioeconomic information, etc. For demand estimation purposes, information is required in several broad categories as listed below: - Transportation Inventory - System Operation - Area Dynamics - Travel Behavior This information is central to the study of the current transportation situation as well as to the estimation of future demand and resulting potential problem areas. 9 Transportation Inventory A transportation inventory is the central requirement for transportation planning and operations. This inventory should include both the highway and public transportation systems, Inventories related to the public transportation system are described in the Transit Planning manual. An initial step related to highway system inventories is the functional classification of streets and highways. This process is described in detail in Chapter One of this manual, "System Functional Classification." The physical features of the highway inventory should generally include: - Physical features, such as width of pavement, condition, and surface type - Traffic and engineering features, such as number of traffic lanes, traffic controls, channelization, parking regulations, and pedestrian control Much of the above information should be available through local agencies such as the Department of Public Works or Traffic Department. In many cases, some of the information will be available through the State Department of Transportation. Quite often, however, the information will not be compiled or centralized to allow a thorough evaluation. The information is central to answering questions relative to determining street needs, short range planning, maintenance budgeting, etc. For maximum utilization of such data, it is best to transfer the data to a computer format such as punched cards or-magnetic tape. There are many sources available concerning the development of street inventories appropriate for operating agencies. For information about the development of a highway system inventory, the reader is referred to manuals: 1A, Determining Street Use; and 5A, Inventory of the Physical Street System; of the series Better Transportation for Your City l/. System Characteristic The operating characteristics of the street and highway system determines the existing levels of service. Included are traffic volume counts, travel times or speeds and the capacity of streets and highways. The operating characteristics of the street and highway system are key determinants of problems and are basic information required for estimating future traffic demand. Good traffic counts are required for the ground count based forecasting process described (See section entitled, "Ground Count Projection Technique."). Likewise, system travel times/ speeds are required if a computerized traffic assignment is to be accomplished. Traffic capacity values are required to assess current and predicted traffic volumes. The operating characteristics should form part of the computer format discussed above for the transportation inventory. Traffic Counts. Traffic counts are needed in several phases of planning. They are an indicator of current problems when related to capacity. They are used in calibration of models and as a check on the results of the models. They 10 are used as a base to forecast future traffic as described in the traffic count based forecasting procedure. Monitoring of traffic volumes will indicate travel growth trends. Sampling methods are used in obtaining average daily volume information since obtaining such volumes for an entire system might be beyond the capability of the transportation planning program. The reader is referred to two sources: - Manual 3A, Measuring Traffic Volumes 1/ - Guide to Urban Traffic Volume Counting 2/ Travel Time. Travel time is an important indicator of system operation and traffic service. Travel time studies, when compared to traffic counts, provide a measure of congestion and also are a measure of vehicle operating costs. Travel time data is necessary in the traffic forecasting phases of a study if computer methods are used for trio distribution and traffic assignment. Travel time "runs" should be made on streets and highways where there are apparent problems. Such"'runs" will pinpoint problems. If travel times are to be made for traffic assignment purposes, it is suggested that a sampling of major routes by area type be undertaken. The classification of speeds (with respect to highway facility vs. location) might be set up as shown in Table 1. The speeds so tabulated can be used to assign speeds to routes not inventoried. Instructions for a study to determine travel time are contained in Manual 3B, Determining Travel Time l/. For study of individual routes, the recommendations concerning the "Required Number of Runs" as described in Manual 3B, are appropriate. However, if coverage is to be undertaken for traffic assignment purposes, then it is recommended that two runs be made in the A,M. peak hour, two runs in the P.M. peak hour, and two runs in off-peak hours (one in the afternoon and one in the evening). Then, to determine an ADT travel time, a formula 3/ that has been used is: 2(Off-Peak Travel Time) + l(Peak Travel Time) ADT Travel Time = ________________________________________________ 3 Traffic Capacity. Traffic capacity determination for streets and highways allows: a determination of present level of service by comparisons with ground counts; and an estimation of future adequacy through Comparisons with forecast volumes. The three major factors influencing capacity are: (1) geometrics (2) traffic characteristics, and traffic controls for intersections (streets) (3) lane widths, and traffic characteristics on controlled access facilities. Procedures for calculating capacity for traffic engineering purposes are described in the Highway Capacity Manual 4/. For traffic assignment and other planning 11 Click HERE for graphic. 12 purposes, less detailed procedures for calculating capacities are appropriate, and are discussed in Chapter Three of this manual. Where parking is an important problem refer to Manual 3C, Conducting a Limited Parking Study 1/, and to the Traffic Planning manual. Area Dynamics The transportation system serves the population and industry of an area, Transportation planning must be responsive to the dynamics of an area in terms of population and employment change, characteristics of the population and employment, such as income, auto ownership, and type of industry, and the distribution of the population and employment across the area. These characteristics must be considered throughout the transportation planning process. For example, trip generation establishes a relationship between travel and the land-use and socioeconomic characteristics. In trip generation the intensity of land-use, the character of the land-use, and location should be considered. Intensity of land-use is the amount of activity to be found in a zone and is usually stated in terms of a density measure, such as employees per square foot of floor area, acre of some specific land-use category, or dwelling units per acre. As an example, the number of trips per dwelling unit generally decreases as the number of dwelling units per residential acre increases. Trip generation also relates to the "character" of land-use. On a household level, character is expressed in socioeconomic terms such as family income and auto ownership. With all other conditions the same, families with higher incomes generally own more automobiles and make more trips. For non-residential land uses, character is usually reflected in the type of activity (e.g., manufacturing, retail, commercial). The location of residential land is important as may be shown by the higher trip rates of a high-rise complex in the suburbs versus rates for a similar complex in the CBD. For trip distribution, area dynamics is considered in terms of the trip generation throughout the metropolitan area and the separation between the activities in terms of time or distance. In mode choice, or transit use estimation, area dynamics is considered in terms of characteristics of the population such as age, income, auto ownership, distance from transit lines, and so on. Sources of Information. The characteristics of the area can be evaluated through the use of available sources of information including topographic data, economic data, demographic/social data, land-use/highway data, and employment data. This information is used to evaluate the current situation as well as a forecast of the probable future situation. 13 The major sources of information are; - Local Tax Assessment Records - Planning and/or Zoning Commission Records - Housing Authority Files - Building, Occupancy, and Construction Permits - Building and Health Inspection Records - Business Licensing Files - Employment Security Commission Records - Bureau of Revenue Files - School District Student Enrollment and Personnel Files - Community Renewal Agency Files - Utility Billings - Motor Vehicle Registrations - U.S. Census Materials - Internal Revenue Service Records - Bureau of Labor Statistics Data - State Planning Commission Files - Community Organization Records - Commercial Data Sources (such as Polk) - Aerial Photography - Public Works Department Files The use of available information is more economical and efficient than the collection of new data. These sources also provide opportunities for monitoring key urban area characteristics (See Monitoring and Forecasting manual.). Forecasts. When considering future projections of social, economic, and demographic characteristics, such factors as population trends, economic development trends, land-use development trends, and geographical and topographical constraints to land and transportation development should be examined. In areas with slow to moderate growth, a single projection for transportation analysis will, in most cases, be appropriate. This projection should be an assessment of what the area will most probably look like in the future. For faster growing areas,, policies for controlling and/or directing growth may be considered. However, growth patterns and trends are difficult to redirect over relatively short periods, such as 5-10 years, and care should be taken to limit transportation testing of alternative growth possibilities. The transportation analysis should be aimed at determining potential problems and system requirements for the most probable future growth and development pattern. Use of map displays and charts, along with land-use plans and economic and demographic projections, provide a starting point for assessing urban area characteristics. Emphasis should be placed on the,chief economic basis for the area and likely residential development patterns. Discussions with key community leaders will often provide a good assessment of future development patterns. Land-Use Forecasting Techniques. Land-use forecasting refers to estimating future amounts of traffic analysis areas such as zones. Several non-computer land-use forecasting techniques have been developed that are 14 practical for transportation planning in small urban areas. In some instances, such as where rapid growth is expected, a simplified computer based procedure may be considered. Land-use forecasting allocates areawide forecasts of the socioeconomic variables to analysis zones. Consideration is given to such zone characteristics as vacant land available for development, zoning, availability of public utilities, accessibility, etc, The traditional allocation technique consists of gathering and analyzing data and then allocating activities based on acceptable planning standards and professional judgment. A further step is to use mathematical models and regression analysis to quantify existing relationships. While this added step makes the allocation process more explicit, planning judgment still plays an important role in the use of the models. For details on land-use forecasting refer to Land Use Forecasting Techniques, for Use in Small Urban Areas 5/. Travel Behavior The amount and type of information collected for understanding travel and travel behavior should be based on assessments of both growth potential and transportation problems. Transportation inventories provide the necessary information-- traffic volumes, speeds, capacities, condition of the streets--for the identification of problems. However, where traffic congestion in a downtown area appears to be caused by trips which have no purpose downtown, a travel survey--limited to the downtown area-- may be warranted. Future estimates of travel must be made through use of the four- step simplified procedure, through ground count factoring, or other suitable traffic forecasting procedures. There may be other instances where travel survey information must be collected, but in most cases, travel behavior and travel can be simulated. The material provided on demand estimation in this chapter allows simulation of most travel characteristics. However, a review of past surveys should be made, and such surveys used as a base, where appropriate. Where other travel surveys must be made, simplification and cost reduction are important considerations. External travel. Since through trips and trips originating or destined to points outside the urban area are a high proportion of total travel in smaller urban areas, an external survey may be necessary. In some cases an alternate approach based on a trip table built from screenline or cordon surveys should be considered. This is described in the section of this chapter on the "Partial Matrix Technique." 15 External travel surveys are described in Urban Origin- Destination Surveys 6/. Most states have personnel with expertise in setting up roadside interview stations, conducting the interviews, counting and classifying traffic, and coding the results. The traditional external survey begins with the development of a traffic cordon line around the study area, Either a loose cordon, literally including rural fringes within it, or a tight cordon, containing only the urbanized area, may be drawn. Each has its advantages and disadvantages, and these are discussed in Urban Origin-Destination Surveys 6./. If the external travel survey is the only survey to be made, a tight cordon will provide more information, but is more expensive, since many routes cross the cordon. If the survey is an update of a previous survey, the same cordon location should be used. If a bypass highway is being considered, the cordon should include the bypass area, The cordon should be located to keep the number of interviews as low as possible. A tight cordon makes repeat surveys more difficult in areas that are contemplated to expand. Once the cordon is located, stations are selected for interview. The stations are selected, traditionally, to account for 85-95% of all traffic crossing the cordon, although in some cases, a smaller sample may be adequate. If some stations are close to each other and are believed to exhibit similar distribution patterns, interviewing on one and using the distribution to represent another may be acceptable. In this manner the number of stations may be minimized. Likewise, for most regional studies, sufficient distribution patterns can be obtained by interviewing at stations that account for approximately 75% of all traffic crossing the cordon. It is recommended that interviewing be accomplished during a 12-14 hour period containing both peak hours (A.M. and P.M.) at all stations. If the external travel survey is being made to forecast traffic, and time of day is not important, it is recommended that interviewing be accomplished in one direction only. This is based on the assumption that inbound and outbound traffic patterns are similar. A counting program must accompany the external travel survey at all external stations and other cordon crossings. For higher volume routes, daytime classifications of vehicle types are also made. On major truck routes, vehicle classification counts may also be necessary at night, when truck traffic predominates. Classification counts permit expansion of the travel sample. In an update situation, new counts in the same locations the external survey was previously conducted, may be an adequate substitute for a new survey-especially in an area with few major land-use changes. The trip distribution from the external survey can be factored using the new count information. There are cost advantages to using automatic counters wherever possible. Recently, problems have occurred in external travel surveys because some drivers avoid survey stations due to forewarning on Citizen Band radios. Care 16 should be taken to monitor that media and observe key diversion points, especially for major highways. Care in avoiding congestion and time delays will eliminate these problems. If the sample is kept small, congestion and delays are minimized, CB diversions avoided, and the cost of handling/coding/keypunching minimized. For stations where the traffic volume is under 10,000 vehicles per day, a sufficient number of personnel should be assigned to each station to interview about 25% of the off-peak traffic. For routes with volumes over 10,000 vehicles per day, about 10% of peak hour traffic should be sampled. If trip length by purpose as well as trip distribution are the major goals of the survey, vehicle type, origin, destination, number of vehicle occupants, garage location, and trip purpose are sufficient. Routing of through trips should also be determined if needed for the development of forecasts or analysis of particular problems. Sample survey forms are shown in Urban Origin- Destination Surveys 6/. Alternative methods of gaining external travel information should be considered. Some are simpler and less expensive, but provide less reliable, but still acceptable, data. The easiest method is a postcard survey, in which a percentage of vehicles is stopped and the driver handed a postcard to fill in and mail back. This typically results in a return rate of about 40%, with the inherent bias of a mail-back survey. However, traffic delays are minimal and manpower requirements are less. For a postcard survey, station locations at stop signs or traffic signals are an advantage. Another option, is a license plate look-up, mail-out postcard survey. License numbers are recorded in full on a tape recorder, on film, or on paper, and matched with motor vehicle files to develop a two-part postcard addressed to the vehicle owner, explaining the survey and where the vehicle was seen, with the form as a detachable return card. These surveys must be made with full cooperation from the State motor vehicle departments. If the simulation of internal travel is contemplated, as discussed in the section on the "Partial Matrix Technique," the possibility of interviewing motorists on two perpendicular screenlines instead of a cordon should be considered. With two perpendicular screenlines, from one edge of the urban area to the other, all through trips and a large portion of both the external- to-internal trips and internal-to-internal trips will be surveyed. The most economical substitute for an external travel survey is a previous external travel survey, expanded through the use of current traffic counts. The expansion of a previous survey should be considered when: - through traffic is not a major traffic problem, - major through or bypass highways have not been built or improved since the last survey, - the cordon for the last survey is still outside the urbanized area, and 17 - major generators of external travel@ such as a new shopping center on the outskirts of town, have not devel- oped since the last survey, If some change has occurred which affects travel, it may be possible to undertake a small survey to check for that impact, such as an external survey only on the newly improved through route, or only in the part of town where the growth has extended beyond the cordon, Internal Travel, Internal travel, that is, travel with both origin and destination within the boundary of the urban area, has traditionally been forecast using distribution models such as the gravity model. Early in the use of distribution models, large samples of travel information were collected to form the basis for the calibration of models. The approach to internal travel estimation recommended in this manual is simulation based upon characteristics of similar communities. Characteristics are selected that relate travel to local land-use and transportation facilities. "Calibration" consists of matching the results of the simulation models to local traffic counts and mode choice. With this approach there is no need to make travel surveys. Where conditions are such that it is necessary to obtain local travel information, alternative methods to a home-interview survey should be considered. Alternative methods to the home-interview-survey are less costly and provide the necessary information. Two such alternative travel behavior surveys are discussed below. One approach is based upon an "on-board auto survey" coupled with an on-board transit survey, if appropriate. Here, a small sample survey, in the range of 500-1,000 households provides sufficient information (See Table 2.). Methods of collection of travel information should be carefully considered, with questioning held to a minimum. Telephone surveys or mail-back interviews should be considered. The other approach is based upon developing a trip table with resulting trip lengths, productions and attractions from screenline surveys expanded to represent the entire area. For truck and taxi trip information, the on-board survey, the screenline Survey or a small sample survey of the type described in Urban Origin-Destination Surveys 6/, should be considered. If there are less than 10 taxis operating, the taxi survey should be eliminated. If a taxi survey is to be undertaken, include all taxis. For trucks, the sample size should be based upon the intended use of the data. If the information is to be used for determining trip length and trip generation, the overall truck sample should be twice the home-interview rate 6/. 18 TABLE 2 SUGGESTED SAMPLE FOR ON-BOARD AUTO SURVEY Sample Households Approximate Approximate Urban Area (Assumes 3 % of Number of Number of Population persons per HH)* Households Autos** 25,000- 50,000 500 4.0 775 50,000- 75,000 550 2.6 850 75,000-100,000 650 2.2 1,010 100,000-125,000 750 2.0 1,165 125,000-150,000 850 1.8 1,320 150,000-175,000 950 1.7 1,475 175,000-200,000 1,000 1.6 1,550 * Source: Urban Origin-Destination Surveys 6/. ** Using an average of 1.55 autos per household and rounded to the nearest 5 households. 19 On-Board Auto Survey, Where local travel information is deemed necessary for either monitoring travel characteristics or for the development of travel models, the information should be obtained at minimum cost and within as short a time period as possible. One such approach is the on-board auto survey, The on-board auto survey is appropriate to determine trip generation, trip length frequencies, and trip purposes. The on- board auto survey parallels the onboard transit survey described in the Transit Planning manual. In addition, to information on trip characteristics, the on-board auto survey measures vehicle use. If transit and highway analyses are to be done independently, then the on-board auto and transit surveys should be separate. If a person trip analysis is desirable. the data sources should be combined. The methodology for combining the data sources is described later. The on-board auto survey combines telephone interviews with a vehicle trip log for each auto in a selected sample. The telephone inquiry solicits household participation and gathers basic socioeconomic information. Sampling is by household, with all vehicles in a selected household part of the survey. Sampling of households is accomplished as described in the manual, Urban Origin-Destination Surveys 6/. As a minimum, it is suggested that 500 households be interviewed, with an upper limit of 1,000 households. Refer also to Table 2. The non-trip information includes the type of data required for anticipated analysis. If, for example, trip generation analysis using cross-classification is contemplated, auto ownership, household income and persons in the household should be obtained. As a minimum, the following non-trip information should be collected and coded: - Interview address - Number of owned passenger cars at address - Number of company-owned or borrowed cars at address - Number of trucks used for personal (non-commercial) trips - Number of persons living in household - Number of persons 5 years old or older living in household - Combined income of all family members In addition, other useful information includes: - Household structure type - Number of persons employed - Sex, race, occupation and industry of each family member - Make, model and year of each auto at household 20 - Number of transit trips by trip purpose made by each family member on the previous day For auto trip data, a vehicle log should be sent for each auto owned or garaged at the household, and for each truck used for personal travel. The vehicle log should be completed for a given 24-hour survey day, The survey day and date should be entered on the forms sent, as well as the household address, Also, if vehicle information (make, model, year, etc.) is requested in the telephone interview, this should be entered on each form sent to the household. The general guidelines for the time of beginning and ending of the travel date (i.e., 4:00 A.M.) should be as described in the Urban Origin-Destination Surveys 6/. The log data for each vehicle in the survey should be limited to the following basic trip information: - Start and end odometer reading for each trip - Start and end time of each trip - Purpose of each trip (plus homebased or non-homebased identifier) - Number of people in vehicle (household members and non- household member passengers) for each trip. A trip for the purpose of the survey is defined as the movement of the vehicle between two points, where a driver or passenger enters or leaves the vehicle at one of the points. As may be noted, information is not obtained with regard to the origin or destination of the trip. This attribute of the technique as well as the minimal interviewer time involved are the major contributors to the low cost per survey. Since the primary focus of the procedure is for model development or monitoring of trip length and trip rates, geocoding of the observed trips is not necessary. It is suggested, however, that the household address be coded to some geographic system. There are many survey form designs possible for the on-board survey. One such design is shown in Figures 1 and 2. These forms allow the determination of the number of auto driver trips by purpose, as well as the number of auto passenger trips by purpose. These may be classified by income and auto ownership and household size. The steps in accumulating the numbers of trips and households to determine, the trip rates is described in Trip Generation Analysis 7/. For trip length distributions,the number of auto driver trips can be accurately determined using the odometer readings for distance in miles and the reported times for travel time in minutes. Trip purpose can be determined directly. An auto occupancy by trip purpose of the auto driver can also be determined. The trip length by purpose for passengers cannot always be accurately determined. If, for example, passengers enter and leave during a trip reported for one auto driver (i.e., driver starts with three passengers, one more is picked up along route, two passengers dropped off further along route, etc.) a determination cannot be made of trip length by purpose of the passengers since the individual passengers 21 Click HERE for graphic. 22 Click HERE for graphic. 23 are not identified. By eliminating reportings of this kind, it is expected that the use of more usual reportings will reflect the universe of travel. If total person trip rates classified by socioeconomic variables (i.e., income, auto ownership) are desired, then the results of an on-board auto and an on-board transit survey are required. In such a case, the results of the on-board auto and on- board transit surveys must be expanded to reflect the universe of trips. Expansion of the on-board auto survey is accomplished in a manner similar to a usual home-interview survey, and the reader is referred to the Urban Origin-Destination Surveys 6/. For expansion of an on-board transit survey the reader is referred to Urban Mass Transportation Travel Surveys 8/. Using the on-board auto survey, as expanded, an estimate is obtained of the number of households by classification category (i.e., income and auto ownership). Likewise, the number of auto driver and auto passenger trips by classification category is determined. The expanded transit survey is also summarized to provide transit trips by classification category. When these matrices are developed, trip rates per household can be obtained by classification category. TRAFFIC ESTIMATION PROCEDURES In this section, the four-step traffic estimation procedure-- trip generation, trip distribution, mode choice and traffic assignment--is described along with alternative techniques for carrying out the procedure. Non-computer techniques are described, mostly by referencing the report entitled: Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters--A Users Guide 9/. Where appropriate, guidelines, hints and simplifications are provided relative to the manual techniques described. A computerized approach is also provided using a minimum number of programs and program options in the Urban Transportation Planning System (UTPS) 10/. The computer application uses a simplified pre-packaging of programs HR (historical record builder), UROAD (highway traffic assignment program), SCAGM (small cities gravity model program) and UMATRIX (matrix manipulation program), as well as inclusion of default paramenters for three sizes of urban areas (i.e., 25,000-50,090 population- 50,000- 100,000 population- and 100,000-200,000 population). 24 Basic Considerations Some basic decisions must be made prior to application of the traffic estimation techniques. These are related to reasons for the analysis, the zone system required, the network detail required, and the choice between manual or computerized approach. Details can be obtained through reference to special manuals on the subjects, such as Traffic Assignment 3/, Trip Generation Analysis 7/, PLANPAC/BACKPAC General Information 11/, etc. For definitions of the terms used in this chapter the reader may refer to the Glossary in PLANPAC/BACKPAC General Information 11/. Number of Traffic Analysis Zones. The number of zones to be used in a transportation network should be based on the purpose of the evaluation and the method of analysis to be used. For areawide analyses to determine potential future problems, zones with up to 30,000 trip ends produce reasonably accurate results. For corridor analysis or major route planning, the number of trip ends per zone should be in the range of 10,000-15,000. The accuracy of trip distribution results is not affected significantly by reduction in the number of zones. Decisions regarding zone systems should be based on the level of acceptable accuracy for the traffic assignment phase of a study. The number of zones should be kept within a maximum of about 40 if non-computer network analyses are to be made. To determine potential future problems on an areawide level, it is suggested that the number of analysis zones approximate that shown in Enable 3. The reader is referred to Traffic Assignment Methods--Applications-Products 3/ for information and definitions regarding zone and analysis network development. Analysis Network. A general guideline for the number of links that should be in a computer network is 8-10 physical (non- directional) links for one analysis area (considering each analysis area has a connection from its centroid to four boundary streets). For areawide analysis to determine problem areas for more detailed study, a good approximation of the number of links needed is shown in Table 4, 25 TABLE 3 NUMBER OF ANALYSIS ZONES FOR AREAWIDE ANALYSES1 Urban Area Number of Population Analysis 25,000 10-15 50,000 20-30 75,000 30-50 100,000 35-60 125,000 45-75 150,000 55-90 175,000 65-100 200,000 75-125 1 Does not include external stations on cordon surrounding the study area. 26 TABLE 4 NUMBER OF ASSIGNMENT LINKS FOR AREAWIDE ANALYSES No. of Links* for No. of Links* for Computer Assignment Manual Assignment Urban Area (includes 4 Centroid (No Centroid Population Connectors) Connectors) 25,000 90-130 20-45 50,000 175-250 40-90 75,000 250-450 60-150 100,000 300-550 70-180 125,000 400-700 90-225 150,000 500-800 110-270 175,000 600-900 130-300 200,000 700-1100 150-375 * These are physical connections (non-directional). If there were no one-way streets in a network, 100 links, as defined above, would represent 200 directional connections. 27 Of course, for slower growing areas, the lower end of the number of analysis zones and link ranges should be considered, and the upper end for faster growing areas. Zone-to-Zone Impedance Values. For modeling purposes, impedance values are required for transportation system segments so that routes can be selected for assignment purposes as well as to provide necessary zone-to-zone impedance values for trip distribution and modal choice procedures. Impedance values represent the separation between zones In terms of time, distance, or cost. Sometimes a combination of these are used. For link impedances, link travel time based upon a sample of observed travel speeds should be used. Travel speeds are usually obtained by employing floating car techniques (See Determining Travel Time 1/). Sampled speeds are generally grouped by facility type and type of area. Average speeds for these groups are then calculated. These speeds are then assigned to links in the system based on facility type and area type. In areas under about 100,000 population, speed runs may be considered for all facilities in the assignment network. For larger areas, a sampling of facilities is appropriate. The number of readings during a measured time period can be safely cut to 2 or 3 from the recommended 6 to 12 if a classification scheme is used. For default values, where speed runs are not made locally, the reader is referred to Table A-6 in the Appendix. Selection of Techniques. With regard to choice between non- computer and computer techniques, the intent of the areawide analysis and the size of the urban area (i.e., number of zones, extent of assignment network, etc.) are important. If alternate plans--both land-use and system--are to be tested,'computer tech- niques are most desirable in almost all cases. If, as suggested in this set of manuals, areawide analyses are accomplished for the sole purpose of identifying problem areas, then manual methods are appropriate in many instances. If the number of zones is limited to about 40 and the number of system links to under about 200, the process can be accomplished manually, especially if a one-purpose trip distribution model is used (see section on one-purpose gravity model in this chapter). Default Values. A great deal of time and effort is required to obtain information for model application and for analysis purposes through individual surveys conducted locally. Certain parameters and characteristics can be transferred between areas. Where a survey has been taken locally in the past, it should be considered for providing local conditions. 28 Appendix A,provides various transportation values which can be used when local information is not available ("default values"). - Trip Generation: - Detailed Trip Generation Characteristics 25,000- 50,000 Population - Table A-1 50,000-100,000 Population - Table A-2 100,000-200,000 Population - Table A-3 - Summary Trip Generation Parameters By Above 3 Population Groups - Table A-4 - Trip Distribution: - Airline Distance to Distribution Factor Conversion 25,100-100,000 Population - Figures A-1 to A-6 100,000-200,000 Population - Figures A-7 to A-12 - Assumptions Used in Developing Distribution Factor Conversion Charts By above 2 Population Groups - Table A-5 - Vehicle Speeds by Facility Type and Location For above 2 Population Groups - Table A-6 - Gravity Model Impedance Exponents For above 2 Population Groups - Table A-7 - Auto Occupancy: - Auto Occupancy by Trip Purpose For above 2 Population Groups - Table A-8 - Auto Occupancy by Land-Use at Destination For above 2 Population Groups - Table A-9 - Time-of-Day Characteristics: - See Chapter Six of Users Guide 9/ - Choice of Travel Mode - See Appendix C 29 "Default" values can be used with non-computer or computer techniques, Several techniques described herein or referenced later in other available rePorts are based on the type of data provided in the default tables and figures listed above. Applications of such default values using non-computer procedures to three scenarios are provided in the Users Guide 1/, The scenarios include a site development impact analysis@ a corridor analysis, and a land-use/highway spacing analysis. Non-Computer Techniques Non-computer techniques are suitable for analyses related to problem identification and location. Although they are most appropriate for subareas or specific sites and corridors, they may be used for an entire urban area if the number of zones and size of network is no more than a maximum of about 40 zones and 200 links. The report, Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters: A Users Guide 9/, provides details regarding the basis of development for the non-computer methods, the data required for application, features and limitations, etc. Trip Generation. Appendix Tables A-1 through A-3, "Detailed Trip Generation Characteristics," are estimates of trip generation rates based upon household income and/or auto ownership. The data are summarized by population groups. Household income is in terms of 1970 dollars and incomes estimated for other time periods must be adjusted to this base when using Appendix Tables A-1 through A-3. To convert household incomes to a 1970 dollar base, the consumer price index is used. The U.S. Department of Labor, Bureau of Labor Statistics, provides this index for the United States as well as for many cities and Standard Metropolitan Statistical Areas (SMSA) 12/. For an example of such a conversion refer to the Users Guide 9/. The most complete forecast of household income was prepared by a joint effort of the Bureau of Economic Analysis of the Department of Commerce, and the Economic Research Service of the Department of Agriculture for the U.S. Water Resources Council 13/. The data includes historical and projected measures of population, employment. personal income and earnings for states, economic areas, Standard Metropolitan Statistical Areas and water resource regions. It is difficult to obtain household income information for areas under 50,000 population. In these cases,car ownership can be used as described in the following example. The data in Appendix Tables A-1 through A-3 can be used in several ways. If an estimate of zonal average autos per household is available, average daily trips per household can be estimated directly. For example, in an urbanized area of 50,000-100,000 population, 11.9 average daily person trips per household can be expected where the average auto ownership is 1.07 per household. Where the number of households by income range is forecast, the following parameters can be estimated: 30 - Average autos per household - Average daily person trips per household - Percent of trips by purpose - Percent of 0, 1, 2 and 3+ auto households - Trips per household for 0, 1, 2 and 3+ auto households A useful application of Appendix Tables A-1 through A-3 is to obtain control totals for an urbanized area based upon income distribution forecasts. These control totals would include the above listed information for the entire urbanized area, thereby providing a basis for controlling estimates derived at the zone or household level. The "Average Daily Person Trips per Household" information in Appendix Table A-4, "Part A - Trip Production Estimates," is based upon the information presented in Tables A-1 through A-3. "% Average Daily Person Trips By Purpose" is an expansion of the 1970 average condition and may vary if the trip rate varies. A review suggests that using these data for another year would not cause too much variation in results. Likewise, the "Auto Person Trips as a % of Total Person Trips" and "Auto Driver Trips as a % of Total Person Trips" can be used for years other than 1970. For many urban areas in the 25,000-200,000 population range, a separate estimate of transit demand is suggested. Except for a rapidly growing area where the potential for transit use for work travel is greater than about 15 percent of person work trips, transit demand should be evaluated separately from highway demand. The relationships shown in Appendix Table A-4 provide the means for converting person trip estimates from Tables A-1 through A-3 to auto vehicle trips. As an example, assume for some zone that the application of Table A-2 results in 13 average daily person trips per household in an area of 65,000 population, with 15 percent of these homebased work (HBW), 62 percent homebased non- work MIN) and 23 percent non-homebased (NHB). Appendix Table A-4 indicates that auto driver trips are about 75 percent of HBW person trips, 52 percent of HBNW person trips and 64 percent of NHB person trips. Estimates by purpose of auto driver trips would be calculated as: HBW: 13 x 0.15 x 0.75 = 1.46 HBNW: 13 x 0.62 x 0.52 = 4.19 NHB: 13 x 0.23 x 0.64 = 1.91 7.56 Auto Driver Trips 31 Also, in cases where there are no large differences in population characteristics between zones, Appendix Table A-4 can be used to generate trips,by zone. Appendix Table A-4, "Part B - Useful Characteristics for Trip Estimation" presents rough-cut approximations which can also be useful. To illustrate the use of the material in Parts A and B of Appendix Table A-4, assume an urbanized area with 65,000 population and 22,000 households. With 14 average daily person trips per household ("Part A"), a total of 308,000 internal plus external trips can be expected. Sixteen percent of these would be HBW trips, or 49,280 daily person trips; 61 percent HBNW or 187,880; and 23 percent NHB or 70,840. Of the total 308,000 person trips, 2 percent or 6,160 can be expected to be transit person trips; 40 percent, or 123,200, auto passenger trips; and 58 percent, or 178,640, auto driver trips. To obtain auto person trips by purpose, the total trips by purpose would be multiplied by 0.96, 0.99 and 0.98, respectively, to obtain HBW, HBNW and NHB trips. The results would be: 47,309 HBW person trips by auto; 186,001 HBNW person trips by auto; and 69,423 NHB person trips by auto. Auto driver (auto vehicle) trips for different purposes would be similarly calculated using 0.70, 0.54 and 0.69, respectively for a total of 184,831 trips. From `Part B," truck trips can be estimated at approximately 49,904 (184,831 auto driver trips x 0.27). Some characteristics of external travel are also presented in Appendix Table A-4. For an area with a population of 65,000, the Table shows the number of external trips passing through the area would approximate 21 percent of total external trips. To apply this percent to cordon counts, which include through trips twice, an adjustment is necessary to determine the absolute number of trips involved. This adjustment is best shown by an example. Assume a cordon count of 100,000 vehicles per day. These are made up of internal- external trips, external-internal trips and external-external trips. The external-external, or through trips, are counted twice. To calculate these, the 21 percent from Appendix Table A-4, "Part B," would be used as follows: Count = Trips x (1 + Proportion of Through Trips) Count = Trips x (1 + 0.21) If the total cordon count is 100,000, for example, the total external trips would be estimated as 100,000/1.21 or 82,645 trips. Of these, 21 percent are through trips, or 17,355. To visualize this graphically, refer to Figure 3. From Appendix Table A-4, "Part B," the number of external trips expected destined for the CBD would be 0.22 x 82,645 or 18,182. Appendix Table A-4, `Part C - Trip Attraction Estimating Relationships" allows estimation of trip attractions. For HBW, HBNW and NHB trip purposes, in 32 Click HERE for graphic. 33 order to adjust the attraction rates to "fit" a particular size urban area, the trip production control total for the area should first be developed using either the material in Appendix Tables A-1 through A-3 or Table A-4, "Part A." For the example described above, total area controls of productions were; HBW 49,280 HBNW 187,880 NHB 70,840 308,000 Daily Person Trips Assume, for example, the following employment-residential mix: Total Employment 25,000 Retail Employment 5,000 (out of the total 25,000) Dwelling Units 22,000 Areawide trip generation totals are next developed using the relationships in Appendix Table A-4, "Part C" as follows: 49,280 F1 = _____________ = 1.16 1.7(25,000) 187,880 F2 = _________________________________________ = 2.29 10(5,000) + 0.5(20,000) + 1.0(22,000) 70,840 F3 = __________________________________________ = 1.00 2(5,000) + 2.5(20,000) + 0.5(22,000) Areawide trip generation tables are used when attraction rates for zones are desired. For example, if a zone has a total employment of 1,000, a retail employment of 200, and 500 dwelling units, the trip attractions would be calculated as follows: HBW = 1.16 x 1.7 x 1,000 = 1,972 HBNW = 2.29 [(10 x 200) + (0.5 x 800) + (1.0 x 500)] = 6,641 NHB = 1.00 [( 2 x 200) + (2.5 x 800) + (0.5 x 500)) = 2,650 To utilize Appendix Tables A-1 through A-3 in estimating trips by household for a study area (entire area, zone, district, corridor, etc.), an estimate of the number of households by income range (1970 dollar base) is required. With this estimate, the user can obtain from the table the average autos per household, average daily person trips per household by number of autos per household, and percent average daily person trips by purpose. For example, if there are 1,000 households in the $6,000-$7,000 income range in an urbanized area of 50,000-100,000 population, the following numbers of trips are estimated to be produced: A. Using Averages of columns 2 and 3 Average autos per household = 1.07 Total autos (1,000 x 1.07) = 1,070 Average trips per household = 11.9 Total trips generated (1,000 x 11.9) = 11,900 34 B. Using Subsequent Columns Households with 0 autos = 15% or 150HH 1 auto = 64% or 640HH 2 autos = 20% or 200KH 3 autos = 1% or 10HH Trips by 0 auto households = 3,0/HH or 450 Trips by 1 auto households = 12.5/HH or 8,000 Trips by 2 auto households = 16.5/HH or 3,300 Trips by 3 auto households = 19.5/HH or 195 Total trips by all households = 11,945 Trips generated: HBW = 18% or 2,150 Trips generated: HBNW = 59% or 7,048 Trips generated: NHB = 23% or 2,747 The HBW and HBNW trips are generated at the household, whereas the NHB trips are generated elsewhere. See Chapter 10, "Scenario for Site Development Impact Analysis: Boise, Idaho," of the Users Guide 9/, which describes an approach to handle this difference. For special generators, such as an airport or a major shopping center, special studies of generation may be required and substituted for the attraction values as may be derived above from Appendix Table A-4, "Part C." Trip Distribution. The non-computer demand analysis procedure of distributing trips uses a desk calculator with an accumulating memory. The procedure shown employs the Gravity Model discussed in the Users Guide 9/. The procedure includes: - short-cut calculation of area-to-area impedances - distribution factors (friction factors) for urban areas for two population groups and, within these, for three trip purposes, namely, HBW, HBNW and NHB. (Distribution factors have also been supplied for a combination of all trip purposes.) The urban area population groups for which distribution factors are provided are as follows: 25,000 - 100,000 population 100,000 - 200,000 population - simplified worksheets for entering required information and calculating the trip interchanges. The use of the non-computer demand analysis procedure for trip distribution is illustrated in Appendix B, "Application of the Manual Trip Distribution Procedure." One-Purpose Gravity-Model. In many communities where growth is expected to be small but concentrated tn certain small areas of a region, a one-purpose 35 gravity model may provide the necessary trip distribution detail and accuracy. it is most appropriate for urban areas of between 25,000 and 150,000 population and where growth is anticipated to be less than 1.0-1.5 percent annually. At least one study, An Evaluation of the Gravity Model Trip Distribution 14/, has found that there is little practical difference in trip distribution from a single purpose, three purpose, or seven purpose model, The study was accomplished in an urban area with 132,000 persons, The basic conclusion was that although stratification of trips into several trip purposes isolate different travel patterns in terms of trip length distributions, each category contains a reduced number of trips and therefore each trip length distribution is subject to greater sample variation. Some estimates of manpower requirements indicate that a 100 zone trip matrix for one purpose can be developed in 120 person- hours, a 70 zone matrix in about 70 hours, and a 50 zone matrix in about 40 hours. A one-purpose model saves considerable calibration time and effort. A study using data from an urban area with approximately two million population found that the use of a one-purpose trip distribution model provided results only very slightly less accurate than the separate distribution of several trip purposes 15/. The calibration of a one-purpose gravity model by computer methods is accomplished in a manner similar to that of a three-or- more-purpose model. Calibration of such models are fully described in PLANPAC/BACKPAC - General Information 11/, and Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters 9/. An example of the use of a one-purpose gravity model is contained in Appendix E, "Details of Traffic Estimation Based Upon Ground Counts." In developing trip production and attraction information for a one-purpose gravity model, three purposes should be used: HBW, HBNW, and NHB. The homebased work and non-work trips would be handled as productions and attractions. The non-homebased trips Would be handled as origins and destinations in the trip generation phase. Thus, at the completion of trip generation, the output estimates would include: homebased work productions, homebased non- work productions and non-homebased origins, and homebased work attractions, homebased non-work attractions and non-homebased destinations, The productions for each zone for the one-purpose model would equal the sum of the homebased productions plus the non-homebased origins. Likewise, the attractions for each zone would equal the sum of the homebased attractions plus the non- homebased destinations. Two sets of friction factors are provided as "default" values for use in the application of a non-computer gravity model. These factors are provided in Appendix Table A-7. Also, for non-computer application using the methodology described in Chapter Three of the report, Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters 9/, conversion graphs for obtaining friction factors from zone-to-zone airline distance are provided in Appendix Figures A-1 through A-6 and Appendix Figures A-7 through A-12, for areas under and over 100,000 population, respectively. 36 The reader is also referred to the information described earlier in this chapter on the number of zones for model purposes (See Table 3). Input Requirements for Gravity Model. The following information is necessary before proceeding with the trip distribution process; 1. A map of the study area showing zone boundaries and their centroids; boundary limits of the CBD, central city and suburban subregions; and arterial streets and highways, with freeways differentiated. It is recommended that the size of the map be not larger than 1":2 mile scale and not smaller than V:4 mile scale, A typical study area map is shown in Figure 4. 2. Production and attraction trip ends by zone for each trip purpose. These figures are output from the trip generation stage and can be rounded to the nearest 100 trips without appreciable loss in trip interchange accuracy. 3. A travel time/distribution factor matrix. This matrix is triangular in that it is assumed that the travel time from zone i to zone i is the s as from j to i. The construction of such a matrix is simplified through the use of the centroid-to-centroid airline distances which are converted to travel time. Knowing the airline distance in miles between the production and attraction centroids and approximating the "over-the-road" freeway- arterial percentage mix for that trip interchange, the total travel times (in-vehicle time plus total origin- destination terminal times) and the corresponding distribution factor for any one of three purposes (or the combined trip purpose) can be read off the graphs (Appendix Figures A-1 through A-12). The derivation of these graphs is documented in the Users Guide 9/, and the use of the graphs is described below. 4. A thorough knowledge of the study area. With a map of the study area, the production and attraction trip ends from the preceding trip generation stage, and the area- to-area travel time/ distribution factor matrix, the user can then distribute trips. Use of Trip Distribution Graphs. Graphs have been developed to reduce substantially the steps involved in obtaining the distribution factors (friction factors) required for the application of the distribution procedure. Rather than select a route and calculate elapsed travel time or distance, the graphical method is based upon measuring a straight line distance (the airline distance) between the centroids of each zone. Two sets of graphs have been developed for each urban area population group (See Appendix Figures A-1 through A-12.). The first set is employed when a trip interchange occurs within a subregion; e.g., CBD-to-CBD,, or central city-to-central city, or suburb-to-suburb. For such an interchange, the travel time read from the appropriate graph is the total travel time including terminal times as well as in-vehicle times. Terminal times include walk and park times at both ends of the trip, as well as travel time on the local streets. 37 Click HERE for graphic. 38 For example, for an urbanized area of 25,000-100,000 population and a trip interchange within central city from analysis area 4 to 12 (Figure 4), the applicable graph is Appendix Figure A- 2. Scaling the airline distance between the centroids of areas 4 and 12, the distance is found to be 7 miles(to the nearest mile). Through past experience of the user, his familiarity with this particular interchange and its surrounding area, it is concluded that 20% of over-the-road travel is made on arterials (and therefore 80% on freeways). For such a mix of facilities, the 20% curve on the graph is used to generate the following information: Total travel time between areas 4 and 12 = 18 minutes Corresponding distribution factor for HBW trips = 0.55 Corresponding distribution factor for HBNW trips = 0.43 Corresponding distribution factor for NHB trips = 0.47 Corresponding distribution factor for all-purpose trips = 0.50 These figures are assumed to hold true for the area 12 to area 4 interchange also. The second set of graphs is utilized when the trip interchange occurs across subregions; e.g., CBD-to-central city, or CBD-to- suburb, or central city-to-suburb, etc. In this case, the travel time read from the graphs is the in-vehicle time and excludes terminal times. Total travel time is then calculated by adding the appropriate total origin and destination terminal times shown tabulated on the second set of graphs for each urbanized area population group. For example, for an urbanized area of 25,000-100,000 population and a trip interchange across subregions, say, central city-to-suburb, i.e., from area 4 to 18 (Figure 4), the applicable graphs are Appendix Figures A-5 and A-6. Scaling the airline distance between the centroids of areas 4 and 18, the total distance is found to be 11 miles, with 2 miles being the central city portion and 9 miles the suburb portion. Through knowledge of the system it is concluded that 5% of over-the-road travel of the 2 mile central city portion is made on arterials; also, 100% of over- the-road travel of the 9 mile suburb portion is made on arterials. For such facility mixes, the 5% curve (interpolated) in Appendix Figure A-5 and 100% curve in Appendix Figure A-9 are used to generate the following parameters: In-vehicle travel time for 2 mile central city portion = 2.8 minutes In-vehicle travel time for 9 mile suburb portion = 18 minutes Total terminal times for central city-suburb-combination = 4.6 minutes Total travel time between area 4 and area 18 = 25 minutes Corresponding distribution factor for HBW trips = 0.27 39 Corresponding distribution factor for NHB trips = 0.19 Corresponding distribution factor for all-purpose trips = 0.20 These figures are assumed to hold true for the 18 to 4 interchange also. Situations arise for trip interchanges where the origin and destination centroids are located within a subregion and yet the centroid-to-centroid straight line crosses subregions. Trip interchange from district 6 to 13 or 13 to 6 (Figure 4), for example, has its origin and destination in the central city but crosses the CBD. For this case, the within set of curves do not apply; the airline distances and the in-vehicle travel times should be computed separately (for the central city and CBD portions) using the across subregion set of curves with the appropriate freeway-arterial mix percentages. The central city-to-central city terminal times are then added to obtain the total times, and the corresponding distribution factors are then obtained. As a further illustration, a situation which entails the use of the across subregion set of curves is an O-D straight line that crosses the CBD, central city and suburbs. For instance, the straight line between areas 18 and 20 (Figure 4) crosses 3 subregions and therefore, 3 in-vehicle times are accordingly determined (for the suburb, central city and CBD portions) and then the suburb-to-suburb terminal times added. It must be pointed out that in cases of the across subregion travel, the distribution factor cannot be read off the graphs for each component of travel time and then summed. The total travel time must be computed first. Then, any one of the appropriate graphs for an urbanized area population group is reentered and the distribution factor (by trip purpose) corresponding to the total travel time is determined. Note that if travel occurs entirely within a zone or district, i.e., an intra-zonal or an intra-district trip, the airline distance is computed by measuring airline distances from the centroid of the area of intra-travel to the centroids of all adjacent zones or districts. These airline distances are then averaged, and then the intra-area airline distance is taken as one- half of this average. The appropriate graph is then entered to obtain the travel times. Notice, however, that in all of the above situations, it is the airline distance between the O-D centroids that is scaled from the map, but the freeway-arterial facility mix is determined from the actual over-the-road conditions irrespective of the magnitude differences between the airline and over-the-road distances. The graphs (Appendix Figures A-1 through A-12) have been designed to accommodate these variations by using built-in factors. Note also that in determining the freeway-arterial facility mixes, the local and collector portion of the trip is not explicitly considered, simply because this factor is also allowed for in the graphs. 40 In the discussion above, it has been presumed that for @he study area in question, a travel time matrix is not available. Of course, if such a matrix has already been compiled, it should be used. it will save considerable time and effort. Also, it may be possible to construct a travel time matrix using other techniques based on knowledge of the area and eliminate the need for applying the airline distance/travel time conversion approach. To construct "Airline Distance vs, Travel Time vs. Distribution Factor" graphs where the user wants a more specific representation (say, for particular local conditions), reference is made to Chapter Three of the Users Guide 9/. Mode Choice Analysis. In areas under 200,00 population, public transportation traditionally has played a relatively minor role in overall transportation service. But with the need to serve segments of the population such as the elderly and handicapped, and to conserve energy and protect the environment, the evaluation of public transportation service is increasingly important. A tech- nical evaluation separate from the automobile-highway system analysis is appropriate in cases where transit usage does not significantly affect automobile usage. The companion technical manual, Transit Planning, describes techniques appropriate for non-computer application-for both general service and service to specific groups. In Appendix C, a sample method is provided for trip interchange mode choice analysis, where such is appropriate, as may be determined from the Transit Planning manual. The method is described for both a three-purpose trip distribution and for a single purpose trip distribution. Trip interchange methods are generally recommended only for areas having significant "choice" riders and requiring an analysis of the relationship between the auto-highway system and transit. Traffic Assignment. Assignment of trips to small networks is described in detail in the Users Guide, Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters 9/, Chapter Seven. The description pertains to assignment to a portion of a network, the concept being that manual methods are extremely time-consuming for large networks. However, in smaller urban areas, the non-computer methods described can be used for complete networks. Here, the justification is that traffic assignment to an areawide network should be accomplished to determine problem locations and the probable extent of the problem. Manual traffic assignment should not be used to test alternatives, land-use or transportation, or an areawide scale. More appropriately, alternative transportation solutions (new facility- a widening, traffic engineering improvements, transit service, etc.) to address specific problems should be evaluated on a more limited corridor or subarea basis. Some evaluation and experience indicates that for typical networks described for an urban area, there should be an approximate ratio of between 2 and 3 physical system links per zone. This assumes no centroid connectors which are 41 not required for non-computer assignment purposes, In the section of this manual entitled, "Basic Considerations," a guideline relative to the number of analysis areas and physical links is provided (See Tables 3 and 4). Summarized in Table 5 are approximate person-hours expected for non-computer assignment analysis, Except for the larger areas, and where alternatives are to be tested on a study area basis, manual assignment can be accomplished in a reasonable amount of time. Truck-Taxi Estimates. Many of the techniques and default values described in this manual are related to person travel. Not addressed in most discussions are methods for estimating truck and commercial auto vehicle travel and taxi travel. In most smaller urban areas taxi travel is a minor part of travel and might not be included in the demand estimation. Special studies as part of paratransit evaluation might be used when necessary. For a truck trip estimate the reader is referred to Appendix Table A-4, the column labelled "Total Areawide Truck Trips as a % of Areawide Auto Driver Trips." Chapter Six of the Users Guide 9/ includes tables reflecting the relationship between total vehicles and internal auto driver trips. Given an internal auto driver matrix, the factors can be applied to account for external travel and commercial autos, trucks and taxis. The appropriate tables in the Users Guide 9/ are Table 18 for areas with 50,000-200,000 population, and Table 23 for areas with 100,000-250,000 population. For areas of between 25,000 and 50,000 population, the values in Table 18 may be safely used. In most cases only the total daily factor at the bottom of each table is necessary. An alternative to the above approach is to take classification counts by facility type and area (i.e., CBD, central city, suburbs) and develop factors by area and facility type relating total vehicles to autos. These factors are applied to assignments of auto drivers. Computer Techniques The four-step transportation planning procedure--trip generation, trip distribution, mode choice and trip assignment--can be accomplished in a simplified fashion by combining appropriate Urban Transportation Planning,System (UTPS) software and appropriate default rates developed by urban area size. The use of these programs to accomplish the four-step transportation planning procedure is illustrated in Figure 5. The programs, shown in Figure 5, have the following functions; Program Function HR Produces a computerized street and highway network description (historical record) from coded highway link data. 42 TABLE 5 WORK EFFORT REQUIRED FOR NON-COMPUTER TRAFFIC ASSIGNMENT Number Number of Approximate Population of Zones Highway Sections Person-Hours 25,000 10- 15 20- 45 3- 5 50,000 20- 30 40- 90 5- 13 75,000 30- 50 60-150 13- 40 100,000 35- 60 70-180 17- 60 125,000 45- 75 90-225 30-100 150,000 55- 90 110-270 50-150 175,000 65-100 130-300 75-200 200,000 75-125 150-375 100-300 43 Click HERE for graphic. 44 Program Function UROAD Produces a matrix of travel times between all zones. SCAGM Applies trip generation rates and a gravity model to produce a trip table. UMATRIX Converts person trips to vehicle trips and incorporates through trips and truck trips. UROAD Converts the production-attraction table to origin-destination format (split 50-50) and assigns trips to the network. Each of the programs described contain multiple options to allow their use in a wide variety of situations. Various types of input data are necessary to utilize the programs. To the extent possible, default data and calculations were incorporated into the program chain. The development of this default data was described previously in this chapter and should be reviewed for its applicability. The UTPS computer program package is a readily available, tested and easy to use planning tool. Many of the programs serve multiple functions which reduce the number of programs and the number of intermediate data files with which the user must become familiar. Some of the programs also have provision for inclusion of user-coded subroutines. The UTPS documentation is essential to any user contemplating the actual application of the programs. The UTPS package (programs and documentation) is available from State or Federal transportation officials. Material has been excerpted from the UTPS documentation to serve as a guide to reading the complete UTPS material and is shown in Appendix D. Validating Non-Computer and Computer Technique Results1 Each model used in transportation planning must be validated to determine the ability of the models to produce realistic results. Trip generation models should be checked first, followed by trip distribution and finally traffic assignment. Differences in traffic assignment results may indicate problems at any stage of the modeling process. Trip generation productions and attractions are checked for reasonableness by comparisons with trip rates per dwelling unit or trips per capita from studies conducted in similar communities. ___________________________ 1 Much of this discussion is based on the work of Marvin Goleberg, S.G. Associates Inc., Newton Centre, Massachusetts. 45 Additionally, areawide travel. estimates (vehicle miles of travel) can be obtained by multiplying productions by average trip length (done separately by purpose and Hummed). This rough estimate should be compared to the areawide travel calculations made for the functional classification of streets and highways (See Chapter One.). Another check that can be made is based on the establishment of cordons around a few selected high activity land-use zones, Counts at the cordon are then compared to the estimates of total productions and attractions. These productions and attractions must be adjusted to account for estimated intrazonal trips. The following high activity zone types should be checked: residential, industrial and shopping. Problems uncovered during the trip generation validation can be addressed in several ways: - If trip rates or areawide VMT are in error, the trip generation relationships or the socioeconomic and land- use inputs must be verified. - If cordon checks are in error, localized data errors may be at fault. In addition, special generators should be examined to determine if special analysis is required for specific zones. Trip distribution can be validated by examining the trip length frequency distribution for each purpose to determine reasonableness. This distribution should be compared with trip length frequency curves from similar communities. Total auto driver trips can be checked by comparison to traffic counts and trips across screenlines, between groups of zones, and into and out of the central business district. A judgment check can be made by examining selected interchanges for logical relationships between production and attraction zones. Problems uncovered in trip distribution can be addressed as follows: - Errors in trip length frequency distribution can be due to 'F-factor' problems or network speed coding. - Screenline and area related problems can be caused by network speed or routing errors or problems with productions and attractions from the trip generation model. The most critical portion of the validation process is performed after traffic assignment has been completed. This is sometimes done first to determine whether errors are large enough to require the evaluation of each individual model, i.e., trip generation and trip distribution. The checks made at this point should be at two levels-macro and micro. Macro checks are performed to make final adjustments, if necessary, to the trip generation and trip distribution models prior to fine tuning the network. 46 They include comparisons of assigned areawide VMT, screenline crossings, CBD cordon crossings-and major cutline crossings to values obtained through counts. Micro checks are made to fine tune the network based on traffic count comparisons by functional class and area type. Typical kinds of network problems are:a) error in assigned volumes for only a specific class of street or highway which may mean that the number of facilities coded is not proper, and b) link volumes are inconsistent along a route; this problem may require speed adjustments to obtain more logical routes. The types of validity analysis that should be performed are shown below. The purpose of,each analysis is indicated, 1. Comparison of Total Screenline Results - Similar over or under-estimates on all screenlines indicate trip generation problems. - Different over or under-estimates indicate a trip distribution, an assignment problem, or a localized trip generation problem. 2. Comparison of Screenlines Between Adjacent Corridors - Indication of trip distribution or assignment problem. 3. Comparison of Screenlines Along_a Corridor - Indication of trip distribution or assignment problem. 4. Comparison of Geographic Area Count and Assignment VMT - Comparison of all functional classes indicates possibility of a localized trip generation problem or a distribution problem. - Comparison by functional class indicates possibility of an assignment problem. - Comparison of study area VMT values indicates possibility of trip generation or trip length problems. 5. Comparison of Assignment and Traffic Counts at External Station - Produces indication of accuracy of external forecasting process. Following this review, an orderly approach is; 1. Account for Assignment Problems Selected link analysis is undertaken to check out problem loadings. The objective is to determine if incorrect routings are present in the network which biases the count-volume comparison. Routing problems are 47 corrected by altering link speeds. The network is rebuilt, new trees built, the network loaded, and the count-volume analysis completed. This should continue until all significant routing problems are eliminated. 2. Account for Trip Generation Problems Areas of under and over-estimates of trips are the focus of the analysis, The trip generation models are studied in detail for those zones to determine the reason for estimation error. Results of this analysis will provide the direction for model revision. If judgment alone is not sufficient to make model adjustments, several possibilities exist for obtaining data for comparing to model results. Two possibilities are (a) an on-board survey and, (b) an external survey expanded by using the partial matrix technique.1 3. Account for Trip Distribution Problems Friction M factors, K-factors and "River Crossing Bias" are the variables which affect the distribution; productions and attractions are independent of the model, Correction of distribution problems is not straight-forward in a synthetic model approach as there are no reference data profiles. Over- assignment on river crossings may indicate the need to use river crossing penalties. Significant under and over- estimates on screenlines may indicate the need for K-factors. Over and under-estimates of required VMT may indicate need to revise F-factors or revise travel times. 4. Review External-Internal Model External-internal estimates are reviewed by external station (corridor). Uniform over or under-estimates across the external stations indicate a characteristic deficiency in the model which can be accounted for by adjusting the coefficient and constants. Differential over and underestimates in adjacent corridors could indicate misassignment to external stations. After all adjustments have been made, the model set is re-run and the comparative analysis repeated. Adjustments to the models should be logical and made in a way they can be related to changes in socioeconomic factors. One of the key assumptions in all models is that the constants and coefficients remain constant. This, in fact, could be an incorrect assumption and can affect all models. If a pattern of adjustments can be discerned and related to causative factors, this will provide the information to modify models for future points in time. ___________________________ 1 The partial matrix technique is described in detail later in this chapter. 48 Ground Count Projection Technique Traffic count data is an important element in transportation planning for monitoring changes in system operation, for checking and calibrating models when necessary, and as a basis for estimating future system loads. For estimating future system loads in areas where large shifts in development patterns are not expected, where growth is expected to be slow to moderate (less than about 3% per year) and where major new facilities are not contemplated, system estimates can be developed utilizing traffic counts as a basis. This approach is most appropriate for smaller urban areas (of less than 50,000 population) and where external travel through the area, and to and from the area is a significant portion of total local travel. However, where conditions are appropriate, the technique may also be used for areas larger than 50,000. Such conditions would include no growth or very slow anticipated growth with no new major facilities planned. The traffic estimation approach is basically a link count factoring approach requiring no model calibration or validation. External travel (through travel and external-internal travel) is forecast separately from internal travel and is actually assigned to the network. Internal travel is forecast by factoring counts, the factors being based upon the trip characteristics generally used for trip generation estimation. A technique is also provided for estimating potential use of in- creased capacity whether through an improvement in an existing facility or the provision of a new facility. The basic ground count factoring procedure obviously cannot consider a new facility since no ground counts are possible. The traffic estimation method is tied directly to an external O-D trip table and link counts. As mentioned above, it is most suitable: - for small urban areas with slow growth; - where external O-D and land-use data are available; - where future alternatives do not include major new facilities on entirely new alignments except when predominantly external travel is expected, for example, on a bypass. The procedure would not be particularly appropriate for estimating future internal travel on a new downtown river bridge if the new bridge were two miles upriver from an existing bridge. The factoring of counts on the existing bridge would not produce relevant data since the new bridge might attract existing trips between internal O-D zones. However, a method is provided to allow consideration of parallel new facilities, which allows smoothing of volumes between facilities. In many instances the factoring procedure may be considered a first step in estimating system loads, If the results indicate no particular system overloading problems, more detailed or sophisticated analysis would not be necessary. Where problems are identified, additional analysis may be warranted. The traffic estimation technique requires no additional data or analysis different from that 49 required by more detailed procedures, except that fewer steps are necessary. The technique can be applied manually or by computer, depending upon the number of zones and system size to be considered. Non-Computer Approach. Appendix E documents the input requirements and the steps in a non-computer application of the traffic estimation methodology. Also, an example application is provided to enable the user to become acquainted with the methodology. Computer Approach. The step-by-step procedure to estimate future system loads by manually factoring link count data is described in Appendix E. In cases where the number of zones and/or the number of nodes in the system are more than a nominal number, the non-computer approach can become laborious and prone to error. A computerized approach would then be necessary. Such a method is described below. The computer programs required are described in Appendix F. The computer approach differs from the non-computer approach in the handling of external travel. The non-computer approach removes assigned external travel from the link count, determines an internal growth factor, factors the internal portion of the link count and adds back assigned external travel for the forecast year. The non-computer approach has several benefits, Negative internal link volumes (where assigned external travel is greater than the link count) can be accounted for with a judgmental process that can also result in improved insight into the validity of the external survey trips, the coded network and the traffic count information. For certain situations, it may not be necessary to go through two gravity models and two traffic assignments. Where the internal contribution to link volumes is very localized, it may be possible to develop growth factors by analyzing the surrounding zonal growth. The completely computerized approach does not allow the flexibility of making judgments on isolated conditions. For simplicity of operation, both internal and external travel are assigned for the base and forecast years. These volumes are used to produce a growth factor by link which expands the base year traffic count to a forecast year value. Generally, the results from the non-computer approach and those from the computer approach should be reasonably close. It is up to the user, however, to decide which approach to use. The input requirements to utilize the computer approach are: (a) base year network with ground counts (b) external travel matrix (base year) (c) socioeconomic base year data for internal zones to allow the calculation of production and attraction estimates (d) growth factors for external travel (e) socioeconomic forecast year data for internal zones 50 All the above items are also required for a non-computer process. The only exception is the base year networks which would have to be coded so that program HR (described in Appendix D) could produce a machine-readable historical record. UTPS programs can be used in conjunction with some non- computer steps. For example, the external trips can be computer- assigned to a network and the results manually subtracted from counts, Computer-developed trips for internal travel can likewise be developed and assigned, The results of a base year and future assignment can be made and the resultant link loads divided to obtain growth factors, and so on. Redistribution of Assigned Volumes Among Available Facilities In any assignment of travel to a highway network, whether by non-computer or computerized methods, the link-assigned volumes will require some redistribution between available facilities to more closely reflect actual operating conditions. Historically, transportation planning procedures have utilized screenlines and auxiliary cutlines to validate and analyze assignment results, and the redistribution technique described below follows the same approach. The technique described to reallocate travel between competing facilities after traffic assignment is based on screenline theory and was developed by R. H. Pratt Associates 16/. The technique requires the analysis of multiple overlapping cutlines of major screenlines within an analysis area. The procedure may appear to be difficult and time consuming, but an analysis area containing ten vertical and ten horizontal major screenlines can be processed and summarized in two person-days. Most analyses will not be so extensive. The redistribution procedure assumes that forecast year volumes on parallel facilities should tend to be distributed proportionately to the volumes as observed on the facilities in the base year. If capacity does not change between the year observations are made and the forecast year, the forecast year volumes on the links intercepted by the screenline will tend to be proportional to the base year volume. All capacity changes to the forecast year system are interpreted as new facilities, including widenings to existing facilities. An example application illustrating the manner in which the post-assignment redistribution technique operates is provided in Appendix C. Partial Matrix Technique (PMT) Transportation planners tend to rely less and less on the traditional surveys and large-scale data collection processes, Synthetic data and simplified models and procedures, developed either for local conditions or transferred from other study areas, are gaining popularity. The reason is the substantial reduction in costs (both monetary and time) for surveys, model calibration, and application. The Partial Matrix Technique (PMT) developed by Neffendorf and Wooton 17/, is one such simplified model, Essentially, the inputs to this model consist of information obtained from roadside interviews at appropriately located screenlines 51 and cordon stations in a particular study area. These interviews yield a partially complete trip matrix of observations. Subsequently, using the unique properties of this partial matrix, one can synthesize the complete matrix, In addition to a complete trip matrix, the PMT produces the trip distribution function of the matrix (i.e., a final set of F- factors and the trip length frequency distribution) and estimates the total trip ends (P's and A's) for each analysis area. The estimate of trip ends enables the development or checking of trip end relationships derived from the preceding analytic step, trip generation. Details regarding this technique are in Appendix H, including; - Theory of the PMT - Computational Steps for Application of the PMT - An Example Application Land-Use/Facility Spacing Relationships The techniques and procedures described for areawide analysis provide the mechanism for determining where potential problems may occur and for assessing changes in demand. Where new developments or large growth is expected in several locations,a simplified technique is desirable to "size" an appropriate transportation system; i.e., determine a system definition. Techniques for system definition include functional classification as described in Chapter One, and land-use density/ highway spacing as will be described here. Where significant growth is occurring in the suburbs and beyond, the following two approaches are most appropriate. The first approach is described in detail, with appropriate scenario descriptions, in Land Use and Arterial Spacing in Suburban Areas 18/. This document provides simplified guidelines for correlating arterial street systems with local development patterns and densities. The guidelines are flexible and easy to apply, and can help to estimate the required arterial street system size and spacing and may be also applied in reverse to determine the magnitude of development that can be supported by a given street system. Three scenarios are provided in the referenced document: 1. Determining street requirements of a proposed new town 2. Reverse application of guidelines to determine the magnitude of development that can be supported by a given street system 3. Evaluating the impact of a proposed new development Another, but more complicated approach, is provided in the report, Quick Response Urban Travel Estimation Manual Techniques and Transferable Parameters (Chapter 9) 9/. This material relates suburban development to estimates of highway levels of service so that assessments can be made of the highway transportation needs of land-use growth and change. As in the previous case, this is a non-computer method. The methodology is designed to produce the number of lane-miles of arterial highway required in an analysis area, given land-use activity, a freeway 52 system and a desired level of arterial traffic service for that analysis area. An estimate of the number of miles of freeway to be provided is made outside the procedure, but the method does indicate where such additional facilities would be desirable to improve the level of transportation service provided. Unlike the first method mentioned, this method provides for input regarding local area trip generation. Computational steps include the following 1. Computation of vehicle trip ends 2. Computations of transit use and auto occupancy adjustments 3. Computation of vehicle miles of travel and external vehicle miles of travel adjustment 4. Computation of vehicle miles of travel on freeways and arterials 5. Computation of average arterial volumes per lane and level of service. A detailed example is provided describing the application of the methodology. Also, a scenario application is included in Chapter 12 9/. As previously noted, the above procedures are most appropriate for system definition where significant growth is expected in a portion of a region or where a new area within a region is rapidly developing. The methods, as discussed herein, are not used to determine problem areas or locations within a region. Such determinations are best made with the other methods described in this chapter. The other methods described in this chapter would be used to evaluate in more detail the system determined by the "sizing"' techniques described above. Again, such a procedure would be required only for newly developing areas or those with rapid growth. Corridor and Site Analysis Much of the material provided for demand estimation is most effectively applied to areawide analysis. However, once problem corridors and subareas of concern are determined through either the application of the demand estimation procedures or through surveillance of the current system, analysis should be in greater detail. Three approaches will be summarized here with reference to appropriate material for more detail. First, areawide analysis techniques can be applied to smaller areas with the greater detail which is usually required for site and corridor analysis. For trip generation, if trip rates have been developed on a cross-classification basis, the techniques can be applied to analysis areas of varying size without any problem. The rates described in this chapter for trip generation and included in Appendix A can be applied at varying levels of detail, For trip distribution, two approaches may be used for study of a corridor. The first is based upon an already developed regional trip table; the second is to develop a trip table based upon the specific problem to be solved. If a trip table is already available, a proportioning process can be easily used to develop trips between smaller analysis units in the corridor of study, To accomplish this without a computer, 53 the reader is referred to Chapter Three, Applications of the Manual Trip Distribution Procedure to Corridor and Site Analysis section of Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters 9/. If using a computer approach, the reader should consider use of program USQUEX, which is part of the UTPS system available.LO/. The alternative to this method is to develop a completely new trip distribution employing a revised set of analysis areas. This would involve using relatively fine analysis areas within the corridor of interest and relatively gross areas outside the area of interest, but would otherwise use the type of techniques previously described in this chapter for trip distribution. For site analysis, using methods described in this document, the reader is referred to the section of the Users Guide 9/ referenced above. Non-computer methods should be seriously considered for site and corridor analysis. Second, the method previously described for land-use/facility spacing 9/, can be utilized for evaluating the impact of proposed major developments on the surrounding street system. The reader is referred to Scenario 3 of the above referenced document. Other scenario applications of this methodology are provided in the "Traffic Generation/Traffic Decay and Street Requirements" section of the Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters 9/ document. Third, other methodologies are available which may be appropriate for subarea and corridor analysis. One such method is described as, "The Traffic Shift Methodology for Corridors," and is also described in the above referenced document. 54 CHAPTER THREE EVALUATING THE TRANSPORTATION SYSTEM IMPACT ANALYSES The results of transportation demand estimation allow an evaluation of transportation system impacts. This evaluation should consider the impacts of the demand relative to social, physical, economic and environmental effects, and are described below. This chapter also describes methods for evaluating transportation service relative to intersection and corridor capacity. Finally, information is provided on determining solutions to transportation problems. Various impacts are associated with the demand for travel, whether by auto or by transit -- social, physical, economic and environmental (See Table 6). This chapter describes impact estimation procedures and indicates how they are to be used in the evaluation of alternative transportation improvements. Due to the fact that -transit environmental impacts are minimal in small urban areas, only auto travel demand impact prediction techniques are presented here. Accessibility and other social benefits of transit improvements are discussed in the Transit Planning manual along with information on costs. To provide a perspective of the impacts described, Table 7 is offered. This table shows the impact sub-elements that may need to be considered. The remainder of this chapter concerns itself with user time costs, operating costs and accident costs. Just as important as determining potential transportation impacts, it is necessary to select appropriate assessment techniques to evaluate those impacts. Simplified impact estimation techniques are discussed below. Table 8 is an example of a comparison of impacts of five alternatives with Alternative 0 being the "do-nothing" alternative. For example, if the total facility cost were the only criterion for judging the impacts of the five alternatives, Alternative 3 would be selected as the optimum (i.e., for alternatives other than the "do-nothing" alternative). Impact estimation is conducted after completion of demand estimation, as described in Chapter Two. Impact estimation requires the following information: - Vehicles (or persons) by highway facility - Zone-to-zone travel time - Average trip distance/time - Vehicle miles of travel (VMT) or Person miles of travel (PMT) - Vehicle hours of travel (VHT) or Person hours of travel (PHT) - Number of auto trips Social Impacts Social impacts are related to the accessibility of segments of the population to various community services, such as the accessibility of low income individuals 55 Click HERE for graphic. 56 Click HERE for graphic. 57 Click HERE for graphic. 58 to medical facilities. Accessibility can be measured relative to both highway and transit service. Such analysis can be made with the computer program SAACCESS 11/. In most cases, non-computer methods are available using trip distribution information (i.e., accessibility indices from a gravity model) and isochronal maps (See Figures 6 and 8). Figure 7 illustrates impact analyses relative to employment accessibility to help evaluate equity between population groups. Economic Costs and Benefits Economic impact analyses are related to user travel time, vehicle operation, and accidents. Since there is much variation in cost values associated with accidents, it is only necessary to predict total numbers of accidents in the evaluation of alternative improvement possibilities (See Table 8). User Time, Cost (Auto). Auto user time and costs are computed in the following manner: Total* Average* Total Total person x trip = Person = user trips time Hours time (PHT) Total $ Value Total PHT x of user = user $ time cost Add Total Total user $ to auto $ (See next section) cost operating costs Total to arrive at areawide $ or report separately. travel costs * could be stratified by trip purpose. The value of user time changes over time and varies by area. For work trips, many have used the average wage in the community. For other types of trips, a lower value is usually assumed. For non- work trips, the value of travel time is calculated by dividing total wages by population. Operating Costs. Auto operating costs for various levels of analysis are shown in Tables 9 and 10. These data are dated from the standpoint of inflation, operating characteristics of vehicles, and so on. The material provides an indication of information for estimating operating costs and potential sources for more current information. Such information can be used by section of facility to determine operating costs by facility type and speed. Figure 9 provides average costs for all travel which can be used to estimate total costs for the community. 59 Click HERE for graphic. 60 Click HERE for graphic. 61 Click HERE for graphic. 62 Table 9 Vehicle Operating Cost on Freeways (1,2) Speed Cost (mph) (cents per mile) 60.0 12.74 55.0 11.88 50.0 11.27 45.0 10.66 40.0 10.23 35.0 9.92 30.0 9.68 25.0 9.56 (1) Cost data based on a vehicle mix which includes trucks and private automobiles: 83.04% passenger cars, 6.81% two-ton trucks, 3.26% six-ton trucks, 3.29% 20-ton trucks, and 3.60% 25-ton trucks. (2) Data based on typical roadway segments which reflect various curves, grades, stops per mile, traffic densities, etc. Notes: Data are in terms of 1973 costs; includes maintenance, tires, oil, gasoline and depreciation, but excludes taxes, insurance, parking, and tolls. Trucks are diesel powered. Source: Bloom, Kent, TRANS-Urban Computer Model (OPGAS), Federal Highway Administrations U.S. Department of Transportation, Washington, D.C., April, 1973 Note: This type of information is quickly outdated. The agency providing the source should be contacted relative to more current information. 63 Table 1-0 Vehicle Operating Cost on Arterial Streets (1,2) Speed Cost (mph) (cents per mile) 35.0 17.03 30.0 16.17 25.0 15.99 20.0 15.44 15.0 14.58 (1) Cost data based on a vehicle mix which includes trucks and private automobiles: 83.04% passenger cars, 6.8% two-ton trucks, 3.26% six-ton trucks, 3.29% 20-ton trucks, and 3.60% 25-ton trucks. (2) Data based on typical roadway segments which reflect various curves, grades, stops per mile, traffic densities, etc. Notes: Data are in terms of 1973 costs; includes maintenance, tires, oil, gasoline, and depreciation, but excludes taxes, insurance, parking, and tolls. Source: Bloom, Kent, TRANS-Urban Computer Model (OPGAS), Federal Highway Administration, U.S. Department of Transportation, Washington, D.C., April, 1973 Note: This type of information is quickly outdated. The agency providing the source should be contacted relative to more current information. 64 Click HERE for graphic. 65 Accidents. Accident rates vary by street and highway design and by such characteristics as speed limits, size of vehicle, right turn on red, etc. Table 11 illustrates the type,of accident information needed for impact analyses. Accidents and accident rates should be monitored to provide information to be used in system analysis. For further information on accidents and accident reporting systems, see the Traffic Planning and Project Programming manuals. Energy Impacts Fuel consumption should be evaluated in system analyses using the following relationship: Total ö Average = Total VMT MPG Gallons Consumed Average fuel consumption rates change and must be monitored to obtain correct values for system analysis. Average MPG can also be obtained from State and Federal highway officials. Some examples of average energy consumption rates are shown in Tables 12 and 13 for freeways and arterials by speed of operation. Environmental Impacts The three main types of auto pollutants to be considered in system analysis are: Carbon Monoxide (CO) Oxides of Nitrogen (NOx) Hydrocarbons (HC) The technology concerning calculation of vehicle emissions is changing rapidly. Whereas previously published material by the U.S. Environmental Protection Agency allowed calculation of emissions in terms of pollutants/VMT by model year and speed, the latest data is provided in terms related to VMT, starts, hot soaks and diurnal evaporative emissions (see footnotes to Tables 14 and 15). Some examples of averages which may be considered are as shown in Tables 14 and 15. These tables are values from the computer program CAPM 10/, and include: emissions from hot stabilized operation on freeways and surf-ace arterials; emissions from auto starts; hot soak emissions; and diurnal evaporative emissions. Table 14 provides this information for 1977 conditions. Table 15 provides estimates for 1995. The footnotes on the tables should be read carefully to fully understand use of the values presented. The values shown in Tables 14 and 15 are also used in computer program UROAD 10/. EVALUATING TRANSPORTATION SERVICE Important considerations in transportation system analysis are the lane-miles of street or highway required to satisfy the estimated travel demand, and the traffic carried by the existing system before it becomes intolerably congested. 66 Click HERE for graphic. 67 Table 12 Vehicle Gasoline Consumption on Freeways (1,3) (2,3) Mixed Vehicle Automobile Speed Consumption Consumption (mph) (gallons per mile) (gallons per mile) 65.0 .09269 .07377 60.0 .08638 .06734 55.0 .08006 .06228 50.0 .07427 .05871 45.0 .07164 .05580 40.0 .07070 .05507 35.0 .06869 .05220 30.0 .06830 .05254 25.0 .06899 .05411 (1) Data based on vehicle mix of 83.04% passenger cars, 6.81% two- ton trucks, 3.26% six-ton trucks, 3.29% twenty-ton trucks, and 3.60% twenty-five ton trucks. (2) Data based on a vehicle mix of 66% standard and 34% compact automobiles. (3) Data based on typical roadway segments which reflect various curves, grades, stops per mile, traffic densities, etc. Sources: Bloom, Kent, TRANS-Urban Computer Model (OPGAS), Federal Highway Administration, U.S. Department of Transportation, Washington, D.C., April, 1973 Claffey, Paul, "Running Costs of Motor Vehicles as Affected by Road Design and Traffic," NCHRP 111, Washington, D.C., 1971 Federal Highway Administration. TRANS: Fuel Consumption for Urban Freeways and Surface Vehicle), U.S. Department of Transportation, Washington, D.C., 1973 Characteristics of Urban Transportation Systems -- A Handbook for Transportation Planners 23/ 68 Table 13 Vehicle Gasoline Consumption on Arterial Streets (1,3) (2,3) Mixed Vehicle Automobile Speed Consumption Consumption (mph) (gallons per mile) (gallons per mile) 40.0 .09326 .07998 35.0 .09312 .07970 30.0 .09497 .08054 25.0 .09428 .07993 20.0 .09102 .07742 15.0 .09244 .07842 (1) Data based on vehicle mix of 83.04% passenger cars, 6.81% two- ton trucks, 3.26% six-ton trucks, 3.29% twenty-ton trucks, and 3.60% twenty-five ton trucks. (2) Data based on a vehicle mix of 66% standard and 34% compact automobiles. (3) Data based on typical roadway segments which reflect various curves, grades, stops per mile. traffic densities, etc. Sources: Bloom, Kent, TRANS-Urban Computer Model (OPGAS), Federal Highway Administration, U.S. Department of Transportation, Washington, D.C., April, 1973 Claffey, Paul, "Running Costs of Motor Vehicles as Affected by Road Design and Traffic," NCHRP 111, Washington, D.C., 1971 Federal Highway Administration. TRANS: Fuel Consumption for Urban Freeways and Surface Vehicle), U.S. Department of Transportation, Washington, D.C., 1973 Characteristics of Urban Transportation Systems -- A Handbook for Transportation Planners 23/ 69 TABLE 14 1 POLLUTANT EMISSION FACTOR COMPONENTS (1977) A. Emissions from Hot Stabilized Operation Freeways and Surface Arterials _______________________________________________________________ Autos Trucks ___________________________ __________________________ Oxides Oxides Carbon Hydro- of Carbon Hydro- of Speed Monoxide Carbons3 Nitrogen Monoxide Carbon3 Nitrogen __________________________________________________________________ (MPH) (Grams Per Mile) (Grams Per Mile) ________________________ _________________________ 60.0 19.11 1.87 4.72 73.34 4.22 12.67 55.0 22.64 2.14 4.23 66.53 4.37 11.26 50.0 24.08 2.25 3.97 62.28 4.48 10.37 45.0 25.30 2.33 3.83 60.97 4.70 9.80 40.0 27.49 2.50 3.73 62.97 5.13 9.41 35.0 31.15 2.78 3.63 68.59 5.83 9.11 30.0 36.44 3.17 3.48 78.31 6.87 8.86 25.0 43.48 3.68 3.30 93.22 8.39 8.67 20.0 53.11 4.37 3.09 116.06 10.70 8.59 15.0 68.94 5.47 2.92 153.59 14.55 8.72 B. Emissions From Auto Starts ________________________________________________________________ Emissions __________________________________________________ Percent of Trips Oxides of Starting Cold Carbon Monoxide Hydrocarbons3 Nitrogen _______________ _______________ _____________ __________ (Grams Per Trip) _______________________________________________________________ 0 15.5 3.9 4.6 10 33.6 4.9 4.6 20 51.9 5.9 4.6 30 69.9 6.9 4.5 40 89.0 7.8 4.5 50 107.0 8.8 4.4 60 126.0 9.8 4.4 70 143.0 10.7 4.4 80 161.0 11.7 4.3 90 180.0 12.7 4.3 100 198.0 13.6 4.3 C. Other Emissions 1. Hot Soak Emissions (HC) 11.8 G/Auto Trip 2. Diurnal Evaporative Emissions (HC): 19.4 G/Auto/Day 3. Diurnal Evaporative Emissions (HC): 21.0 G/Truck/Day ___________________________ 1, 2, 3 See footnotes page 72 70 TABLE 15 POLLUTANT EMISSION FACTOR COMPONENTS' (1995) A. Emissions from Hot Stabilized Operation Freeways and Surface Arterials _______________________________________________________________ Autos Trucks ___________________________ __________________________ Oxides Oxides Carbon Hydro- of Carbon Hydro- of Speed Monoxide Carbons3 Nitrogen Monoxide Carbon3 Nitrogen __________________________________________________________________ (MPH) (Grams Per Mile) (Grams Per Mile) ________________________ _________________________ 60.0 2.01 0.16 2.38 26.48 0.75 6.52 55.0 3.30 0.23 2.09 27.56 0.88 5.56 50.0 3.64 0.26 1.93 27.82 0.97 4.96 45.0 3.62 0.27 1.85 28.14 1.09 4.58 40.0 3.75 0.29 1.80 29.24 1.28 4.31 35.0 4.24 0.34 1.73 31.44 1.59 4.09 30.0 5.11 0.41 1.63 34.88 2.06 3.91 25.0 6.26 0.50 1.50 39.64 2.75 3.75 20.0 7.52 0.61 1.35 45.91 3.72 3.64 15.0 9.02 0.77 1.20 54.74 5.16 3.63 B. Emissions From Auto Starts ________________________________________________________________ Emissions __________________________________________________ Percent of Trips Oxides of Starting Cold Carbon Monoxide Hydrocarbons3 Nitrogen _______________ _______________ _____________ __________ (Grams Per Trip) _______________________________________________________________ 0 11.1 3.7 1.2 10 20.4 4.0 1.4 20 29.7 4.4 1.6 30 39.0 4.8 1.8 40 48.3 5.1 2.0 50 57.6 5.5 2.2 60 66.9 5.9 2.4 70 76.2 6.2 2.7 80 85.5 6.6 2.9 90 94.8 7.0 3.1 100 104.0 7.3 3.3 C. Other Emissions 1. Hot Soak: 1.1 G/Auto Trip 2. Diurnal : 1.1 G/Auto/Day 3. Diurnal : 2.2 G/Truck/Day ___________________________ 1, 2, 3 See footnotes page 72. 71 Tables 14 and 15: Footnotes 1 Mobile source emissions are generated in four ways: 1. From vehicles traveling in hot, stabilized mode; that is, after the engine and catalytic converter (if any) have warmed up to their most efficient operating temperature range. (CO, HC, and NOx,emissions) 2. From vehicle starts; additional emissions arise when an engine is started, regardless of the travel distance. (CO, HC, and NOx emissions) 3. From hot soaks; when an engine is turned off, hydrocarbons are evaporated from unburned fuel in the crankcase. (HC only) 4. From diurnal evaporation; daily temperature cycles cause evaporation of hydrocarbons from fuel tanks, whether or not the vehicles are used. (HC only) 2 Because of data limitations, it is currently impossible to split truck start-up and hot soak emissions from hot stabilized emissions. Therefore, these truck factors include three components, excluding only diurnal emissions. 3 Hydrocarbon emissions include reactive hydrocarbons only; methane is excluded. 4 For vehicles with catalytic converters, any engine started more than one hour since its last operation is classified in the cold start mode. For non-catalytic vehicles, the interval is four hours. In the absence of local data, a default value of 53 percent cold starts.may be used for 1977. This value increases with calendar year as the auto fleet is increasingly populated by catalyst equipped vehicles. For 1995 it is estimated at 9 . Note: All factors have been computed for an assumed temperature of 75'F. Sources: Environmental Protection Agency, Mobile Source Emission Factors, Washington, D.C., March 1978. Federal Highway Administration, How to Prepare the Transportation Portion of Your State Air Quality Implementation Plan, Washington, D.C., November 1978. 72 To determine street and highway levels of service, volume/capacity (V/C) evaluations are conducted at two levels of detail. The more detailed of the two is for street and highway design and operation. The other is for conducting system analysis to assess the ability of the streets and highways to move the required amount of traffic at a satisfactory service level. It is the latter application that should be addressed with the procedures presented in this chapter. The techniques included have been selected to respond to intersection problems and corridor problems, The intersection, by definition, is site specific and is viewed as an entity, whereas the corridor problem is analyzed by comparing total traffic volume and traffic capacity within the corridor. The operating condition of streets and highways is stated using levels of service concepts of the Highway Capacity Manual 4/ as follows, Level of Service Operating Conditions A Free flow, low volume, high operating speed, high maneuverability. B Stable flow, moderate volume, speed somewhat restricted by traffic conditions, high maneuverability. C Stable flow, high volume, speed and maneuverability determined by traffic conditions. D Unstable flow, high volumes, tolerable but fluctuating operating speed and maneuverability. E Unstable flow, high volumes approaching roadway capacity, limited speed (- 30 m.p.h.), intermittent vehicle queuing. F Forced flow, volumes lower than capacity due to very low speeds. Heavy queuing of vehicles, frequent stoppages. The transportation planning team must decide the level of service to be used as the goal of the community. The decision is dependent upon several factors including financial resources to correct deficiencies and established policy. Historically, Level of Service "C" has been used as this benchmark and is used herein for capacity analysis. The user should differentiate between a desirable operating capacity and the physical or maximum capacity of a facility. The analyst may select a desirable operating capacity not to be exceeded -- such as Level of Service "C" or "D" -but physical capacity is defined at Level of Service "E" 4/. There are two basic and independent indicators of level of service -- the volume-to-capacity (V/C) ratio and the operating speed. The procedures described 73 below rely on the V/C ratio to assess service levels, For further discussion of the measurement of traffic capacity, refer to the Highway Capacity Manual Evaluating-Intersection Capacity The intersection capacity analysis described is based on the "Critical Movement Summation" technique developed by McInerney and Petersen 24/, and adopted by many local and regional planning agencies for use in development impact studies 25/. In using the technique, a critical intersection volume is calculated and compared with the benchmark intersection capacity that is stratified by level of service. Table 16 shows the benchmark capacity ranges for intersections at each level of service. Information Required. It is assumed that an intersection capacity analysis is undertaken after a study of existing conditions and after vehicle trips have been assigned to the highway network. The following information is required for the analysis: - Number of approach lanes - defined as the number of through lanes. - Exclusive use lanes (left turn, etc.). - Peak hour traffic volumes and turn volumes for all intersection movements. (Refer to Chapter Six, "Time-of- Day Characteristics," of the Users Guide 9/ for method to convert ADT volumes to peak hour volumes.) - Special operating characteristics that might affect lane volumes, such as free right turns. It is desirable for traffic volumes to be obtained for each lane. If lane volumes are not available, the total directional traffic flow can be converted into lane volume using the following factors: Approach Lanes Lane Use Factor 1 1.00 2 0.55 3 0.40 4 0.30 The lane use factors allow an estimation of the volume in the heaviest lane from the total approach volume. The "Critical Movement Summation" Technique This technique defines critical movement volumes as "The volume of travel represented by the highest lane volumes of opposing travel (through and left turn) from both the north-south and east-west directions that occur during the "peak hour." 74 TABLE 16 INTERSECTION CAPACITY*BY LEVEL OF SERVICE Range of Capacity (VPH) ____________________ Level of Service Low High A 0 900 B 901 1050 C 1051 1200 D 1201 1350 E 1351 1500 F (Special Case) 1500 * Capacity as defined by single lane through movement plus opposing single lane left turn movement. See discussion on "Critical Movement Summation" technique. Source: Highway Capacity Manual 4/, Guidelines for Transportation Impact Analyses as to the of Public Facilities Associated with Preliminary Plans of Subdivisions 25/ 75 The technique is best illustrated by an example application, such as shown in Figure 10. Volume numbers shown represent peak hour traffic. If only ADT volumes are available, these must be converted to hourly volumes. For default values see Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters: Users Guide 9/. The following steps will result in the intersection level of service. 1. Determine the net approach volume (through volume) and multiply by the appropriate lane use factor to get lane volume. (If lane volumes are available this step is not necessary.) Direction Net Approach Vol. Lane Use Factor Lane Volume Northbound 1,000 0.55 550 Southbound 800 0.55 440 Eastbound 750 0.55 413 Westbound 700 0.55 385 2. Determine the critical lane volume for each approach as follows: N-bound S-bound E-bound W-bound Through Volume 550 440 413 385 Opposing Left Turn Volume 175 200 150 100 ____ ____ ____ ____ TOTAL 725 640 563 485 3. Select maximum of N-S volumes and E-W volumes and sum to determine "Critical Movement Summation" (CMS). CMS = Northbound + Eastbound CMS = 725 + 563 CMS = 1,288 vehicles 4. Compare CMS to volume ranges listed in Table 16 to determine that the intersection is operating at Level of Service D. The example is representative of conditions at most intersection configurations, the exception being the unprotected left turn. For that case calculate the lane volumes as previously described but add the left turn volume to the critical lane volume to account for the interruption caused in the through lane by left turning vehicles. Evaluating Corridor Capacity The objective of the corridor analysis is to produce an assessment of volume to capacity (V/C) relationships for the corridor. The technique to accomplish this consists of "cutline evaluations" and preparation of a "Facilities Stress Diagram." Each is described below. 76 Click HERE for graphic. 77 Cutlines are strategically placed lines perpendicular to the direction of the travel being Analyze. Figure 11 shows a schematic representation of a corridor with five cutlines. The "Facilities Stress Diagram" is a plot of the V/C ratio for either the entire corridor -- after summing volumes and capacities across each cutline -or for a major street or highway (as in Figure 12 for routes A, B or C). An example of a stress diagram is shown in Figure 12. The diagram represents the corridor shown in Figure 11. The datum (horizontal line) is the benchmark level of service, measured as a V/C ratio. The diagram illustrates where the system is congested (over capacity) and where surplus capacity exists in the corridor. One conclusion from the diagram is that a large volume of traffic is leaving the corridor--probably from the freeway at the interchange between cutlines 1 and 2--resulting in a surplus of capacity between cutlines 2 and 4. Traffic begins to build up between cutlines 3 and 4 probably due to entering freeway volumes at the southern most interchange. Data Requirements. The corridor capacity analysis procedures require information to allow a V/C calculation for the corridor and each facility. Traffic assignment estimates of traffic volumes are needed in the form of directional peak hour volumes for capacity analysis. Refer to the Users Guide, Chapter Six 9/, for information which will allow conversion of ADT assignment volumes into hourly directional volumes. Also required is descriptive information for the streets and highways being analyzed. The needed descriptive information is as indicated in the calculation sections that follow. Freeway/Expressway Capacity Calculation. Access and parking are controlled on freeways and expressways and, consequently, the capacity determinations for freeways are less complicated than for arterial streets and highways. To facilitate the calculation of freeway capacity, assumptions have been made regarding average operating conditions and these are shown in the footnotes to Table 17. Table 17 also indicates tile resulting capacities in a form that requires the user to provide only the number of lanes per direction in order to determine facility capacity. The capacity quantities assume a peak hour factor of 0.85*, and 10 percent trucks. Volume-to-capacity (V/C) ratios are provided to give an idea of the maximum usage rates a freeway should experience for any given service level. It must be noted that V/C ratios and operating speed actually determine level of service: therefore, the relationships shown are approximations based only on V/C ratios. Table 17 contains directional capacity data, by level of service, for 2, 3 and 4 lanes of freeway and an additional column for "add-on" capacity for each lane over four lanes. ___________________________ * Peak hour factor is the ratio of the peak hour flow to twelve times the peak 5-minute flow that occurs during the peak hour. 78 Click HERE for graphic. 79 Click HERE for graphic. 80 Click HERE for graphic. 81 Urban Arterial Street Capacity Calculation. For capacity determination, urban arterial require more information than freeways and expressways. The capacity tables shown in Tables 18 through 22 allow the consideration of various operating characteristics. The capacity of an urban arterial street or highway is generally controlled by the capacity of intersections. Obviously, there are occasional situations where mid-block conditions control capacity. In such cases judgement will have to be exercised. Except in cases where mid-block conditions control capacity, the user should be guided by the rule: capacity is equal to the capacity of the most restricted intersection. The following list represents the information needed to determine urban street capacity: 1 - Approach width. Approach width is equal to the curb-to- curb street width for one-way streets and the painted centerline- to-curb street width for two-way streets. Boulevards are considered as two one-way streets with the approach width of each measured from the median edge of the roadway to the curb. Approach width excludes any special purpose (i.e., turning bays) lanes. 2 - Load factor is assumed equal to 1.0 in the tables provided. This corresponds to a Level of Service of "E" which represents available capacity. Load factor is defined as the number of green cycles that are fully used in the peak hour divided by the number of available green cycles occurring during the peak hour. 3 - Peak Hour Factor (PHF) is equal to the observed peak hour volume divided by four times the volume observed during the peak 15 minutes of the peak hour. The tables assume an average PHF of 0.85 26/. 4 - Percent commercial vehicles, percent left turns and percent right turns. The following example describes the arterial street and highway capacity calculations. Approach Width - 20 feet Commercial Vehicles - 15% Left Turns - 15% Right Turns - 10% Street Operation - 2-Way with Parking Volume - 900 Vehicles Reference to Table 18 indicates a capacity in the range of 950 vehicles per hour. The V/C ratio (see next section, "Determination of Level of Service") is calculated as 0.95. This provides a level of service between "D" and "E" which is indicative of the operation of the street, slow speeds - heavy congestion. It is anticipated that over the next five years the volume will increase to about 1,200 vehicles per hour. It is desired that a Level of Service `C" be provided (V/C = 0.8). With a volume of 1,200, a capacity of about 1,500 is required. Review of Tables 18 through 22 indicates the possibility for improvement. 82 Click HERE for graphic. 83 Click HERE for graphic. 84 Click HERE for graphic. 85 Click HERE for graphic. 86 Click HERE for graphic. 87 If the street is not widened, Table 18 indicates that a 1,500 vehicle capacity can be achieved if parking is removed (2-way no parking) and commercial vehicles routed elsewhere. This would provide a capacity of about 1,533. Another possible solution is removal of parking and eliminating left turns, and these measures provide a capacity of about 1,525. If parallel streets are available for one-way pairs, a one-way operation with no parking is a possible solution. If a two-way operation must be maintained with on-street parking, the charts indicate that the street must be widened to provide an approach width of greater than 40 feet. Determining Levels of Service. For urban arterial streets and highways, the maximum V/C ratio for each level of service is shown below: LEVEL OF SERVICE MAXIMUM V/C RATIO A 0.6 B 0.7 C 0.8 D 0.9 E 1.0 F (varies) Figure 13 is an extended representation of Figure 12 to show the impact of specifying a level of service to determine parts of the street or highway system where capacity is exceeded. The impact of establishing a Level of Service "C" instead of using Level of Service "D" as the datum can be readily seen in Figure 13. To gain the higher service level will require capital or maintenance funds to remedy the deficiencies. While a service level of "C" may be the least cost solution, there may be a need to reassess the situation in light of available funds. DETERMINING SOLUTIONS The development of solutions to problems revealed in system analysis involves the evaluation of alternatives. Heavy traffic volumes through a downtown corridor, for example, might indicate several potential solutions such as: re-routing all through traffic or trucks only; building a bypass; eliminating on-street parking; and the development of one-way pairs. The types of solutions to be considered will vary by the conditions surrounding the problems, the goals and objectives established for the area, costs, economic implications, and so on. The material provided in this and other manuals of this series allow the evaluation of alternative solutions. Problems to be addressed are indicated by the demand estimation and traffic assignments to the forecast period as well as from surveillance of the current system operation. A set of problems will emerge from this analysis that will exhibit the following considerations: 88 Click HERE for graphic. 89 - Current problems that do not appear to worsen over time - Current problems that worsen over time - Future problems that are not problems now The problems uncovered should be solved one-by-one. The long range analysis should be viewed as providing an indication of future corridors of demand, the type, locations, and severity of possible deficiencies, and should form a framework for the development of solutions. Current problems that do not appear to worsen over time can be resolved by traffic engineering improvements, other Transportation System Management (TSM) measures, and higher capital investment such as widenings, a bypass, etc. Current problems that are likely to worsen over time require solutions that may be staged over time, an interim solution to be improved upon later, or an ultimate improvement expected to satisfy current and anticipated future demand. In the latter case, a TSM strategy may make a capital-intensive solution unnecessary, postpone its implementation, or alter its design. Conversely, early stages of a programmed long range element could affect the type of TSM strategy proposed as an interim measure. For example, congestion on a major arterial can be relieved in several ways: adding lanes, channelizing intersections, eliminating parking, or improving signal systems. If the long range analysis reflects a need for additional lanes, channelized intersections may be the appropriate short range solution, because it permits street widening at a future time. If a new facility is needed to relieve the traffic in the longer range, a new signal system may be appropriate now. In addition, areawide actions such as parking and other terminal and transfer facility needs, carpooling programs, systemwide traffic engineering and transit technology studies, and paratransit projects need to be analyzed, and their impact on the long range elements measured. Finally, future problems that are not now problems must be carefully considered. Many future problems will require short range solutions such as TSM actions including traffic engineering improvements. These future problems can be monitored through time and appropriate actions can be taken when required. Other future problems may warrant capital investments and solutions that take many years to implement, such as a new bypass route. Again, an immediate action might be the acquisition of right-of-way. A few veers ahead. analyses may continue to indicate the bypass solution and design work might be undertaken. Or, perhaps future analysis indicates the anticipated problem is not occurring and consideration of the bypass might be dropped in favor of less capital-intensive solutions. The above discussion should indicate that alternative short range actions are developed from near-term needs and early stages of major capital projects reflected in the refined long range plan element. The balance between TSM and more capital intensive solutions will reflect such things as city-size, problem complexity, and growth potential. It is vital to the effectiveness of planning that a great deal of emphasis be placed on short range planning and that it be made clear that the remaining stages are subject to 90 further study as planning continues. To this end, it is necessary to make a comprehensive study of short range needs to arrive at a sound and practical short range program for highway and transit. Most likely, such a study will be based on an analysis of today's conditions and a comprehensive regional analysis of short range TSM alternatives. In other cases, the short range program could be composed of TSM strategies developed by the local implementors or operators. Depending on local programming policies, this short range period may cover from 3 to 8 years. To determine potentials for traffic engineering improvements of a TSM nature, the reader is referred to the Traffic Planning manual of this series. The Transit Planning manual indicates transit improvement potentials. Earlier sections of this chapter provide capacity analysis methods to evaluate alternate improvements. Chapter One of this manual suggests improvements related to system functions. To evaluate alternative solutions to a particular problem, the reader is referred to material in this chapter. The evaluation should be based on costs and both positive and negative impacts. 91 REFERENCES 1. National Committee on Urban Transportation (NCUT), Better Transportation for Your City: A Guide to the Factual Development of Urban Transportation Plans, (Brattleboro, Vermont, Public Administration Service, 1958). Also 17 Technical "Procedural Manuals" on various activities. Recommended ones as referenced in the text of this document. 2. U.S. Department of Transportation, Guide to Urban Traffic Volume Counting, Washington, D.C. 3. U.S. Department of Transportation, Traffic Assignment-- Methods-Applications-Products, (Washington, D.C.,FH14A, August 1973). 4. Highway Research Board, Highway Capacity Manual, Highway Research Board Special Report No. 87, (Washington, D.C., 1965). 5. U.S. Department of Transportation, Land Use Forecasting Techniques for Use in Small Urban Areas, Vols. 1 and 2, (Washington, D.C., FHWA, August 1977). 6. U.S. Department of Transportation, Urban Origin-Destination Surveys, (Washington, D.C.). Dwelling Unit Survey, Truck and Taxi Surveys,'External Survey. 7. U.S. Department of Transportation, Trip Generation Analysis, (Washington, D.C., FHWA, August 1975). 8. U.S. Department of Transportation, Urban Mass Transportation Travel Surveys, (Washington, D.C., MIA and UMTA, August 1972).. 9. COMSIS Corporation, Quick-Response Urban Travel Estimation Manual Techniques and Transferable Parameters: Users Guide, National Cooperative Highway Research Program, Report 187, (Washington, D.C., 1978). 10. U.S. Department of Transportation, Transportation Planning System (UTPS) Reference Manual, (Washington, D.C., periodically released). 11. U.S. Department of 'Transportation, PLANPAC/BACKPAC General Information, (Washington, D.C., Federal Highway Administration, April 1977). 12. Department of Labor, Handbook of Labor Statistics, 1972 Bulletin 1735, (Bureau of Labor Statistics, 1972). 13. U.S. Department of Commerce and U.S. Department of Agriculture, OBERS Projections, Economic Activity in the U.S., Vols. I-VIII, (Washington, D.C., Bureau of Economic Analysis and Economic Research Service for the U.S. Water Resources Council, 1972). 14. Long, Gary D., An Evaluation of the Gravity Model Trip Distribution, (Texas Transportation Institute, 1968). 92 15. Highway Research Board, Effect of Zone Size on Traffic Assignment and Trip Distribution, Highway Research Record, No. 392, (1972). 16. Pratt Associates, R.H., A Method for Distributing Traffic Volumes Among Competing Highway Facilities, (Kensington, Maryland, 1976). Unpublished working paper. 17. Neffendorf, H. and Wooton, H.J., A Travel Estimation Model Based on Screenline Interviews, PTRC Summer Annual Meeting (1974). 18. U.S. Department of Transportation, Land-Use and Arterial Spacing in Suburban Areas, (Washington, D.C., FHWA, May 1977). 19. COMSIS Corporation, Travel Estimation Procedures for Quick Response to Urban Policy Issues, National Cooperative Highway Research Program, Report 186, (Washington, D.C., 1978). 20. National Cooperative Highway Research Program, Transportation Decision Making--A Guide to Social and Environmental Considerations, Report 156, (Washington, D.C., 1975). 21. U.S. Department of Transportation, Accessibility--Its Use as an Evaluation Criterion in Testing and Evaluating Alternative Transportation Systems, (Washington, D.C., FHWA, July 1972). Highway Planning Technical Report. 22. U.S. Department of Transportation, Cost of Owning and Operating an Automobile ] 1979, (Washington, D.C., FHWA, 1980). 23. U.S. Department of Transportation, Characteristics of Urban Transportation Systems--A Handbook for Transportation Planners, (Washington, D.C., FHWA and UMTA, July 1977). 24. McInerney, H. and Petersen, S..,"Intersection Capacity Measurement Through Critical Movement Summations: A Planning Tool,"Traffic Engineering, (January 1971). 25. Maryland-National Capital Park and Planning Commission, Guidelines for Transportation Impact Analyses as to the Adequacy of Public Facilities Associated with Preliminary Plans of Subdivisions, (September 1974). Unpublished Technical Memorandum. 26. Pignataro, Louis J., Traffic Engineering--Theory and Practice, (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1973). 93
