A Performance-Based Highway Geometric Design Process (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

68 This section of the report outlines a performance-based geometric highway design process. The highway design process is a major part of the overall transportation project development process. Successful completion of key elements and tasks of the overall process are essential to the success of the highway design process. Figure 22 illustrates a typical, simplified project development process that is modeled after many state DOT processes. NCHRP Report 480 (Neuman et al. 2002) documented a framework for project development that reflects the evolution of highway projects over the past 30 years. The critical success factors to highway and road project completion are fourfold: • Employ effective decision-making process; • Reflect community values (i.e., include stakeholders); • Be environmentally sensitive; and • Implement safe and feasible solutions. These critical success factors align with the guiding principles discussed earlier. The importance of geometric design decision making reflects the accountability of the agency to its customers and to those providing the resources and funds. Reflection of community values and sensitivity to the environment emphasize the importance of context as broadly defined in what constitutes reasonable outcomes. Finally, the notion of safety as being central to successful design policy is self-evident; and the term feasible applies to both the individual project and its role or contribution to the overall road network. Feasibility encompasses both the physical aspects, footprint, and the affordability of the project. These critical success factors are reflected in the highway design project development steps outlined below for the revised a performance-based geometric road design process (see Figure 23). 5.1 Step 1—Define the Transportation Problem or Need All street, highway, or transportation projects have common characteristics. The first and foremost step is to define and articulate a transportation need. There must be a reason why the owner is spending time and money. Agency programs are developed around the continual oversight of the transportation performance measures of mobility and access (congestion, LOS), safety (crashes and their outcomes), and state-of-good repair (pavement and bridge asset management). A project may also emerge from a need to fulfill a mission or vision of a community such as redevelopment or urban renewal, which could create an accessibility to land problem. Historically, the highway design process has, in many cases, been performed independently of the statement of purpose and need. For reasons of economy, efficiency, and a focus on C h a p t e r 5 Performance-Based Highway Design Process

performance-Based highway Design process 69 cost-effective solutions, the geometric road design process should be strongly bound to and driven specifically by the problem or needs being addressed. Clearly defining the problem is the first step in an effective decision process. A feasible solu- tion can only be one that clearly addresses the problem. AASHTO reinforces the importance of this first step in its policy document on highway design flexibility (AASHTO 2004). Defining the problem in all its transportation aspects is a critical initial step. Only after a clear understanding has been reached by all stakeholders regarding the need for the proj- ect, including what transportation issues and problems it is intended to address, can progress toward problem solving and actual implementation occur. AASHTO Guide for Achieving Flexibility in Highway Design Figure 22. Roadway design development process. Figure 23. The steps of the performance-based highway design process.

70 a performance-Based highway Geometric Design process Of great importance is the need for each project’s problem statement to be objective. Problems with pavements may be rutting, lack of drainage, or cracking. Bridge condition problems may vary from substructure condition, to vertical clearance, to superstructure condition, or load-bearing capability. Transportation mobility problems should be defined in explicit, performance-based terms (e.g., travel time during x hours of the day, delays at location y, distance to walk from A to B). Substantive safety problems should be defined in terms of the frequency of specific types of crashes at specific locations. Accessibility should be addressed in terms of modes and specific origins and destinations of trips. Subjective and vague problem statements referring to user perceptions, or livability, are unhelpful and should be avoided absent further explanation of what the terms mean. Problems based on objective performance measures should directly reflect agency policies and priorities. These are translated and communicated through reference benchmarks that are statistically valid and relevant to the specifics of the project. Thus, a pavement considered in need of repair earns that designation through objective condition assessments and references to the agency’s asset management approaches to pavement. Safety problems are identified through the appropriate statistical analyses of crash data by frequency, type, and severity. Qualitative or opinion-subjective-based problem statements should be avoided as they offer no basis for validating the expenditure associated with a geometric or other solution. This thought also applies to agency programs with specific performance focuses, for example, the systemic safety approach. The importance of a strong, objective problem statement in the context of geometric design cannot be overstated. An appropriate, geometric design solution is one that will address the problem in meaningful, measurable ways. Moreover, as the nature of the problem is unique to the location, the specific geometric design solution may be unique, or at least may be different from the solution applied at another location with similar context features but with different problems. 5.2 Step 2—Identify and Charter All Project Stakeholders Highway and transportation projects are intended to provide a service and benefit to the general public. For the most part, they will be funded using public resources. They may require the conversion of privately owned land to public right-of-way. They will require the ongoing allocation of funds for the upkeep, maintenance, and operation of the roadway. They may create impacts, some of which may be adverse, on adjacent properties or entities. For all of these reasons, the highway geometric design process must be viewed as an open, transparent, and ultimately public process, and not just the purview of the technical staff of the transportation agency. Every highway project regardless of size or location involves multiple stakeholders. The term stakeholders refers to any individual or organization having a stake in the project. The stake may be direct as is the case for the owner of the road, the users of the road, and those directly influenced by its design and construction. Some stakeholders may have more indirect stake in the project. Indirect effects may include socioeconomic impacts in the area or region, expected changes in land values or utility in proximity to the project, aggregate regional environmental effects, or other similar impacts that may affect people but are less measurable directly or are uncertain in their timing and extent. A highway design project team cannot possibly be successful without understanding and incorporating the meaningful input of key stakeholders in all relevant aspects of the project.

performance-Based highway Design process 71 5.2.1 Internal Agency Stakeholders Within the agency that owns the project, the stakeholders include all different departments from which data, input, or assistance is needed. In addition to the road design department, other internal stakeholders will typically include traffic engineering, safety, maintenance, construction, public relations, and environmental planning. City and county internal stakeholders also may include departments responsible for planning and zoning, economic development, and law enforcement. Any department that may be asked to provide resources to implement and maintain the project is an internal stakeholder that should be fully engaged or represented throughout the project. 5.2.2 External Agency Stakeholders Other governmental bodies may be affected by or have regulatory power over one or more aspects of the project. External agency stakeholders may include a local government or county department of public works, the U.S. Army Corps of Engineers, state natural resource and environmental agencies, state cultural resource agencies, and others. These stakeholders may be responsible for review and approval of project plans, permitting, and ensuring mitigation of identified adverse project effects. Other key external agencies include law enforcement and emergency service providers, local transit agencies and providers, and public utilities. Their common stake is in the use of the road or road network to conduct their jobs, locate their facilities, and serve their customers. 5.2.3 Other External Stakeholder Groups or Agencies Other external stakeholders may include organizations, groups, or agencies that provide data or input, give advice and counsel, or represent general interests having to do with transportation or land use and social policy. These may include: • Political jurisdictions within which the roadway passes; • Metropolitan planning organizations (MPOs); • Non-governmental organizations; • Local environmental agencies; • Social outreach agencies; • Chambers of Commerce and other business groups; • Schools and school districts; • Community interest groups; and • Road user groups such as AAA, the freight and goods community, and bicycle advocacy groups. These external stakeholders will often become heavily engaged in larger projects of a regional nature. External stakeholders typically do not have a direct role in project decision making other than advisory or reaction and comment, but their influence individually and collectively will often substantially shape the design solution. 5.2.4 Directly Affected Stakeholders Stakeholders who are directly affected include those whose property may be taken, or access affected, during construction or in the final plan. The input of these stakeholders has a direct influence on important road design decisions. Permanent spatial impacts to properties and access, to mobility in the corridor for all users, and impacts during construction (road closures and detours, noise and dust, and construction vehicle routings) are all common issues that highway designers must consider.

72 a performance-Based highway Geometric Design process 5.2.5 Stakeholder Chartering The concept of chartering refers to formal acknowledgement of each stakeholder’s presence and reason for having an interest in the project. It includes agreement on the roles of stakeholders, which may range from advisory to decision making; timing and methods of communication; and, most importantly, validation of the problem being addressed. Chartering as a subprocess is scalable. For many projects, it may consist of a single meeting at which the basics of the project are discussed, communication protocols set, schedules discussed, and agreement reached on timing and nature of each individual’s input. Some projects may involve mostly internal agency stakeholders and hence chartering may become an internal routine. An important aspect of chartering regardless of the project size is the documentation and joint ownership of the project by all involved. This is often accomplished by having stakeholders sign and acknowledge the notes or meeting minutes that document decisions and actions. Chartering as described here can be included as part of Step 3, project scoping. 5.3 Step 3—Develop the Project Scope The term scoping refers to the setting of the parameters of the project. Scoping includes engaging the many stakeholders with varying interests and concerns. Regardless of the purpose and need, these concerns will shape the alternatives and ultimately drive or influence the selection of the preferred alternative. Broadly stated, stakeholder concerns will involve one or more of the following areas: • Implementation and life-cycle costs including maintenance; • Need for right-of-way (amount, location, conversion of existing use); • Environmental considerations (air, noise, water quality, socioeconomic effects, visual effects; construction issues such as dust, truck movements); • Specific community concerns such as key landmarks, important land uses, or buildings; • Transportation operations and service (types and modes of travel, quality of service, access to abutting properties); and • Public safety (users of the road or facility, safety interactions of stakeholders with users). Depending on the project, many of these issues may be conducted within a strict set of regula- tory protocols or requirements. Scoping serves to define the amount, type, and eventual use of data gathering and analysis steps. In the case of programmatic projects, such as a systemic safety project, the scope will include the jurisdictions and road types to be included. Within the context of geometric design, there are three key elements to project scoping and all of the elements have an impact on each other—(1) refinement and confirmation of the problem statement and definition with stakeholders, (2) determination of project type, and (3) setting of project and study limits and major study parameters and constraints. 5.3.1 Refinement and Confirmation of Problem or Needs Statement The transportation problem or needs statement developed in Step 1 should be presented, discussed, confirmed, and/or revised based on input from key stakeholders both inside and outside the owning agency. The importance of this step is twofold. First, efficiency demands those involved in the project understand and maintain the focus on why the project is being performed. Limitations in resources and schedule pressures are such that every effort should be made to avoid diversion from the core mission. The sharing of the problem can uncover oppor- tunities to refine or expand the project to include other objectively defined problems within the project that are directly related to or coincident with the road or portions thereof. Second, the

performance-Based highway Design process 73 specific definition of the problem should translate directly into the appropriate geometric design process, as discussed in detail below. Stakeholder discussions should be facilitated in a manner that generates objective and well- understood problems expressed in neutral ways (i.e., that do not bias or presuppose a solution). Stakeholders should understand agency data and analysis methods that express problems associated with congestion, delay, or safety performance. Vaguely described concepts expressed in terms such as sustainability or livability are often not helpful. Opinions (the road is unsafe) must be supported by objective relevant data to be useful. Expressing problems as lack of xx should be avoided, as the solution is implied in the way the problem is expressed. Finally, to the extent possible, problems should be referenced to the location or locations where they exist or are perceived, or to the specific type of conflict or crash type. A useful statement of purpose and need will direct the development and evaluation of potential solutions. It will facilitate the selection of a solution that produces measurable value in terms of solving all or part of the problem for the resources invested. Confirmation of the problem in objective and data-driven terms clearly communicates to all stakeholders what the agency’s approach will be in the geometric design process. Problem statements should be understood as driving the approach to the geometric design solution. This entails geometric elements to include, their dimensions, and locations (e.g., intersection, curve of specific concern). A refined, formal problem statement is the key outcome from Step 3. This should be high- lighted in the chartering document. It also will form the basis of the purpose and need, which is a central element of the environmental process. 5.4 Step 4—Determine the Project Type and Design Development Parameters Once the stakeholders have been consulted and the problem statement published, the agency makes a determination of the type of project. Project type refers to the pre-existing condition and fundamental nature of the identified problem(s) or needs. The performance-based geometric design process encompasses these three basic project types: • New roadway (on new alignment with newly acquired right-of-way); • Reconstructed roadway (along existing right-of-way but with substantial change in the roadway characteristics and potential additional right-of-way); and • 3R roadway [resurfacing, restoration, and rehabilitation of an existing roadway within the existing right-of-way, with only minimal changes to the road’s geometric characteristics (at most)]. Project scoping should include the confirmation of the type of project, which should direct and drive the highway design process and ultimately the solution. New roadway projects are fundamentally different from reconstruction and 3R projects in ways of sufficient import to warrant treating them differently. To review, by definition, a reconstruction project involves a roadway already in place and functioning. Similarly, a new construction project (one on new alignment) introduces a new transportation corridor where none exists. The following differences are evident: • With an existing road, the abutting land use has developed around the mobility and access provided by the road; with a road on new alignment the road itself always produces significant change (positive and potentially negative).

74 a performance-Based highway Geometric Design process • With an existing road, there is a clear and well-understood operational and safety performance history; none exists for a road on new alignment. • The costs of construction and constructability of each project type are based on significantly different factors. 5.4.1 Unique Characteristics of New Roads New roads or projects on new alignment are different from other project types. First, a new road is typically built to address problems of access and/or mobility only (and not to address an existing safety problem). New roads will substantially change the surrounding land use. In many cases, such changes are not only positive but intentional. Providing access to land creates development opportunities. Segmentation of large parcels such as agricultural land changes the economics and potential value of such lands. Existing regional and local travel patterns will substantially change, as the new road presents rerouting opportunities. Forecast traffic volumes and travel patterns are inherently more uncertain for new roads. Travel behavior including speed, crashes, and conflicts may be estimated based on tools and models, but the actual route choices of potential users are unknown prior to opening of the road. The costs and difficulties of constructing new roads are typically associated with context features of the terrain, geology, and water courses. Alignment siting studies focus on avoidance of conflicts with concurrent utilities, and respecting property boundaries by avoiding unusable remnant parcels and assuring public access. Cost estimates for construction are primarily based on the quantities of major elements including earthwork, structural features, and pavement. Given the uncertainties inherent in new road construction, the geometric design approach may differ from existing roads. Agencies may reasonably acquire sufficient right-of-way to enable future changes to the road based on potential unforeseen development and traffic growth beyond the nominal design year. Bridges and other major structures may be built to accommodate potential widening beyond the design year, especially when the incremental costs of construction of wider crossings is minor. Finally, a geometric design process utilizing traffic operational and safety performance models should consider the uncertainty in the input data to the models and their overall quality or reliability. 5.4.2 Unique Characteristics of Reconstruction Projects Properties with site development and access directly tied to the existing road geometry within a fixed right-of-way footprint makes reconstruction projects different from roads on new align- ment. This difference tends to be magnified in built-up urban and suburban areas. Where real estate is highly valued, developments will use every foot of available land, and public rights-of-way will be limited to the area needed for all infrastructure. The abutting land use will often be fully occupied by buildings, parking, planned developments, and plazas. In-place developments will reflect local ordinances for setbacks, parking requirements, storm water mitigation, and other features. Utilities, both underground and aboveground, are situated with reference to the existing road. 5.4.2.1 Effects of Additional Right-of-Way Acquisition Mature developments are the economic and social fabric of the community, and are the source of employment and tax revenues. Reconstruction of an existing road bears a special, significant burden of minimizing damage or disruption to property owners (businesses, residents, govern- ment agencies) who invested in their land based on the right-of-way and footprint of the exist- ing road. Reconstruction projects that require taking of new right-of-way, even as much as a

performance-Based highway Design process 75 relatively small strip taking of 5 feet will often produce adverse impacts beyond the direct cost of the right-of-way itself. Such impacts may include: • Elimination of building setbacks required by local ordinance, thus rendering the property as non-conforming to such ordinance; • Loss of mature landscaping and plantings serving as visual buffers between the road and development; • Loss of site parking, again rendering the property as non-conforming and potentially damaging the economic viability of the development; • Newly created conflicts with existing utilities (both subsurface and aboveground); and • Reduction of the sidewalk widths/offsets to the travelway. It is not uncommon for the taking of even a small strip to be considered so adverse as to render the remaining property uneconomic and hence requiring a whole taking. 5.4.2.2 Changes in Access and Their Effect on Adjacent Properties Reconstruction projects will also frequently produce changes in access. Such changes may be related to necessary realignment, safety enhancements, and/or operational improvements. In most cases, changes in access (e.g., loss of driveways, closure of medians, consolidation, or relocation of access) will be considered adverse by the abutting property owners. Depending on the nature of the land use (commercial and various types of retail), such changes, unless mitigated, can have measurable, permanent adverse effects on businesses. 5.4.2.3 Disruption to Traffic and Access During Reconstruction Produces Adverse Economic Impacts Reconstruction may involve changes to the three dimensions of the roadway. In most cases, maintaining traffic during construction along the corridor is required. Access to properties by emergency vehicles at all times is necessary. These important factors often result in even relatively straightforward geometric design projects requiring more than 12 months to complete construc- tion. Even when through traffic is maintained and temporary driveway connections provided, the quality of service during construction is much worse than before construction. When this condition exists for extended time periods, the economic impacts to businesses can be substantial. This problem may exist even for projects reconstructed completely within the right-of-way. This issue also occurs with 3R projects, but resurfacing and restoration (and not major changes in alignment or profile) take much less time to construct, and thus generally result in lesser economic burdens to adjacent landowners. 5.4.2.4 Costs of Construction Unlike new roads, the costs of reconstructing existing roads are based on many factors beyond the earthwork, pavement, and structural quantities. As noted above, the need to maintain traffic greatly influences the number of separate construction phases, viability of means and methods of construction, and length of time to construct. Site access to the existing corridor by construction vehicles may be highly restricted or limited (avoiding truck traffic on local streets or commercial areas). Time-of-day restrictions on truck activity are common. These necessary constraints can have substantial effects on the unit costs of earthwork, concrete, and other materials for the specific project. As utilities typically occur within the existing right-of-way, conflicts arise and the need to relocate the right-of-way (thus incurring additional phases and time) is inevitable. While these factors are typical, their relative importance and direct influence on an optimal design are highly context sensitive. For reconstructed roads, an optimal or minimal construction cost is a much more involved and variable exercise than for a comparable road constructed on new alignment.

76 a performance-Based highway Geometric Design process 5.4.2.5 Transportation Performance Is Known and Measurable Finally and most importantly, roads to be reconstructed have a known, observable traffic operational and safety profile. Indeed, it is the measurable performance of the road combined with agency policies and performance benchmarks that should have been used to frame the problem statement and define the need. Actual traffic volumes and patterns, crash frequency and severity, and speeds are measurable. The availability of such data is both unique and critical to reconstruction (and 3R) project design development. Traffic forecasts are generally more reliable, particularly where the road network and development are mature. The ability to use location-specific data of adequate quality is a unique, valuable aspect of reconstruction projects as compared with new alignment projects. 5.4.3 Unique Characteristics of 3R Projects The third type of project is 3R, which involves addressing the state of good repair of the infra- structure and only involves existing facilities. The land use and access impacts of reconstruction projects discussed above also apply to 3R projects. A project for which the state-of-good repair is the core problem is potentially eligible for 3R designation. A key stakeholder input process step may include discussion of whether or not other prob- lems identified by stakeholders could or should be incorporated into the project. For example, an urban street repair project to address failing pavement may be viewed by some stakeholders as an opportunity to provide or enhance pedestrian accessibility or mobility by incorporating sidewalks, intersection ramps, pedestrian signal upgrades, or restriping of the pavement to include bicycle lanes. These measures would address a stated problem associated with access and/or mobility of pedestrians or bicyclists. (NCHRP Project 15-50 is currently considering revised geometric design guidelines for 3R proj- ects to replace those developed in the later 1980s for TRB Special Report 214. The vision and approach of the authors of that effort are generally consistent with the approach suggested here.) 5.4.4 Project Types and Transportation Problems Figure 24 demonstrates the range of possible problems associated with the three project types. A project on new location will always involve addressing a mobility problem, accessibility issue, or both. An important process step for new location projects is a full stakeholder discussion and agreement on how the travel modes will be accommodated, including motor vehicles, freight, transit, pedestrians, and bicycles. Confirming and potentially expanding the problem definition based on stakeholder input is a key process step. A 3R project may proceed with additional features added without changing the basic project definition or purpose of the project. But depending on the project specifics and X X X X Project Type Mobility Safety State-of-goodRepair New Locaon X X 3R X Reconstrucon Transportaon Problem Access Figure 24. Project types and transportation problems.

performance-Based highway Design process 77 additional problems suggested, the scope of the project, applicable type, and design approach may significantly change. An effective and responsive highway design process resolves problem definition and designates the appropriate project type early in the process. A reconstruction project may involve any of the fundamental problems or combinations thereof. Indeed, what differentiates a 3R from a reconstruction project will be the identified problem(s) or need(s) associated with traffic operations, safety performance, or both. As with a 3R project, stakeholders should be consulted with respect to enhancing the problem definition to include the full range of multimodal interests. 5.4.5 Setting of Project Limits and Major Study Parameters or Controls Each project has geographic limits or boundaries. These can be as small as a single intersection, a street in front of a school, a short segment along a highway; or a corridor of some length that may contain one or more routes. The context of the project’s location should shape and determine study approaches, feasibility of alternatives, stakeholder involvement strategies, and performance characteristics. Agency stakeholders responsible for specific resources shape the scope of the project. They may identify resources to be avoided, remind project staff of regulatory requirements, and help establish technical protocols. The project limits and indeed those segments of existing road to which the project ties also are important design process features. A road segment being reconstructed should be designed with cross-section dimensions compatible with the sections of road to which it ties. This concept is referred to as the design domain, an element of road design decision making in Canada. Thus, for example, a long segment of road may have 10-foot lanes and 3-foot shoulders. If one portion of that road is to be reconstructed, with sections on either end to remain, reconstruction using 12-foot lanes and 6-foot shoulders (which may be the applicable standard) may not make sense operationally, and be more costly than using lesser dimensions more compatible with the existing roadway (this of course depends on the problem being addressed and whether the cross-section dimensions are attributed to the problem). Project limits for new roads and major reconstruction are subject to environmental regula- tions that require the project have independent utility. Knowledgeable environmental process stakeholders should participate in setting limits for such projects. 5.4.6 Determining Environmental Process and Documentation Requirements The scoping process in Step 3 includes the expected level of environmental analysis and documentation (federal, state, or local), which reflects the overall scope and understanding of constraints and expected impacts. It is based on input from key environmental stakeholders both inside and outside the agency. An important consideration is the extent to which new right- of-way acquisition is expected. This type of impact may in itself require a project to be performed as an environmental assessment (EA) rather than a categorical exclusion (CE). Recognizing that right-of-way impacts are uncertain at the beginning of the design process, the type of problem and type of project should provide firm direction to the geometric design team. Addressing pavement condition for a two-lane rural highway with no observable safety performance problems should generally not involve right-of-way, and there should be no compelling reason to (1) widen the lanes or shoulders, (2) flatten a horizontal curve, (3) lengthen a vertical curve, or (4) flatten a roadside slope or increase the clear zone.

78 a performance-Based highway Geometric Design process 5.4.7 Establishment of Planning Level Implementation Budget With the project type and limits established, a project construction or implementation budget and schedule should be readily set based on historic data. The budget should reflect the reasoned judgments of the type and nature of the constructed project elements. This budget should be a reference point against which the project manager and team will be evaluated during project tasks and following completion. Many major projects historically suffer from unrealistically low initial budgets. This issue has been studied and recent efforts led by the FHWA have focused on how such budgets are developed, and how they are monitored during project development (Federal Highway Administration Office of Innovative Project Delivery nd). Major projects seeking federal funding cannot proceed without a detailed financial plan, which matches financing sources with estimated project costs. Having a reasonable budget that reflects the unique project conditions, and that is continuously referenced throughout the project development process, is an essential element of a geometric design process. The importance of including a preliminary budget as part of the design process is to remind the project team of their project’s role within a greater agency program. A project designated as 3R should have a level of investment well understood by all involved. Changes in project circumstances that would produce substantive changes in the budget should be discussed thoroughly prior to investing resources or proceeding ahead. 5.5 Step 5—Establish the Project’s Context and Geometric Design Framework The geometric design framework is composed of external controls and constraints (the context), input data and assumptions, design controls that are choices or at the discretion of the designer or his/her agency, applicable policies of the owning agency, and decision-making processes and responsibilities for the project. Some aspects of the framework apply to all similar projects performed by the agency, and others are specific to the project. The project context is the set of conditions, controls, and constraints outside the control of the designers (givens). These include location, terrain, climate, land use, and political or jurisdictional boundaries. Current AASHTO design policy addresses the context in just two dimensions, characterizing location as urban or rural, and terrain as level, rolling, or mountainous. This framework has been in place since the first versions of the AASHTO policies in the 1950s. It is based primarily on motor vehicle operational needs and the costs and practicality of project implementation for roads on new alignment in differing area types and terrain. A much more robust framework is needed that further differentiates important external con- trols on appropriate geometric solutions. The need for a more robust framework is primarily associated with the need to tailor the geometric design process for contexts in which multimodal travel, and in particular pedestrian and other vulnerable user mobility and safety, should be the predominant concern. 5.5.1 Framework for Geometric Design Process— New Construction and Reconstruction This section outlines a suggested framework and straw-man demonstration of its application in the development of a revised basis for geometric design criteria for new construction and reconstruction projects. The framework incorporates the basic principles discussed above. It

performance-Based highway Design process 79 considers both transportation function and value, as well as costs and cost-effective principles. Figure 25 shows the format of a matrix for the context framework for geometric design criteria. The two-dimensional context framework includes road type and an expanded, land use context definition. 5.5.1.1 Land Use Context The more nuanced context definitions better reflect the range of land uses, and most impor- tantly, address the expected presence of and unique requirements associated with pedestrians and bicyclists. Research has established a basis for quantifying the level of activity associated with land use by type and density. As mentioned in section 3.2.2, the best current model for converting land use into usable context definitions for a geometric design framework is that developed in the ITE, Designing Walkable Urban Thoroughfares: A Context Sensitive Approach—ITE Recommended Practice (Institute of Transportation Engineers 2010), which is recommended for adoption (Figure 26). Roads in rural context zones are traveled primarily by motor vehicles. The types, speed regime, and operating conditions will vary based on road type. Two types of rural context zones are pro- posed, consistent with the ITE informational report. • Rural natural zones are those with highly sensitive and valued environmental features. Road users are made aware of the special nature of the surroundings. Rural natural zones may include Roadway Type Rural Natural Zone Rural Zone Suburban Zone General Urban Zone Urban Center Zone Urban Core Zone Special Purpose Roads Local Collector Arterial Freeway Figure 25. Geometric design context framework. Source: (Institute of Transportation Engineers 2010), image courtesy of Duany Plater-Zyberk and Company. Figure 26. Roadway context zones.

80 a performance-Based highway Geometric Design process national and state parks, national forests, and lands proximate to such lands with similar envi- ronmental features. For all road types and functions, geometric design elements are purposely minimized dimensionally to reinforce the primacy of protecting the character of the zone. Travel is mainly by automobile. Bicycle and pedestrian facilities are primarily for recreational purposes. • Rural general zones are all other areas beyond urbanized jurisdictions. The automobile is still the predominant travel mode. Any developed areas generally have large setbacks not conducive to transit or non-motorized travel. • Suburban zones are areas in which development, land use, and population density are such that motor vehicles are the primary means of mobility, but the presence of other users (pedestrians and bicyclists) becomes evident and must be considered. Most suburban neigh- borhoods have sidewalks or paths for pedestrian and/or bicycle travel among the residences, schools, parks, and the few commercial uses in those neighborhoods. Design for transit (primarily buses) also may be a design consideration in suburban zones. Roads in urban context zones include contexts in which land development is relatively dense, but can be varying in nature. Urban areas include residential contexts including single family and multifamily, industrial and commercial districts, office parks, and retail. Zones within the urban context would include consideration of multimodal users including specifically pedestrians, bicyclists, and transit vehicles (buses). • General urban zones include land uses outside the city center, with generally less dense land development. Industrial parks and land activities will be within the general urban zone. • Urban center zones include primarily higher value, more densely developed land, including multifamily housing, offices, and retail or commercial lands. Limited parking and congestion make automobile travel less efficient. Urban centers are normally served by local bus or rail transit service. • Urban core zones represent the densest developed lands. The urban core is usually served by long distance buses and trains. Local bus or rail transit is the predominant mode of travel for workers and shoppers. For shorter trips in the area, the predominant mode of travel is walking. Assuming the ITE context zone definitions apply, the design framework as described here would include the selection or confirmation of the context zone or zones within which the project exists. In practical terms, this would be predetermined as part of the local or regional planning process conducted by either the owning agency or MPO. Context zone boundaries could be set using objective measures of land use definitions related to their propensity or relationship to generation of non-motorized traffic. Such boundaries should reflect planned future (i.e., formally approved plan) conditions. Stakeholders would confirm the context zone definitions for the proj- ect, or revise as necessary based on special conditions. 5.5.1.2 Road Type Road types generally mirror the functional classes currently used by AASHTO, with some additions. Road types describe the basic purposes of the road. The types of road users and manner in which service is provided vary with the context, which is defined in terms of land use and location. • Local Roads serve as access to adjoining properties; • Collectors serve as intermediate roadways linking local roads with arterials; • Arterials serve primarily to provide mobility to road users; • Freeways and other Controlled Access Facilities serve longer distance traffic, including specifi- cally freight movement; and

performance-Based highway Design process 81 • Special Purpose Roads serve unique, designated road functions or users. These may be transit- only corridors, resource recovery roads, and agricultural roads. Their applicability will clearly vary by land use and location context. Local roads and collectors will primarily be two-lane roads (serving both directions of travel). Arterials may be two-lane roads in rural context; but in urban contexts these will mostly be multilane roads. Terrain is not part of the framework. Terrain has historically been a surrogate for cost or difficulty of construction, and a simplifying descriptor for dimensional geometric criteria. The geometric design process envisioned employs direct measures of construction or implementation cost, thus eliminating the need for this indirect context descriptor. NCHRP Project 15-52 is a current active project titled “Developing a Context-Sensitive Functional Classification System for More Flexibility in Geometric Design.” The objective of this research is to identify potential improvements to the traditional functional classification system to better incorporate the context, user needs, and functions of the roadway facility. The potential improvements should lead to a flexible framework that can be used by planners and designers in the development of optimal geometric design solutions. As of the date of this project’s final report, NCHRP Project 15-52 was focusing on a functional classification approach very similar to that presented here. The AASHTO Green Book Task Force should be able to combine both efforts going forward with little difficulty. 5.5.1.3 Geometric Design Framework and Transportation Performance 5.5.1.3.1 Substantive Safety Performance and Land Use Context Zones. The importance of a more robust context zone definition is illustrated in Figure 27. The research team, with permission from the Illinois DOT, performed an analysis of the safety performance of roads in Cook County, Illinois (which includes the City of Chicago) using the context zone definitions in Figure 26 and data from Illinois DOT for the years 2007 to 2009. The downtown core of the city of Chicago was defined as Context Zone 6. Outside the downtown, the remainder of the city was considered Context Zone 4 or 5; and suburban Cook County was considered Context Zones 3, 4, and 5. Labeling of the context zones was based on the research team’s knowledge of the area and not a rigorous review per the ITE guidance. The analysis includes characterization of fatal and serious injury crashes (KA) by type of crash (single-vehicle or multivehicle) and by location (intersection versus road segment). For comparative purposes, data describing crash type characteristics for rural roads (Context Zones 1 and 2) were obtained from default values in the AASHTO HSM (Glennon 1987b). The figures show the proportion of crashes by type and the relative frequency of road segment versus intersections for the same context zones, indicated by the areas of the circles. Within Context Zone 6, the city core, the following is evident: • Of the 129 KA crashes, 78 (60 percent) occurred at intersections; • A total of 56 percent of intersection crashes and 37 percent of road segment crashes involved vulnerable users; • The next most prevalent serious crash type was multivehicle crashes (40 percent); and • Serious single-vehicle crashes were a small proportion of serious crashes. The distribution of crash types shows an intuitive trend for other context zones. For areas clas- sified as Context Zones 3 and 4, the proportion of vulnerable user crashes decreases to 16 percent, and the relative proportion of intersection versus segment crashes is closer to 50 percent. Finally, using HSM default statistics for rural context zones, the vulnerable user crashes are negligible, and single-vehicle crashes become more predominant.

Context Zones 1 & 2 -- Natural and General Rural Zones Context Zone 3 -- Suburban Zone Context Zone 4 -- General Urban Zone Context Zone 5 -- Urban Center Context Zone 6 -- Urban Core Characteristics of Fatal and Injury Crashes by Context Zone -- A Demonstration Ped/Bike 17% SV 34% MV 49% K and A-Injury Crashes for Road Segments (2007-2009) 2,659 Severe Crashes on 17,563.5 Lane-Miles Ped/Bike 16% SV 7% MV 77% 2,735 Severe Crashes on 47,008 Intersections K and A-Injury Crashes for Intersections (2007-2009) Ped/Bike 28% SV 33% MV 39% Roadway Segments in Chicago, Context Zones 4, 5 & 6 3,293 Severe Crashes on 8,666.5 Lane-Miles K and A-Injury Crashes for Segments (2007-2009) Ped/Bike 36% SV 5% MV 59% Intersections in Chicago, Context Zones 4, 5 & 6 K and A-Injury Crashes for Intersections (2007-2009) 3,132 Severe Crashes on 23,455 Intersections Ped/Bike 37% SV 22% MV 41% Roadway Segments in Downtown Chicago, Context Zone 6 K and A-Injury Crashes for Segments in Chicago, Urban Core (2007-2009) 51 Severe Crashes on ~61.75 Lane-Miles Ped/Bike 56% SV 5% MV 39% Intersections in Downtown Chicago, Context Zone 6 78 Severe Crashes on 154 Intersections K and A-Injury Crashes for Intersections in Chicago, Urban Core (2007-2009) Ped/Bike 0.2%SV 10.6% MV 89.2% Intersections, Context Zones 1 & 2, Highway Safety Manual*Ped/Bike 1.1% SV 62.7% MV 36.2% Roadway Segments, Context Zones 1 & 2, Highway Safety Manual* *Table 10-4. Default Distribution by Coll ision Type for Specific Crash Levels on Rural Two-Lane, Two-Way Roadway Segments (Total Fatal and Injury) *Table 10-6. Default Distribution for Collision Type and Manner of Collision at Rural, Two-Way, Four-Leg, Stop Controlled Intersections (Total Fatal and Injury) Roadway Segments in Cook County (Chicago excluded), Context Zones 3 & 4 Intersections in Cook County (Chicago Excluded), Context Zones 3 & 4 Figure 27. Profile of fatal and injury crashes by type for ITE context zones for Cook County, IL (MV = multivehicle and SV = single vehicle).

performance-Based highway Design process 83 This analysis of city of Chicago and Cook County data could be confirmed and further refined using data from other cities and more rigor or objective descriptors of zone definitions. However, it clearly demonstrates the importance and value of a more robust land use definition to apply to the geometric design control task. The substantive risk profile as it relates to road users and crash type throughout the road net- work is highly sensitive to context. Design policy and process in its formulation and application should be adapted to provide appropriate focus, design controls, dimensions, and approaches that address the unique and differing inherent risks associated with each type of roadway in each of the land use contexts. Traffic Operational Performance. Within Context Zones 1 and 2 the primary transportation modes are motor vehicles—both passenger cars and trucks. Context Zone 1 may include special vehicle types associated with the nature of the land use. These may include logging or other resource recovery trucks, recreational vehicles including towed carriers for boats and other equipment, and bicycles. Although pedestrians are not excluded, their presence in Context Zones 1 and 2 is typically of sufficient rarity that their presence does not create special roadway design demands. Context Zone 3 has sufficient development by type and density such that non-motorized road users are present with some regularity. As such, the design and operation of the roadway must consider their presence. In the suburban environment the primary operational concerns remain the movement of vehicles, but (1) transit buses may be present, (2) pedestrians crossing major intersections are a concern; and (3) bicycle traffic may be present, particularly where such routes serve or are adjacent to land uses such as schools and parks. With Context Zone 4 land use type and density progress to more urban forms. Pedestrian travel within the right-of-way should be expected, including the crossing of pedestrians at intersections. Intersection frequency may be greater, driveways more prevalent, and in some cases on-street parking common or necessary to support the adjacent land use. Within this context zone buses and trucks also operate, with many of the latter supporting retail and commercial development. Because of the prevalence of pedestrian activity, traffic operations in Context Zone 4 should emphasize lower speeds to minimize both the frequency and severity of vehicle/pedestrian conflicts. With Context Zones 5 and 6, the land use types and density are pure urban form. The street environment serves high volumes of pedestrians. Transit is prevalent and may be the most sig- nificant vehicular element. Trucks and freight vehicles are present to serve the commercial land uses, but their operations may be restricted to off-peak time periods. Within these context zones, road design issues typically include allocation of cross section and intersection design. Inter- section design and operations should favor the safe accommodation of pedestrian travel. This may mean limiting the use of turning lanes, provision for pedestrian-only signal phasing, and signal progression that produces very low speeds. Within these zones the concept of LOS for vehicles as classically understood does not apply. 5.5.1.4 Transportation Functions and Road Users in Context Zones Figure 28 describes the functional operational needs, values, and road user types associated with each cell in the basic framework. Functionality ranges from accessibility to mobility; with the latter further described by reliability. Road users for which the design should routinely serve range from one vehicle type (for special purpose roads), to motor vehicles only, to the full range of users including vulnerable road users. Based on the analysis in Figure 27 and other information contained in the AASHTO HSM, a safety performance framework is evident to guide the thought process and priorities of design engineers. This is shown in Figure 29.

84 a performance-Based highway Geometric Design process Roadway Type Rural Natural Zone Rural Zone SuburbanZone General Urban Zone Urban Center Zone Urban Core Zone Local Collector Arterial Freeway Accessibility to adjacent land uses with minimal cost and environmental disruption Access to land uses for motor vehicles and vulnerable users Access to land uses by pedestrians, transit users, and bicyclists; access for freight and goods delivery.Mobility and reliability of traffic service (travel ‚me and travel ‚me variance) for reasonable range of vehicle types Mobility for full range of road users including motor vehicles, bicycles, and pedestrians Travel ‚me reliability for transit buses and taxis; mobility for pedestrians Minimiza‚on and reliability of minimiza‚on of total costs of motor vehicle trips of all types (including especially freight) such costs to include both vehicle opera‚ng and travel ‚me costs Figure 28. Generalized profile of typical or critical operational issues governing geometric design by context. Roadway Type Rural Natural Zone Rural Zone Suburban Zone General Urban Zone Urban Center Zone Urban Core Zone Local Collector Arterial Freeway Mulvehicle intersecon and driveway-related; pedestrian and bicycle; low speed Mulvehicle intersecon and driveway-related; median and access related Single-vehicle run- off road; weaving, entering and exiting (interchange related) Mulvehicle weaving, entering and exing; congeson-related rear-end and sideswipe Single-vehicle run-off-road (low speed, low frequency) Single-vehicle run-off-road (high speed, higher frequency); multivehicle intersecon-related Single-vehicle run-off-road; truck involved; merging and exiting (interchanges); cross median Pedestrian -- intersecons and mid-block Pedestrian -- intersecons and mid-block; multivehicle intersecon-related Figure 29. Generalized profile of typical or critical substantive safety issues governing geometric design by context. Considering a land use context zone as part of the geometric design framework would serve to direct the appropriate approach to geometric design based on objective measures of safety performance associated with the full range of road users, and specifically vulnerable users. The selection of appropriate design or target speeds, intersection operating conditions, design vehicles, and design levels of service are influenced by the land use or context definition. Tables 10 through 15 show such guidance. For each roadway type, the design priorities, speed regime, and influence of the physical context are presented. Priorities are expressed in terms of access and mobility, and safety, with the latter related to the relative frequency of severe crashes.

performance-Based highway Design process 85 Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Generally Not Applicable Local Access to adjacent properties (employers, residents, delivery vehicles); Safety of pedestrians 20 to 50 mph Maintain character of the natural environment. Cross section (1 or 2 lanes only) and alignment to meet minimum operational requirements of critical design vehicles at very low speeds; 10- foot lanes and nominal shoulders may be used on lower-volume facilities Collector Access/mobility for motor vehicles; consider needs of recreational bicyclists but separate from travel lanes for motor vehicles 30 to 50 mph based on terrain Maintain character of the natural environment. Cross section (maximum 2 lanes) and alignment to meet operational requirements of typical vehicles. Design speed based on terrain (30 mph for mountainous to 50 mph for level) Arterial Mobility for motor vehicles passing through, including trucks; consider needs of recreational bicyclists but separate from travel lanes for motor vehicles 50 to 70 mph based on terrain Respect the character of the natural environment; 2 to 4 maximum lanes. Employ roundabouts and grade separations and avoid at-grade signalized intersections. Employ 11.5- foot to 12-foot lanes and full right shoulders; use open medians or barrier medians. Develop alignment that reinforces selected speed limit; design speed based on terrain (50 mph for mountainous to 70 mph for level) Freeway Mobility and reliability of mobility for vehicles passing through, including trucks 50 to 70 mph based on terrain Respect the character of the natural environment; employ 12-foot lanes and full shoulders; medians generally open and of sufficient width to not require barrier where possible. Develop alignment that reinforces speed limit and minimizes speed differentials of heavy vehicles; provide interchanges with spacing of 3 miles or more and avoid weaving sections Table 10. Geometric design guidance for roads and highways in rural natural zones. Table 11. Geometric design guidance for roads and highways in rural general zones. Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Generally Not Applicable Local Access to adjacent properties (employers, residents, delivery vehicles); Safety of pedestrians 20 to 50 mph Maintain character of the natural environment. Cross section (1 or 2 lanes only) and alignment to meet minimum operational requirements of critical design vehicles at very low speeds; 10- foot lanes and nominal shoulders may be used on lower-volume facilities Collector Access/mobility for motor vehicles; consider needs of recreational bicyclists but separate from travel lanes for motor vehicles 30 to 50 mph based on terrain Maintain character of the natural environment. Cross section (maximum 2 lanes) and alignment to meet operational requirements of typical vehicles. Design speed based on terrain (30 mph for mountainous to 50 mph for level) Arterial Mobility for motor vehicles passing through including trucks; consider needs of recreational bicyclists but separate from travel lanes for motor vehicles 50 to 70 mph based on terrain Respect the character of the natural environment; 2 to 4 maximum lanes. Employ roundabouts and grade separations and avoid at-grade signalized intersections. Employ 11.5- ft to 12-ft lanes and full right shoulders; use open medians or barrier medians. Develop alignment that reinforces selected speed limit; design speed based on terrain (50 mph for mountainous to 70 mph for level) (continued on next page)

86 a performance-Based highway Geometric Design process Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Generally Not Applicable Local Access for motor vehicles to designated locations; accessibility and mobility for pedestrians and bicyclists 20 to 35 mph Limit to 2 lanes; provide on-street parking per land use and local policies; employ minimum lane widths of 10 feet to 11 feet. Provide all-stop, 2-way stop or mini-roundabout intersections with minimal corner radii for speed control. Employ traffic calming near pedestrian-generating land uses such as schools, parks, places of worship, and libraries. Provide on-street parking on at least one side of the road Collector Access/mobility for motor vehicles; accessibility and mobility for pedestrians and bicyclists 30 to 40 mph Limit to 4 lanes; provide on-street parking per land use and local policies; employ signalized intersections with arterials. Provide width sufficient for bicycles (either shared-use or separate lane); provide for trees and streetscapes as desired with sufficient offsets from traveled way; employ medians for either landscaping or left-turn access based on land use Arterial Mobility for all motor vehicles; consider needs of bicyclists; provide for safety of pedestrians crossing at intersections 35 to 50 mph Employ lane widths of 11 feet to 12 feet based on speed, volume, and large vehicle presence; employ medians with width for left-turning vehicles and pedestrian refuge; prefer raised medians where possible given land use and provide median refuge for pedestrian crossings for 6 lanes or more. Discourage on-street parking (provide off-street to support commercial needs); avoid on-street parking along higher speed arterials. Design intersections for reasonably frequent vehicles (transit bus, delivery trucks); signalize to provide progression at desired speeds; provide adequate timing for pedestrian crossing and consider pedestrian lead signals for right-turn conflicts; programmatically prohibit right turn on red; design intersections to enable landscaping without conflicting with sight lines Freeway Mobility and reliability of mobility for vehicles passing through, including trucks 60 to 70 mph Employ lane widths of 11.5 feet to 12 feet; provide full shoulders for incident management. Minimize interchanges to 1.5 mile spacing; avoid weaving sections requiring CD or ramp braids. Favor depressed versus raised cross section for noise and visual effects; Employ efficient service interchange configurations [partial cloverleafs (parclos), diverging diamond interchange (DDI), diamonds]; adapt interchange design to facilitate pedestrian activity along crossroad; employ traveler information ITS and integrate with adjacent arterial network Table 12. Geometric design guidance for roads and highways in suburban zones. Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Freeway Mobility and reliability of mobility for vehicles passing through including trucks 50 to 70 mph based on terrain Respect the character of the natural environment; employ 12-foot lanes and full shoulders; medians generally open and of sufficient width to not require barrier where possible. Develop alignment that reinforces speed limit and minimizes speed differentials of heavy vehicles; provide interchanges with spacing of 3 miles or more and avoid weaving sections Table 11. (Continued).

performance-Based highway Design process 87 Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Mobility and access for special vehicle types (transit buses, pedestrians, bicyclists) 20 to 30 mph Special purpose defines spatial needs (transit, pedestrian mall, etc.); maintain pedestrian accessibility to adjacent land uses Local Access for motor vehicle to designated locations; accessibility and mobility for pedestrians 20 to 35 mph Limit to 2 lanes with 10 foot widths sufficient; provide on-street parking per land use and local policies; employ minimum lane widths. Provide all-stop, 2- way stop, or mini-roundabout intersections with minimal corner radii for speed control. Employ traffic calming near pedestrian- generating land uses such as schools, parks, places of worship, and libraries Collector Access/mobility for motor vehicles; accessibility and mobility for pedestrians and potentially bicyclists 25 to 35 mph Limit to four lanes; provide on-street parking per land use and local policies; employ signalized intersections with arterials. Minimize lane widths—11 foot maximum; consider provision for bicycles within traveled way; provide for trees and streetscapes as desired with sufficient offsets from traveled way Arterial Mobility for all motor vehicles; consider needs of bicyclists; provide for safety of pedestrians crossing at intersections and potentially mid-block 25 to 35 mph Minimize footprint using 11-foot lanes; employ medians with width for left-turning vehicles and pedestrian refuge. Design intersections for reasonably frequent vehicles (transit bus, delivery trucks); signalize to provide progression at lower speeds; provide adequate timing for pedestrian crossing and consider pedestrian lead signals for right-turn conflicts; consider prohibition of right turn on red; employ space and design intersections to enable landscaping without conflicting with sight lines Freeway Mobility and reliability of mobility for vehicles passing through including trucks 50 to 60 mph Favor depressed versus raised cross section for noise and visual effects; consider lane widths less than 12 feet and minimal shoulders to maximize capacity within limited right-of-way; minimize interchanges to avoid weaving sections; use tight diamond and similar configurations and low design speeds for ramps. Generally design for LOS E Table 13. Geometric design guidance for roads and highways in urban general zones.

88 a performance-Based highway Geometric Design process Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Mobility and access for special vehicle types (transit buses, pedestrians, bicyclists, truck deliveries) 10 to 25 mph Special purpose defines spatial needs (transit, pedestrian mall, etc.); maintain pedestrian accessibility to adjacent land uses Local Access for motor vehicle to designated locations; accessibility and mobility for pedestrians 20 to 30 mph Limit to 2 lanes; provide on-street parking per land use and local policies; employ minimum lane widths. Provide all-stop, 2- way stop intersections with minimal corner radii for speed control. Employ traffic calming near pedestrian-generating land uses such as schools, parks, places of worship, and libraries Collector Access/mobility for motor vehicles; accessibility and mobility for pedestrians and potentially bicyclists 20 to 35 mph Limit to 2 lanes; provide on-street parking per land use and local policies; employ signalized intersections with arterials. Minimize lane widths; consider provision for bicycles within traveled way; provide for trees and streetscapes as desired with sufficient offsets from traveled way Arterial Mobility for all motor vehicles; consider needs of bicyclists; provide for safety of pedestrians crossing at intersections and potentially mid-block 25 to 35 mph Minimize footprint using 10-foot to 11- foot lanes; employ medians with width for left-turning vehicles and pedestrian refuge. Design intersections for reasonably frequent larger vehicles (transit bus, delivery trucks); signalize to provide progression at lower speeds; provide adequate timing for pedestrian crossing and consider pedestrian lead signals for right- turn conflicts; programmatically prohibit right turn on red; employ space and design intersections to enable landscaping without conflicting with sight lines Freeway Mobility and reliability of mobility for vehicles 40 to 50 mph Favor depressed versus raised cross section for noise and visual effects; consider lane passing through including trucks widths less than 12 feet and minimal shoulders to maximize capacity within limited right-of-way; minimize interchanges within urban core to avoid weaving sections; use tight diamond and similar configurations and low design speeds for ramps. Generally design for LOS E Table 14. Geometric design guidance for roads and highways in urban center zones.

performance-Based highway Design process 89 Transportation Design Priorities Typical Speed Regime Context Influence on the Approach to Roadway Design Special Purpose Mobility and access for special vehicle types (transit buses, pedestrians) 10 to 25 mph Special purpose defines spatial needs (transit, pedestrian mall, etc.); maintain pedestrian accessibility to adjacent land uses Local Access for motor vehicles to designated locations including delivery vehicles; accessibility and mobility for pedestrians 20 to 30 mph Limit to 2 lanes with 10-foot lanes; provide width for curb parking, loading zones and bus stops. Provide all-stop, or 2-way stop intersections with minimal corner radii for speed control Collector Accessibility and mobility for pedestrians and potentially bicyclists; access/mobility for motor vehicles 20 to 35 mph Limit to 2 lanes; provide width for curb parking, loading zones, and bus stops; employ signalized intersections with arterials. Employ 10-foot lane widths; provide for trees and streetscapes as desired with sufficient offsets from traveled way Arterial Provide for safety of pedestrians crossing at intersections and potentially mid-block; mobility for all motor vehicles 20 to 35 mph Minimize footprint using 10-foot or 11-foot lanes; employ medians with width for left-turning vehicles and pedestrian refuge. Design intersections for reasonably frequent vehicles (transit bus, delivery trucks); signalize to provide progression at lower speeds; provide adequate timing for pedestrian crossing and consider pedestrian lead signals for right-turn conflicts; programmatically prohibit right turn on red; employ space and design intersections to enable landscaping without conflicting with sight lines Freeway 45 to 55 mphMobility and reliability of mobility for vehicles passing through including trucks Favor depressed versus raised cross section for noise and visual effects; consider lane widths less than 12 feet and minimal shoulders to maximize capacity within limited right-of-way; minimize interchanges within urban core to avoid weaving sections; use tight diamond and similar configurations and low design speeds for ramps. Generally design for LOS E Table 15. Geometric design guidance for roads and highways in urban core zones. 5.5.1.5 Other Land Use Considerations Another aspect of the land use context is the consideration of specific properties adjacent to, in the vicinity of, or affected by the project. These would include public health and safety sites (hospitals, fire stations, and police stations), schools, public parks, playgrounds and recreational facilities, and places of worship. These may influence the type of design solution, the dimensions of the road, or special operating requirements to be incorporated into the design. For many of these special land uses, pedestrian accessibility is a key feature. The geometric design process, particularly as it relates to development of alternatives and consideration of road user impacts during construction, should include the explicit compilation and analysis of such land uses. For 3R and reconstruction projects, the accessibility to important public facilities and vice versa will often be critical considerations in maintenance of traffic plan. The geometric design process should include a process step that informs designers of critical land uses. This step should be required regardless of the environmental process applicable to the project. Every project

90 a performance-Based highway Geometric Design process should identify, acknowledge, and address changes in access, travel time, or location of public health and safety facilities either permanently or during construction periods. Protected lands such as environmentally sensitive wetlands, forests, cemeteries, historically significant properties, and unique local features should be noted and identified as external con- trols. Regulatory stakeholders and experts in each environmental field are typically the source of information. Information should be of sufficient detail and quality to provide meaningful direction to the designer on the criticality of avoidance versus mitigation per impact. Roadway designers should be fully briefed on the nature and criticality of such information prior to initiating the geometric design process. Criticality relates to the type of land use, its quality, the applicable regulations, and the input of the stakeholders responsible for oversight and approvals or permits. Geometric designers also should understand the agency’s philosophy and approach to each type of feature or control. Avoidance potentially increases adverse impacts on lands not designated as special in nature. The performance of the road may be adversely affected by pure avoidance approaches (but such an alternative may be necessary to develop and evaluate). Impact and mitigation minimization also are approaches to dealing with important controls. Some important controls may not be specifically protected by environmental processes, but can be critical to understand. Examples include drainage tiles and flow patterns on agricultural lands, and holders of farm leases (versus ownership) that relate to the patterns of access between properties and their economic viability. For new alignment or reconstruction projects of any sizable length, there will virtually always be some significant properties or land uses for which the geometric design may be adjusted in three dimensions to avoid or minimize an adverse effect. A central aspect of the geometric design process is the comparison of design solutions that may differ in their effects on adjacent lands, their expected transportation performance, and implementation cost. 5.5.2 Develop Project Evaluation Criteria Within the Context Framework Every project regardless of type, size, and problem(s) addressed should be approached from the perspective that there is more than one reasonable design solution. The more complex the problem and the more stakeholders involved, the greater the importance of considering multiple alternatives. Moreover, the more complex the project and number of stakeholders, the greater the chances that adverse effects will be apparent with each design alternative. Regardless of project type and context, every road design project requires the designer to assess and manage trade-offs among important variables of interest to stakeholders and the owning agency. A key aspect of the design framework is the set of stated priorities and preferences that will form the basis for selecting a preferred design among many alternatives. This takes the form of project evaluation criteria. The evaluation criteria express the performance and impact trade-offs involved. Given the location, land use context, project type, and problem(s) being addressed, such criteria should be tailored to the project. The following shows typical trade-offs from which performance-based evaluation criteria may be developed to direct decision making. • Trees along the roadside to provide shade and improve aesthetics vs. removing all potential roadside hazards. • Inclusion of a left-turn lane at a signalized intersection vs. avoid right-of-way taking and/or conflicts with adjacent land uses.

performance-Based highway Design process 91 • Open median or continuous two-way left-turn lane to provide access to properties between intersections vs. raised median to control access. • Shoulder use on freeways to increase capacity and enhance operations vs. maintenance and emergency use, drainage. • Right turn on red (creating conflicts with pedestrians) vs. reduce delay for right-turning vehicles at signalized intersections. • Shoulders on two-lane highways with rumble strips to warn drivers vs. without rumble strips to facilitate shoulder use by bicyclists and/or to avoid noise impacts from shoulder encroachments. • Use of permissive only signal phasing vs. protected signal phasing for left-turn lanes at signal- ized intersections. • 3:1 side slopes vs. 4:1 or flatter to reduce construction costs. Evaluation criteria are categorized in one of the following areas: • Construction or implementation cost and time (including relevant M&O costs); • Right-of-way (separate from cost of right-of-way; may include amount, type, and relocations); • Traffic operations; • Public safety; and • Environmental effects. Evaluation criteria can include prohibitions or “fatal flaw” attributes or can include specific requirements. These may be associated with requirements of regulatory agencies (including specifically environmental agencies) or significant preferences of key stakeholders involved with the project. The criteria should be stated in quantitative terms relevant to the project size and scope and problems being addressed. Appendix A is a performance criteria memorandum devel- oped for a project involving the replacement of a river bridge. The criteria specify the threshold values, means of measuring, and relative importance of each. Agency assurances that a particular stakeholder preference will be addressed or accommo- dated are policy decisions. The agency bears the responsibility of defending or explaining their assurances, and demonstrating they understand the trade-offs or impacts borne as a result of the accommodation. In the interests of transparency, agencies should reveal such assurances and provide relevant data or substantive reasons for such assurances that demonstrate their relative value as part of the overall evaluation criteria. Indeed, in some cases meeting a stakeholder con- cern may be part of the core defined transportation problem (e.g., maintaining full accessibility to property x). It has historically been common practice for evaluation criteria to be developed by agency design engineers after the initial development of design alternatives. The performance-based geometric design process calls for this step to be conducted prior to design project development. This is for two important reasons. First, awareness of the criteria serves to direct the work flows, data and analysis processes, and level of detail in the alternatives development stage. The criteria also provide focus to the designers by informing them of what constitutes success. Second, the design team avoids the potential perception of bias from external stakeholders, which may occur if evaluation criteria are developed after alternatives development. For complex design projects involving multiple trade-offs, it is pointless to begin design without first arriving at a common point of reference or consensus around what constitutes an optimal design solution. An optimal design solution implies a solution consistent with the priorities established for and constraints present within the project. The evaluation criteria should be objectively stated, uniquely defined, and tailored to the project. Qualitative or subjective measures should be avoided in favor of specific, numeric, or objective measures. They may include criteria that relate to just one portion or segment of the

92 a performance-Based highway Geometric Design process project (e.g., an intersection or curve). The importance of the evaluation criteria should be expressed (i.e., what are the most important factors driving the final decision). The process where criteria may be weighted and formal processes for calculating an optimal project score established is known as Multiattribute Utility Analysis and is used by some agencies to inform decision making (AASHTO 2004). Relatively small or straightforward projects will involve design choices and trade-offs. A 3R proj- ect that involves addressing a pavement condition problem may include analysis of different pavement types and pavement designs, each of which may produce differences in vertical alignment and drainage, length of time to construct, useful service life, and total initial construction costs. Depending on the project, the criteria and weighting should be developed through consultation with stakeholders, typically in one or more workshop settings. As a minimum, the criteria should be shared with stakeholders so they understand the basis on which the decision will be reached. 5.5.3 Establish Decision-making Roles and Responsibilities Stakeholders and project staff need to understand who is making the significant decisions, including the selection of the preferred geometric design. Confirmation and communication of the decision maker(s) should be specific, and not just within the DOT as there are many types of projects that involve different levels within the organization. For some projects the agency’s project manager or lead engineer may make the decision; for others the local or district engineer may make the ultimate decision; or the decision may rest with the agency’s senior management. Highway engineering staff responsible for the geometric design of alternatives may not be the final decision makers on projects, particularly more complex ones. Engineering staff must be fully aware of the constraints, values, and the evaluation criteria that define success. They should fully understand the policies of their agency, promises to external stakeholders, and the agency’s overall program and resources. Agency leadership will rely on the technical skills and creativity of the highway engineering staff to develop the solutions that best meet project needs and respect the community values. This knowledge and understanding should be reached prior to initiating project alternative designs. External stakeholders may serve in a consultation or advisory role. Some project decisions may involve the stakeholders directly. Project issues that affect the operation or performance of another agency’s infrastructure should include that agency. Everyone’s role and responsibility should be confirmed at the project outset. 5.5.4 Determine Basic Geometric Design Controls— Design or Target Speed Geometric design controls involve those technical inputs that will directly or indirectly con- trol the type and nature of the design. Many of these controls will be directed by the context (described above) and project type. However, an effective geometric design process is one that provides designers with choices in the controls to be used. The importance and necessity of speed as a direct input to roadway design is confirmed. The geometric design process will require speed input as a means of developing design criteria, select- ing appropriate design features, and establishing traffic control. For contexts in which design speed applies, the selection of an appropriate design speed is a control that a designer should have a range of speeds from which to select. The original concept of design speed, dating to the 1940s, was that it was independently developed by the designer. The legal posted speed, which historically was set by traffic engineers,

performance-Based highway Design process 93 was typically associated with an 85th percentile observed free speed. It has become common practice now for posted speeds to be set by governmental policy or even legislation, outside of an engineering analysis of speed behavior or roadway context. Moreover, an outcome of the older AASHTO design policies for horizontal alignment is that the original relationship between the 85th percentile speed and design speed no longer holds true irrespective of posted speed limits. Given AASHTO’s definition of design speed there is no substantive reason why design and posted speed need to be linked. Indeed, under the current design process and methods for deter- mining alignment and sight distance, the linking of the two creates conditions in which the road design encourages higher speeds than are desired or appropriate (Tables 10–15). However, the legacy of the term and its direct linkage by policy of FHWA and many DOTs may preclude changes to such policies. ITE adopted the term target speed in its Designing Walkable Urban Thoroughfares: A Context Sensitive Approach—ITE Recommended Practice (Institute of Transportation Engineers 2010). This term may be useful in expressing the desired speed outcome of a geometric design solution for which a speed metric is needed to determine a physical dimension. The geometric design process could utilize available tools to predict speeds, and the body of literature on speed effects of treatments or designs (e.g., traffic calming devices or narrowing of roads), to test and refine a design to produce the desired target speed for the specific context. It is clear that some measure or uses of speed in the development and application of design criteria should remain within a revised design process. However, what is needed is a more nuanced approach to speed that is sensitive to road type and context. A revised design process should emphasize and promote the following: • Speeds appropriate for both driver expectations and the context, with the latter defining a full range of speeds; • The de-linking of design speed with quality, which translates to how to provide high-quality design approaches in the urban environment for which lower speeds are appropriate; • The inclusion of measures of speed related to crash risk, including speed differentials and speed changes; and • Full consideration of terrain and land use (i.e., selecting a practical design speed). Design policy should avoid communicating the concept that design speed is a surrogate for design quality and that promotion of higher speeds regardless of context will often produce undesirable performance (not to mention higher cost). In urban contexts where pedestrians are prevalent, designing a road to meet the desires of drivers to operate at higher speeds increases vulnerable user crash severity risk. Another design policy challenge is the institutionalized relationship between the design speed and posted speed. As originally developed, the concept of design speed was selected based on the AASHTO context (urban/rural; terrain); and the geometric design would produce driver behavior consistent with the design speed and codified through speed studies in setting posted speed limits. Over the past 40 years, both the outdated nature of alignment policy model assumptions and the injection of public policy into setting of artificially low speed limits has made the initial concept of design speed unworkable, to the extent that common practice is now to set a design speed based on what the current law or policy says what the posted speed limit is or should be. Finally, the concept of design speed as originally developed was rural centric. The alignment and cross section of roads in urban and suburban areas has little influence (except in unusual, extreme cases) on speed behavior. Rather, it is the presence, frequency, and spacing of intersections, and in particular signalized intersections, that influence speed and speed behavior.

94 a performance-Based highway Geometric Design process 5.5.4.1 Design or Target Speed and Road Geometry Speed currently directly influences horizontal alignment design, vertical curvature (through application of SSD, which is based on speed), intersection sight distance (ISD), and roadside design. 5.5.4.1.1 Design or Target Speed Regime—Urban Nonfreeway Roads. Roads and streets in urban contexts should be designed and operated to serve multimodal traffic within the right- of-way. A primary safety performance concern in all urban land contexts is the exposure of pedestrians to crashes. The second most important safety concern is the crossing conflict associated with intersections, operated in a variety of ways. A design process with a basis of overall safety performance should be centered on the develop- ment and application of design criteria in a moderate to low-speed environment. The general urban, urban center, and urban core zones are characterized by fully developed land use, and six to 10 intersections occurring per mile, with intermittent driveways in some cases. The following is evident: • The concept of SSD (stopping to avoid an object in the road) is functionally irrelevant in this environment. Sightline-related conflict avoidance primarily concerns crossing conflicts at intersections and driveways, such conflicts to include pedestrians and bicyclists. Drivers essentially navigate from one intersection to the next. For residential streets and those adjoining parks and similar spaces, consideration should be given to providing SSD for pedestrians in the street. • Posted speed limits will usually be at most 40 mph, and often 30 to 35 mph, for nonfreeway road types. • The concept of providing for clear zones has limited application in roadside design for urban roads. Consequently, there have been changes in the 2011 AASHTO policy. Right-of-way outside the edge of the pavement is limited, often to less than 10 feet. Curbs create the poten- tial for vaulting and loss of control when struck. Fixed objects such as roadside furniture, light and utility poles, and trees will be within 5 feet, and pedestrians will often travel within 5 feet of the edge of pavement. Despite these conditions, the relative risk of serious urban single- vehicle crashes is much less than other crash types precisely because of the lower prevailing speeds (Figure 26). • There is generally insufficient space to allow for transition from roadway elevations and adjacent driveways and walkways, making it important for the vertical alignment of the road to match the adjacent terrain. • The concept of providing an alignment that is comfortable to travel at some speed when unimpeded by traffic does not apply. • Visual distractions to drivers from signs, lights, and activities are typically numerous and continuous. For the urban environment, a CSD process that incorporates speed in relevant ways should focus on sightline conflicts and sight distance. The following is suggested: • By policy, all nonfreeway roads within each defined urban context zones would be designed to a single or small range of speed. The selected speed would be sufficiently high to produce appropriate calculations of sight lines for all forms of ISD. • The only application of this speed on the road geometry would be in determination of sight distances, which would be ISD related. Design controls for horizontal and vertical alignment would be based on vehicle operational limitations independent of speed and provision for drainage. Table 16 shows an example set of design or target speed guidelines applicable to urban context zones.

performance-Based highway Design process 95 5.5.4.1.2 Design or Target Speed Regime—Rural Nonfreeway Roads. The use of a speed metric for rural roads appears necessary. The relationships between crash severity and speed, and between user operating costs and speed, both support speed as a vital metric. How speed is actually applied to alignment and sight distance design may differ from current policy. 5.5.5 Determine Basic Design Controls—Design Traffic Volumes The design process being inherently forward looking, all projects are designed based on a projection or forecast of future traffic. Design year traffic directly influences pavement design, sizing of the roadway and intersections, and may influence roadway geometry. Design year traffic forecasts should be developed and applied in a manner consistent with the current best practices and methods for safety and operational analysis. Under current practices this means that as a minimum AADT, design hour volume (DHV), vehicle classification (T percent), and turning movement count and DHV for intersections. For many higher-volume roads in congested contexts, additional basic data may be necessary to enable complete evaluation of a geometric design. Peak-hour traffic has historically served as the basis for sizing decisions. Traffic analyses to support design decisions or differentiate among alternatives have focused on operations during the peak hour or hours. In most major cities, however, it has long been impractical, and is now considered poor public policy, to even attempt to design to meet the demand represented by peak-hour traffic. As designing for LOS E becomes commonplace, additional traffic parameters and metrics or approaches have become necessary to fully understand and differentiate among alternatives. The geometric design process relies on transportation planning subprocesses that forecast vehicle traffic volumes based on future land use, road network characteristics, and research on human behavior and response to travel choices. Forecasts reflect the socioeconomics and major policies of the jurisdiction in the future. Weekday traffic distribution data describing the variation in demand across a 24-hour day are a necessary basic traffic volume input to design. Many MPOs no longer forecast average daily traffic, but rather peak-hour traffic, or even traffic throughout a typical 24-hour weekday. For example, Figure 30 shows the modeling approach used by the Metropolitan Planning Agency for Chicago, which involves the forecasting of demand volumes for eight time periods within a typical weekday. The development of traffic forecasts of multiple time periods for design purposes, or for operational analyses in alternatives consideration, may be appropriate for certain project types or contexts. General Urban Zone Urban Center Zone Urban Core Zone Local 35 35 35 Collector 35 35 35 Arterial 40 40 40 *Speeds for design of intersecon sight distance only Non-freeway Urban Context Speed Regime* Table 16. Target speeds for urban context zones by highway type.

96 a performance-Based highway Geometric Design process Transportation agency planning policies will generally establish both a timeframe and basis for design year traffic, with certain parameters not subject to judgment or choice. These would include average daily traffic and percentages of heavy vehicles, which are used in pavement design. Other parameters may involve a technical or policy decision by the designer or agency. Certain roadway features are designed based on hourly traffic volumes, with the DHV selected to represent a critical threshold for design purposes. This number, expressed relative to the annual distribution of hourly volumes, may be influenced by historic count data, but the selection of the hourly basis is a choice that should reflect the road type and context. For example, an hourly volume representing the 30th, 50th, 100th, or 200th highest volume for a year may be the basis for design. Designers may select and test more than one hourly volume, particularly in urban areas where design for congestion is necessary, or for roadways serving highly peaked or recreational traffic. The forecast data become direct inputs to fundamental geometric design decisions as illustrated below: • Basic number of through lanes reflects design year average daily traffic; • Intersection traffic control reflects peak-period approach and turning traffic; • Intersection channelization (number of left- and right-turn lanes) reflects peak-period approach and turning traffic; • Intersection channelization (lengths of turn lanes) reflects approach volumes and signal-cycle lengths that are related to speed, phasing, and other operational parameters; • Warrants for grade separations or interchanges reflect crossing average daily traffic; • Roadside design criteria and policies are based on average daily traffic; • Minimum cross section dimensions may be based on average daily traffic; and • Crosswalk and pedestrian signing treatments reflect average daily traffic. Forecast traffic volumes also form the basis for estimating quantitative benefits of design solutions. Operating costs, travel time, and crash costs are directly related to the volume of 8:00 PM - 6:00 AM 6:00 AM - 7:00 AM 7:00 AM - 9:00 AM 9:00 AM - 10:00 AM 10:00 AM - 2:00 PM 2:00 PM - 4:00 PM 4:00 PM - 6:00 PM 6:00 PM - 8:00 PM 0 2 4 6 8 10 12 Ho ur s Time Periods CMAP Time-of-Day Traffic Assignment Periods Figure 30. Traffic assignment periods employed by the Chicago Metropolitan Agency of Planning (CMAP).

performance-Based highway Design process 97 vehicles (and persons) using the roadway. Evaluation methods and models may employ hourly volumes or daily volumes. The HSM methodologies are currently based on AADT values; with additional data necessary such as the number of hours per day that lane volume exceeds 1,000 vehicles per hour for freeway safety predictions. Use of such tools requires 24-hour week- day distribution data. To summarize, a revised geometric design process requires that sufficient detail and specificity be developed in forecast traffic, consistent with the operational analysis methods representing best practices. 5.5.5.1 A New Parameter—Service Life Traffic A historic anomaly of the geometric design process is the disconnect between the nominal forecast design year (typically 30 years or less) and the expected useful physical life of roadway infrastructure. Right-of-way, road grading, storm sewer, and structural elements have practical service lives of 75 years or more. A current trend in pavement design is for longer-lasting pavement—up to 50 years or more. Major river crossings such as tunnels or long span bridges are designed assuming at least 100 years of life. These much longer functional lives conflict with the much shorter timeframes for which design year traffic is typically developed. To summarize, the sizing and geometric design of roads is commonly based on a level of traffic (and hence intended performance) far short of that compared to the expected life of the physical infrastructure. The suggested geometric design process incorporates the full life cycle of a project as defined by the functional performance of the project’s infrastructure for as far as is reasonable to estimate. Roadway infrastructure provides measurable benefits and requires the incurring of real costs, well beyond a 20 to 30 year nominal design life. A new service life traffic design control is proposed to enable the estimation of these benefits and costs beyond the nominal design year as part of the geometric design process. Service life traffic volume would be an annual traffic volume considered appropriate for analysis purposes to represent the operations of the roadway at a time period consistent with the intended physical service life of the roadway infrastructure. The application of service life traffic is: • Based on the project type, designers (or agencies by policy) would assign a service life to the project. Table 17 presents a set of guidelines that would reflect the nature of the construction and context. Longer service lives are associated with newly acquired right-of-way, new bridges, and new major infrastructure such as tunnels or major river crossings. Shorter service lives may be associated with bridge repair, replacement of pavement (reconstruction), and projects that do not include new right-of-way or utility relocations. Figure 31 demonstrates the relation- ship between the typical agency costs and the transportation benefits through the service life of the facility. The transportation benefits increase over the years as the traffic volume served Project Type Recommended Guidance for Service Life of Roadway Infrastructure 3r (Pavement Resurfacing) 20 to 30 years Roadway Reconstruction 75 to 100 years New Alignment Roadway 75 to 100 years Bridges, Walls, and Related Infrastructure 50 to 75 years Major Watercourse Crossings 100 years Table 17. Guidelines for assignment of service life for roadway infrastructure projects.

98 a performance-Based highway Geometric Design process by the facility increases. The agency costs for maintenance increase as the facility gets older. Periodic costs for rehabilitation will be incurred over the life of the facility. • Service life traffic volumes would be developed to be used in cost-benefit analyses. Such volumes would not be forecast in the traditional sense, as the premise in service life volume is that the applicable timeframe is beyond that for which travel forecasts can be reliably made. Rather, the traffic volumes would be set by agency policy to reflect an operating set of assumptions considered reasonable by the agency for the purpose of calculating or estimating societal travel benefits. Depending on the project and context, a number of reasonable approaches may be taken. These may include (1) holding design year traffic constant from the design year to the end of service life, (2) extrapolating traffic growth at the same rate as the ratio of design year to current year traffic implications, and (3) extrapolating traffic growth at some rate between no growth and that growth implied by the ratio of design year to initial year traffic. • Calculation of benefits (safety, traffic operations, time, and delay) and costs (O&M) sensitive to traffic volume would be made assuming the service life traffic volumes beyond the design year to the service life year. • A terminal life also could be assumed at the end of the service life. This may include the value of the right-of-way and road grading, and some portion of major bridge infrastructure. Careful thought and judgment are necessary to develop reasonable service life traffic volumes. Some additional thinking and research may be appropriate. Yet, the notion that highway projects provide meaningful benefits and produce ongoing costs beyond the nominal design year is correct. A geometric design process reliant on principles of cost effectiveness and financial sus- tainability must include consideration of these benefits and costs. The relative importance of capturing costs and benefits beyond the design life will depend on the discount or interest rate used in the analysis. In financial environments in which low interest rates prevail, calculated cash flow benefits and costs will be meaningful for 50 years or more. Current U.S. Government Office of Management and Budget guidance for federal infrastructure cost effectiveness analysis using constant dollar (no inflation) values calls for a discount rate of 1.9 percent for 30 years (U.S. Government Office of Management and Budget, nd). Typical discount rates in the range of 2 to 4 percent will adjust for the inherent uncertainty associated with forecasts beyond 30 years. Figure 31. Agency costs—transportation benefits.

performance-Based highway Design process 99 5.5.6 Determine Basic Design Controls—Design LOS (or Operating Condition) Current design practice applies the concept of levels of service to the roadway design. Such levels of service are defined and their basis for application described in the 2010 HCM (TRB 2010). Levels of services range from A to F. The definition of LOS and its calculation varies with the type of road and, in some cases, feature of the road (Figure 32). Under current design practice a roadway is sized and designed to provide the selected LOS using the appropriate methodologies per the HCM (for road segment type, intersection, ramp terminal, roundabout). Source: HCM Exhibit 2-2 and 2-3 Figure 32. Level-of-service definitions for road types.

100 a performance-Based highway Geometric Design process LOS is a design control that is the choice of the designer or agency. Current design policy in the AASHTO Green Book provides guidance, but the text and intent of the policy is for this to be a choice. An effective design process requires methods that use meaningful and understandable measures of transportation performance. The historic design process uses design hour traffic and LOS merely for sizing purposes. Also, it may be used in other evaluations such as air quality and noise calculations. Design hour traffic also is used to develop performance measures for delay, fuel consumption, and other metrics using microsimulation programs such as VISSIM and CORSIM. The concept of reliability has been developed and is now included in the HCM as an addi- tional attribute or control in traffic operations. Chapters 36 and 37 of the HCM present methods that can be used to describe how often particular operational conditions occur and how bad conditions can get. The framework proposed here can incorporate reliability measures to the extent that they can be translated into operational-performance measures (travel time, operating cost, and crash costs). 5.5.7 Determine Basic Design Controls—Road User Attributes Multimodal road users include all motor vehicle types (passenger cars, buses, trucks, and special vehicles such as emergency service, agricultural, or resource recovery), bicyclists, and pedestrians. 5.5.7.1 Design Motor Vehicles The design vehicle is that vehicle or vehicles that may be legally operated on the roadway. They have physical and/or operating characteristics more critical than other vehicles, and should be used to control the design of the relevant geometric element. Characteristics include: • Width (feet)—lane width, shoulder width; • Turning radii and offtracking (feet)—corner radius, offset, width of turning roadway; • Center of gravity (feet)—propensity to overturn at a combination of speed and curvature; • Height of eye (feet)—sight distance; • Acceleration (feet/second/second)—length of speed change lanes; • Gradability (wt/hp)—maximum speed on grade by length; and • Length (feet)—queue lengths. 5.5.7.2 Design Driver AASHTO policy has developed and applied the concept of a “design driver” for many years. In its traditional form, the design driver is one considered to be of sufficient minimal capabilities to legally operate a motor vehicle. The design driver is sober (not under the influence of drugs or alcohol). The design driver per AASHTO is one with a 95th percentile ability to react to external stimuli by applying the brake and/or steering to avoid a conflict. The design driver has suffi- cient capabilities to see and perceive the roadway and its environs. Human factors research has historically been applied to establish basic driver inputs to design, e.g., perception and reaction time, unsignalized intersection gap acceptance, and navigational or decision time. In addition to the above basic design driver attributes, there are other relevant attributes that describe the range of driver response to roadway situations. These may include driver response to traffic signal changes (“yellow deceleration rate”), speed selection, car-following and headway acceptance, merging gap acceptance and acceleration selection, and passing behavior. Some of this behavior is regional in nature, and some demographic (e.g., elderly drivers).

performance-Based highway Design process 101 Both the knowledge base and traffic operational technology now in practice allow for the highway designer to adjust a design, or to evaluate the operation of a design based on a range of potential driver behavior scenarios. Agency design standards may be based on less aggressive or more “conservative” driver assumptions, but different driver profiles in practice may produce different operations. The software currently available for the evaluation and simulation of traffic operations enables selection of various driver types. One commonly used proprietary software package allows selection of as many as 10 different driver types based on 14 input parameters. 5.5.7.3 The Design Bicycle In some contexts and road types, design for bicyclists is a potential design control. The operational requirements for bicycles would control the geometric design of bike paths. Where bicycles operate within the roadway used by motor vehicles, three elements may be influenced by the operation of bicycles: • Width (feet)—shared lane width, bicycle lane width, • Gradability (combination of grade and length traversable by cyclist), and • Physical separation and/or barrier to provide protection from higher speed vehicle traffic. 5.5.7.4 The Design Pedestrian For certain contexts, the operating needs of pedestrians may control the geometric design of roadway elements. The relevant characteristic is walking speed (feet per second). Walking speed translates to the minimum crossing time at a signalized intersection. Elderly and disabled pedestrians travel at slower speeds, and pedestrian walk times in inclement weather are slower. Designers would select a walking speed compatible with the demographic characteristics of the area and climate. An updated design process also should include the identification of where and how the pres- ence of pedestrians is of sufficient importance to directly affect the geometry or operation of the road. This may be set by agency policies, for example, linked to land use context, based on counts or estimates of pedestrians, and/or proximity to specific land uses. Pedestrian presence as a design control may influence selection of the design vehicle and design radii for intersections, islands, channelization, median widths, a design assumption of no right turn on red, pedestrian phasing, and signal-cycle assumptions. 5.5.7.5 Non-motorized Traffic Operational and Design Controls In some land use contexts, the presence of pedestrians and bicyclists becomes an explicit design control. The geometric design process may require counts or estimates of bicycle volume, pedestrian volume, and pedestrian walking speeds. The presence of transit routes, designated stops, and arrival headways also may be controls. For urban core context zones or their equivalent, explicit operating controls for intersection design may be considered either as agency policy or on a project-specific basis, with such controls to influence the road design. Following is a sample list of the controls and may include any or all of the following: • Prohibition of right turn on red at signals (to facilitate pedestrian crossing), • Pedestrian-only phasing and/or all-red pedestrian-only lead phasing, • Cycle length to accommodate progression for certain movements, • Provision for (or preclusion of) pedestrian crosswalks, • Lane widths of 10 feet preferred (to minimize pedestrian crossing distances and promote lower speeds of traffic),

102 a performance-Based highway Geometric Design process • City bus or single unit truck assumed for intersection design, • Small corner radii to slow turning traffic and increase queuing space for pedestrians at intersections, • Raised medians for streets with more than six lanes (three each direction) to facilitate pedestrian crossings, and • Space for transit stops (near side or far side) within intersections. Setting these as design controls will translate directly to geometric elements such as lengths of left- and right-turn lanes (produced by queuing behavior), radii at intersections, total width of the roadway, and others. 5.6 Step 6—Apply the Appropriate Geometric Design Process and Criteria Under the proposed geometric design process there would be unique published geometric design criteria and processes applicable to the project as defined by its type and problem being addressed. Figure 33 demonstrates the simple decision tree for the types of design criteria envisioned. The process is based on the aforementioned unique aspects of projects as defined by type and the problem being addressed. Figure 33 illustrates the process using the core documents, including the HCM, HSM, and the AASHTO Green Book. The intent is not to specifically call only these documents out, but rather to demonstrate that the geometric design process will include not only published dimensional criteria, but will rely on methodologies, tools, and best-practice information that characterize the operational and substantive safety effects of geometric elements and their dimensions. This simple diagram outlines a decision framework focusing on the type of project and nature of the problem or issue being addressed. Both the design process and applicable design criteria Figure 33. Problem-based unified geometric design process.

performance-Based highway Design process 103 differ based on the applicable approach. (The possibility that multiple problems may exist is acknowledged. For the purposes of outlining the basic philosophy and approach, consider that the process would be driven by the problem that is most central or critical to the project’s programming and budget allocation.) 5.6.1 Roads on New Alignment Are Designed with a Unique Process and Criteria The decision tree of Figure 33 shows two initial possible paths based on the type of project. A road on new alignment would be designed in accordance with a revised AASHTO Green Book for Roads on New Alignment or its equivalent. The geometric criteria in this Green Book would be updated to reflect operational and safety relationships and cost-effective principles, including the incorporation of traffic volume in formulation of criteria for the full range of cross section, alignment, and sight distances. The derivation of the criteria would be based on cost-effective approaches using the concept of service life. Chapter 3 of this report provides an overview of suggested approaches to revising AASHTO criteria for new roads. Designers would use other references and tools such as the HCM and HSM as they are used today, to size the roadway and test and compare alternative designs. Revised design criteria reflecting context, sensitivity to measurable performance outcomes, and traffic volume would be developed with greater depth and potential important differences based on context compared with the current Green Book’s contents. 5.6.2 Design of Projects Involving Existing Roads The second major path in the decision tree is for an existing road. Here, the geometric design process proceeds based on the type of problem(s) that are behind the reason for the project. • If the project is solely to address infrastructure state of good repair, it would be considered a 3R project. No revisions to geometry would normally be considered, rather, the design focus would be on addressing the infrastructure condition need. Agencies by policy may choose to programmatically include safety enhancements for which the marginal additional costs are low or negligible (e.g., adding rumble strips as part of pavement resurfacing). They also may undertake programmatic measures such as pavement marking upgrades, sign replacements, and utility pole relocations. • If the project need is fundamentally to improve safety or traffic operations, the geometric design process is focused on the body of substantive safety knowledge or traffic operational knowledge (as published in the AASHTO HSM or other technical references designated by the owning agency). A project may involve both operational and substantive safety needs, in which case both sets of references would apply. Systemic safety improvements should be included to any projects that have the identified risk factors. • For reconstructed roads, the recommended process further differentiates between roads to remain as the same fundamental type, and roads to be reconstructed within a right-of-way corridor, but to a fundamentally different type. Changes in type would include addition of continuous basic lanes (e.g., expanding a two-lane highway to a four-lane highway), converting from a semi- or uncontrolled access facility, and constructing an interchange or roundabout to replace an intersection. 5.6.2.1 Reconstructed Roads to Remain the Same Type The geometric design process for reconstructed roads to remain the same type focuses on development of cost-effective solutions, with the actual safety and operational performance of

104 a performance-Based highway Geometric Design process the existing road serving as the baseline for alternative development. The actual crash history is directly relevant and applicable to the expected safety performance, from which crash cost benefits are computed. Actual speed and travel time data also are used to compute user operating and travel time costs. Proposed improvements to the road to address the identified needs would come from reference documents of proven effective treatments such as the HCM and HSM. The envisioned process for such projects would be a cost-effective analysis approach (versus an application of dimensional criteria). The project context plays a dominant role in determining cost effectiveness; hence, tailored solutions would be produced. For example, projects involving a horizontal curve with a high potential for safety improvement (PSI) may have different solutions based on the specifics of the site. In some cases, curve flattening may be the best solution; in others widening and paving the shoulder, treating the roadside, or implementing speed reduction measures may be more cost effective. • Geometric revisions would be made only for those roadway elements directly related to the identified problem. • A design process would apply science-based knowledge on operation and safety performance, with the designer developing alternatives using the tools and methods from the HCM, HSM, or other similar references, with benefits calculated for the service life of the project. • The location-specific costs of construction and maintenance would be calculated over the service life, and an optimized cost-effective solution obtained from the benefits and costs. • The roadway geometry at the project limits would provide a strong basis or reference for potential revisions. • The process would naturally favor solutions that do not require additional right-of-way (or require readily available right-of-way in return for a significant improvement in performance). The process also would not require the comparison of the solution with published criteria and a design exception. 5.6.2.2 Benefit/Cost Process for Reconstructed Roads The following summarizes the geometric design process based on comparison of the benefits and costs of alternative design solutions. The process relies on objective measures of transportation performance: • Crash costs would be derived from AASHTO HSM or other approved references for effects of roadway geometry and treatments on crash frequency and severity. • Crash costs would be based on valuations for fatalities, injuries, and property-damage- only crashes as published in research (e.g., FHWA) and adopted by policy of the owning agency. • User operating costs would be based on research and tools developed from the AASHTO Manual on User Benefit Analysis for Highways 3rd edition (AASHTO 2011c) or other similar work, which provides operating cost curves for different vehicle types as a function of operating parameters such as speed (Figure 34). • User travel time costs would be also based on research from the same AASHTO publication or other similar work, which provides travel time costs as a function of trip type and time saved. (For both operating costs and travel time, incorporation of costs for the range of vehicle types is possible.) • Initial construction costs would be used per standard agency approaches for cost estimating. • Annual M&O costs would be obtained from agency asset management systems and data- bases as well as research on the costs of maintenance as a function of context and road geometry. The above valuations and methods would be set by policy and regularly updated for inflation and as additional research is available. Agency staff would be provided software to

performance-Based highway Design process 105 enable efficient and consistent application of the analysis approaches briefly summarized as follows: Step 1: Develop annual costs for base alternative.* • Base alternative will typically be the existing condition or no-build. • Determine service life of project and discount rates.# • Establish service life traffic projections. Calculate Annualized User Costs^ crash costs vehicle operating costs user travel time costs for service life.∑= ++  ∑= + +    Calculate Agency Roadway Costs* initial construction intermediate construction annual M&O throughout service life. Intermediate construction would entail infrequent, major construction such as bridge deck replacement and resurfacing. *Agency roadway costs should assume a level of construction during the service life to maintain a minimum state of good repair by policy and/or regulation. This may include periodic resurfacing and bridge repairs. Agency costs for M&O would be developed from asset man- agement programs to a level of detail supported by the agency’s databases and management resources. #Annual costs require the setting of discount or interest rates for analysis purposes. These would be set by agency policy, and may vary to reflect the level of risk or uncertainty associated with Figure 34. Fuel costs (cents per mile) by speed—automobiles (AASHTO 2003, figure 5-1).

106 a performance-Based highway Geometric Design process a project type or types. The lower the discount rate, the more future projected costs and benefits will represent meaningful values in the analysis. ^Agency crash costs would be based on application of agency processes and models as well as valuations for injuries and fatalities per agency policies. Vehicle operating and user travel time costs would reflect traffic type distribution (e.g., freight and transit) and agency policies for valu- ation of travel time. And established operating cost relationships would be based on the literature. Step 2: Develop Alternative Solutions–Consult appropriate references for geometric design and traffic operational solutions that address the problem(s); select and design the context. Designers would be provided initial guidance on the most effective solutions given the traffic volume levels, crash types, and expected costs. Effective solutions will be expected to generate lower annual crash costs, lower vehicle costs, lower user travel time costs, or a combination of these. They would select a reasonable number of solutions, perform sufficient engineering design to establish the three-dimensional footprint, estimate right-of-way, and estimate construction costs. Based on the design solutions they would perform the following: Calculate Annualized User Costs for Alternative crash costs vehicle operating costs user travel time costs for service life.∑= ++  ∑= + +    Calculate Annualized Agency Roadway Costs* initial construction intermediate construction annual M&O Effective solutions may involve initial construction including right-of-way beyond the base condition. Intermediate construction costs would reflect the timing and extent of the improve- ment. Annual M&O costs for the alternative may be greater or less than the base condition depending on the specific solutions. In general, the annualized agency costs for an improvement will be greater than the base condition annualized costs. *Agency roadway costs should assume a level of construction during the service life to maintain a minimum state of good repair by policy and/or regulation. This may include periodic resurfacing and bridge repairs. Agency costs for M&O would be developed from asset management programs to a level of detail supported by the agency’s databases and management resources. Step 3: Determine cost effectiveness of the alternative. Annualized benefit to cost B/C ratio User Benefits Increased Agency Costs.∑ ∑( ) = ∑∑ ( ) ( )= − + −   + − Where User Benefits crash costs for base alternative crash costs for alternative vehicle operating costs for base alternative vehicle operating costs for alternative travel time costs for base alternative travel time costs for alternative Increased Annualized Agency Costs total agency roadway costs for base alternative total agency roadway costs for alternative .∑∑ = −   Multiple design alternatives would be tested against each other, with the optimal alternative among build solutions producing a marginal B/C ratio > 1.0.

performance-Based highway Design process 107 Agencies would need to establish policies for minimally acceptable B/C ratios based on their total available resources and priorities. The quantitative benefits will generally accrue to agency customers—the road users. In many cases, these benefits will require an increase in agency costs through initial construction, long-term maintenance, or both. The optimal solution among multiple alternatives may require the greatest long-term investment by the agency. For these reasons, a minimum project threshold for B/C ratio > 1.0 may be programmatically untenable for the agency. The setting of minimal threshold B/C ratios assures that every project will be affordable over time (at least to the extent the agency is able to forecast its future costs and revenue). The threshold policies could change over time as an agency’s available resources change. Indeed, the level of an affordable B/C ratio would be a key management metric that would be useful to those responsible for providing funding and resources to the agency. This process does not rely on externally published dimensional criteria. It therefore would not necessitate the characterization or approval of an exception to design policy for any solution. Given this framework, the only exception may be the selection and implementation of an alternative that does not meet minimal agency thresholds for cost effectiveness. Assuming that agencies adopt policies that value fatalities and injuries in a manner consistent with that given in the HSM, this process will naturally result in designers focusing on alternatives that produce measurable reductions in serious crashes, and that do so with minimal construction. This outcome is the central objective of agency policies such as practical design. 5.6.2.3 Design of Reconstructed Roads for Conversion to New Road Types A separate AASHTO policy for reconstruction of existing roads is suggested for roads fully reconstructed within existing right-of-way to a different road type (e.g., two-lane rural to multilane, three-lane urban to multilane, road for motor vehicles only converted to one for multimodal use). The proposed AASHTO Green Book for Reconstruction of Roads would include design criteria similar in nature to the New Roads Green Book, but it would describe a design process in which appropriate dimensions are established that reflect the practical limitations of the location and the incremental costs and benefits of dimensions in the specific context. If the nature of the primary problem being addressed is traffic operational, the design process would be diagnostic, using the knowledge, tools, and methods representing best practices (e.g., the TRB HCM or other manuals or methodologies endorsed by the owning agency). In many cases the traffic operational problem may be traffic demand exceeding the capacity of the road during certain times of the year, in which case the HCM would serve as the basis for provid- ing additional capacity, managing demand, or optimizing the operation of the road. Operational improvements might include construction of medians for left-turn refuge, retiming of traffic signals to facilitate pedestrian trips, and restriping or widening of the cross section to enable bicycle operations. The dimensions associated with achieving improvements may vary based on the context, for example, with narrower left-turn lanes adopted should this be the only way to provide left-turn capacity given limited space. If the problem is a substantive safety problem, the design process would be diagnostic, using the knowledge, tools, and methods representing best practices (i.e., the AASHTO HSM). Geometry unrelated to the specified problem can be presumptively retained. Changes in the operational performance of the roadway associated with the safety improvements would be characterized and included in the final documentation of the selected design solution. For those projects in which both safety and operational problems need addressing, the design process would direct designers to consult both primary references. Agencies may develop shortcut procedures or establish hierarchies of preferred approaches that combine operational and safety issues. Differences among agencies may be expected based on their unique contexts,

108 a performance-Based highway Geometric Design process budget priorities, and policies, but the underlying approaches would be consistent assuming the agencies are referencing the same body of technical knowledge on the geometric effects on safety and operational performance. Regardless of the problem(s) being addressed, designers would have the flexibility to adjust or redesign alignment or cross section, but their emphasis should be on solving the problem given the unique site constraints, including the availability and costs of right-of-way and impacts during construction to adjacent landowners. The design process would directly incorporate such costs in the determination of the optimal solution. The reconstruction design process would automatically promote the concept that an appro- priate solution for a given problem in one context may be unworkable in another, thus requiring a different solution (or perhaps different dimensions). With an expanded knowledge base and training, designers would learn to quickly focus on those solutions most likely to provide the optimal value. Moreover, their recommended design solution would be optimal for the context and thus not in need of being labeled a design exception. Finally, road geometry incidental or unrelated to the identified problem could presumptively remain unless there is evidence from performance models of a meaningful loss of performance value with a given alternative. 5.6.2.4 Design of 3R Roads If the problem being addressed is only infrastructure condition, the road would be designated as 3R. The basic geometry would remain unchanged, unless the designer can demonstrate that the revisions will be cost effective in terms of the return on investment using the established benefit/cost analysis procedures. The 3R process can allow for designers to include low-cost safety improvements when the marginal cost of such improvements is clearly small or negligible. For example, in resurfacing a shoulder, the inclusion of rumble strips can produce meaningful benefits with minimal additional costs. 5.6.3 Develop Project Technical Approach The project technical approach is the detailed work plan for the project team. This would include what data are needed and to what level of detail or precision (survey, traffic, crash, utility information, environmental data, and construction cost). The project technical approach also would include the outlining of interim and decision deliverables leading up to the selection of the preferred design plan. 5.7 Step 7—Designing the Geometric Alternatives Only after the previous four steps have been carefully and fully completed should the geomet- ric design process proceed. First and foremost, the process must be understood and executed as an alternatives development process. Regardless of the project size, scope, and problems being addressed, there is always more than one reasonable geometric solution. The process itself involves multiple steps and subprocesses outlined below. 5.7.1 Assemble an Inclusive and Interdisciplinary Team Geometric design is the technical means by which the intended performance of the roadway is accomplished, within the bounds of affordability and respect for the context. Geometric design- ers should have strong working knowledge and/or access to other individuals who can work with them in the disciplines of traffic operations, hydrology and drainage, bridge and structures, geotechnical engineering, and construction. In particular, traffic engineering and operations

performance-Based highway Design process 109 expertise is needed. In many cases the optimal solution will include stakeholder engagement activities, and combined geometric design and operational components. 5.7.2 Focus on and Address the Need or Solve the Problem(s) Within the Context Conditions and Constraints The geometric design process should be focused on solving the objective, identified problem(s) confirmed in Step 3. Solutions that are cost effective reflect the unique context conditions and constraints established in Step 4. For projects involving existing roads the starting point is the existing geometry and footprint of the road. For projects on new alignment, the road designer has a clean slate and thus requires some basic guidance to initiate alternatives development. For these project types, designers require basic geometric design criteria as a core starting point. Such criteria by necessity must be expressed as physical dimensions. The use of design criteria expressed as dimensions should be understood as a means to an end, the end being performance (operation and safety). The costs associated with constructing given dimensions, and the resultant performance of the road built to such dimensions, can vary based on the unique context conditions. For these reasons, the geometric design process for roads on new alignment may produce unique or tailored designs more often than not. 5.7.2.1 Unique Designs and Driver Expectations The concept of unique versus standard or typical designs requires explanation, as some designers express concern over driver expectations when a unique design is offered. The concept of uniqueness addresses the context-specific features of the roadway environment that influ- ence how well it respects the surrounding land use and features, and what the costs to build and maintain it may be. Table 18 shows driver expectations from human factors research (Campbell et al. 2012). They can be summarized as follows: • Design of decision points or route choices (turns at intersections, exiting, through movements); • Placement, type, and messages of signs and traffic control devices; • Colors and patterns of traffic control devices and features; • Appearance of the alignment (in particular, horizontal alignment) and consistency with preceding alignment; • Abruptness and severity of changes in width; and • Notice to drivers of significant changes in the road character. Done properly, there is no inherent risk to driver expectations associated with meaningful revisions in the cross section (lane widths, medians and median width, roadside character) or alignment as long as sufficient notice through both sight lines and traffic control devices is pro- vided. Should a unique context feature or features exist, the design process for new roads should be sufficiently flexible not only to allow but encourage the designer to test alternative geometric solutions and select the optimal design based on the performance and cost analysis of the range of possible solutions. Done properly, such a process should result in the preferred solution and not one that is labeled an exception. 5.7.2.2 Exercise Design Flexibility—Choices and Trade-offs AASHTO has promoted the notion of design flexibility for some years, including most nota- bly in its policy document A Guide to Achieving Flexibility in Highway Design. In practical terms, design flexibility means that designers have choices and not mandates. Choices relate to the level

110 a performance-Based highway Geometric Design process of transportation service provided for all modes, the inclusion or exclusion of specific features (lanes by type and usage, medians by type, on-street parking), treatments (intersections versus roundabouts, types of intersections, types of interchanges), and dimensions for each element. The updated geometric design process for new and reconstructed roads will ideally promote the hierarchy of choices with respect to their relative importance in influencing operational and safety performance. Figure 35 illustrates this hierarchy. The most important decisions involve Arrangement of Roadway Features and Navigation or Route Choices Arrangement of Roadway Elements Influencing Speed, Acceleration, and Deceleration Behavior Location, Design, and Messages of Traffic Control Devices to Communicate and Reinforce Appropriate Driver Behavior Upcoming freeway exits will be on the right-hand side of the road When a minor and a major road cross, the stop control will be on the road that appears to be the minor road Mix of two-way and all-way stop control along a route Consistent interchange exit type (before or after cross road structure) Speed limit changing for a short segment or dropping by more than 10 mph Consistent roundabout signing and lane use At splits where the off-route movement is to the left or where there is an optional lane split Tight curve after long tangent Traffic signal placement and number of heads provided – mast arm mounted signal heads When approaching an intersection, drivers must be in the left lane to make a left turn at the cross street, vice versa for right Narrowing of lanes, shoulders, or other cross-section elements Consistent sign size and placement Intersection, access point, horizontal curve after crest vertical curve Freeway acceleration lane reduction made so far downstream that motorists become accustomed to a number of lanes and are surprised by the reduction Uses of colors on signs, pavement markings and traffic signals to denote core messages Trap lane, through lane becoming turn only or special purpose lane Uniform application of traffic control devices with respect to the amount of change in the roadway alignment conveys a consistent message Consistency of navigational signing (type and frequency of signing) A continuous through lane (on a freeway or arterial) will not end at an interchange or intersection junction Prohibited turn movements Source: Derived from information in NCHRP Report 600: Human Factors Guidelines for Road Systems (Campbell et al. 2012). Table 18. Design and traffic control measures to address driver expectations. Figure 35. Hierarchy of roadway design decisions and their relative importance.

performance-Based highway Design process 111 roadway type (both segments and intersections, including the extent of access control and number of lanes). Of secondary importance is the presence and arrangement of roadway features such as medians, turn lanes, and intersections. Of lesser importance are the dimensions associ- ated with the roadway elements (i.e., within the normal or accepted range of those dimensions). For example, the speed, capacity, and crash-risk profile of a road in a rural context is much more influenced by the basic type (two-lane, multilane with partial access control, multilane with full access control) than the dimensions of lane width, shoulder width, curve radius, or SSD. Designers should understand that their fundamental mission is to provide the optimal design for the given roadway type. Flexibility and creativity contribute to the assembly of the three dimensions of the road to fit the context, maximize transportation value, and minimize life-cycle cost. This approach conflicts with the traditional design thought process as employed by most highway engineers and promoted by most agencies. The traditional approach focuses on design dimensions as the most important feature; moreover, it reflects the mindset of transportation performance and value illustrated by the green line in Figure 15. Minimum design criteria, typically expressed in dimensional values related to a design speed or other control, are automatically selected as the preferred solution. This approach applies to cross sectional elements and SSD (the latter value is not directly designed but rather translated into dimensions for vertical cur- vature and horizontal offsets). Some designers do not explore or test greater widths or longer vertical curves other than the minimum values because (1) they believe they will automatically cost more to construct, (2) influential stakeholders push for the impacts to be minimized, and (3) they are not believed to provide additional transportation value. Many designers do employ a full range of values for both horizontal and vertical alignment, but even this design thought process reflects a basic construction cost focus, with curves and grades selected to fit the terrain and tie into other roads and to avoid physical conflicts or minimize retaining walls and other costly structural elements. Among the most significant changes in the geometric design process should be the adoption of the mental approach to design decision making illustrated by the orange line in Figure 15 shown previously. A highway designer needs a starting point or base condition, but the process of dimensional choices and trade-offs should drive the designer in all aspects of the design (cross section, horizontal alignment, and vertical alignment). 5.7.2.3 Design to an Appropriate Level of Detail Throughout the Design Process The roadway design process, of which geometric design is a centerpiece, is actually a series of subprocesses that translate thoughts and ideas into design data and drawings to a level of detail sufficient for construction. Individual phases of the design process serve varying needs and involve different types and criticality of decisions. The geometric process of alternatives is a continuum as demonstrated in Figure 36. Initial designs may be developed in plan-view only, referenced to aerial photography with simple sup- porting sketches. Skilled designers can effectively portray a plan that accommodates the third dimension (elevations) without actually designing such dimension, even for relatively complex projects such as interchanges. Taking full advantage of computer-aided engineering tools, this process facilitates development of multiple solutions. Early in the alternatives design process, the design engineer works in appropriate scales for the context and applies as much detail as is necessary to confirm key impacts such as right-of-way conflicts to enable screening down to the most promising designs. Many engineering details are purposely ignored or deferred for investigation in later phases. Design drawings are generated both for internal working purposes and external stakeholder interaction.

112 a performance-Based highway Geometric Design process The concept of using proper scales in design is important regardless of the ability for the designers to print or plot their plans at any scale. Working at too large a scale (1 inch equals 20 feet) in early concept planning is highly inefficient as it is difficult for designers to absorb the informa- tion, and easy for the designers to have their attention diverted to details that do not matter early in the process. Also, stakeholders (many of whom will be non-technical in their background) will require varying levels of detail and scale to perform their reviews, and the appearance of deliverables for their use should be tailored to maximize their understanding and facilitate their input. The concept of varying the scales and appearance of deliverables is illustrated in Figure 37. Multiple alternatives can be evaluated for their effects and costs, enabling screening and selection of the most promising alternatives. As geometric design progresses and the number of Figure 36. The continuum of geometric design. Figure 37. Tailoring of the design approach and deliverables to the major project phases.

performance-Based highway Design process 113 alternatives is narrowed, the level of effort and detail is increased. The third dimension may be developed over digital terrain mapping (vertical alignment) with supporting engineering showing basic structure dimensions. Additional insights such as visualizations or “virtual drive throughs,” estimates of earthwork quantities, and modeling of environmental impacts such as noise can be completed. When the process has reached the point at which the number of viable solutions is limited, further study detail (geometric design and performance) can be efficiently performed for the remaining viable solutions. This concept is described in Figure 38. For the most promising alternatives, the level of detail should be sufficient to fully characterize the necessary right-of-way and basic construction elements and quantities. As a minimum, the geometric design involves the horizontal alignment of control line geometry and edges of pave- ment, the typical section, and the vertical alignment of the control line and edges of pavement. The geometry should be referenced to the terrain developed from mapping to a sufficient level of quality such that right-of-way lines can be set and evaluated. 5.7.2.4 Maintenance and Operations The recommended process for all project types should incorporate explicitly life-cycle costs of O&M activities. Section 8.3 contains a detailed discussion of the relationship of maintenance to roadway design. Note that maintenance needs and issues are strongly context sensitive. O&M needs and costs will vary by road type, terrain, climate, and traffic levels. The knowledge base on the incremental or unique O&M costs associated with highway design features or incremental dimensions is limited. The Handbook of Road Safety Measures contains some information on this topic, but there is clearly much more research needed that is beyond the scope of this research project. 5.7.2.5 Performance Iteration in Geometric Design The conventional or traditional geometric design process has been an iterative process in which the highway designer’s efforts are focused on minimizing the construction quantities and right-of-way for a given geometric solution. Figure 38. Tailoring the technical approach of performance evaluation to the phase of alternatives development.

114 a performance-Based highway Geometric Design process Advances in the knowledge base of both traffic operations and safety are such that the nominal safety mindset of Figure 15 is demonstrably not correct. The measurable performance of a design can and does vary as dimensions and elements are tested and iterated. Alternatives fully meet- ing applicable design criteria may differ significantly in their expected safety and operational performance. The suggested geometric design process incorporates science-based analytical models and methods that provide predicted quantitative outcomes based on the geometry. Such outcomes would ideally include: • Expected cost to construct, • Expected crash frequency over project life, • Expected crash severity over project life, • Expected travel time(s) by time of day over project life, • Expected cost to operate vehicles over the project life, and • Expected annual maintenance over project life. As shown in Figures 37 and 38, performance-based tools tailored to the phase in the process can provide sufficient information to each phase. Look-up tables, default values, and shortcut analysis procedures may be used to screen to a limited few designs, and full modeling (microsimulation for operations; HSM predictive methods via FHWA’s IHSDM for safety) for those few surviving alternatives is an efficient approach. For simple projects, full simulation may not be necessary. Transportation performance outcomes for a given iteration could be translated to annual user costs (crash costs, travel time, operating cost) and to agency life-cycle costs (annualized construction and annual maintenance costs) for the entire service life of the project. A cost-effectiveness function or index value could be produced for the alternative. Subsequent iterations in any aspect of the design would produce differing performance and cost values, and differing cost-effective functions or index values. 5.8 Step 8—Design Decision Making and Documentation The ultimate decision regarding the preferred geometric design solution is made by the agency responsible for funding, construction, and maintenance. Many projects will be performed within an environmental regulatory framework that requires complete analysis and documentation of expected impacts and mitigation measures. The recommended geometric design process adds the expected traffic operational and safety performance benefits of the optimized plan. Assum- ing Steps 1 to 7 are conducted as discussed, the recommended plan should be evident from its optimized features. The most important documentation should be the statement of the problem being addressed and summary of how the recommended design will address the solution. In the context of the U.S. environmental process, this is referred to as meeting the purpose and need. This documentation and discussion should be quantitative in nature. The documentation should also include the geometric design framework, design controls, constraints, and decision processes, with particular attention paid to describing the values of the community and decision makers and how they influenced the selection of the preferred alternative. This documentation should be referenced in the ROD which concludes an environmental impact statement, and the FONSI, which concludes the EA. 5.8.1 Independent Quality and Risk Management Processes A number of important independent design processes have evolved over the past 30 years and are now considered fundamental to roadway project development. These include VE, road safety

performance-Based highway Design process 115 audits (RSAs), cost-estimating validation processes (CEVPs), and alternative technical concepts (ATCs). These processes evolved as a result of a perceived lack of: • Optimization in value received as part of the traditional design project development process, • Sensitivity to safety performance by road designers, and • Foreseen circumstances or conditions that greatly increased the cost of projects (usually large and/or complex projects). Many agencies purposely separate Steps 8 and 9 either internally or in their use of design consultants. In the intervening periods of re-assigning the project to others, these processes are often undertaken. The following is a discussion of how these independent processes may fit or best be performed in the revised geometric design process. 5.8.1.1 Value Engineering VE is a specific process required for projects using federal funds above a minimum con- structed value threshold. VE involves a carefully assembled multidisciplinary team that meets over a one-week period, facilitated and managed by a trained, professional VE specialist. The concept behind VE is to challenge and test the assumptions, controls, and the specific solutions to a project in an independent manner. The VE participants seek to increase value, reduce cost (without degrading value), or both. Their work product is a report to the owner that outlines specific refinements or changes to the plan, with an assessment of the potential value increase or cost saving. Project owners then decide to accept or reject each suggestion. VE is most effective when incorporated earlier in final project design development. VE work- shops that are held at the outset of the preliminary and final engineering stages offer the time to implement change and avoid wasting or having to redo final design efforts. The suggested design process that is more performance based and that focuses on alternatives and the trade-offs blends very well in that performance or value (what is the problem in objective terms, what solution was reached, how well will the solutions address the problem in objective terms) is highlighted. This serves as a firm benchmark on which the VE process can build. Even a well-completed project can benefit from trained, independent observers within a well-run VE workshop setting. 5.8.1.2 Road Safety Audits RSAs have become a standard tool for many agencies. An RSA is conducted by a multidisciplinary team of traffic engineers, safety specialists, law enforcement, human factors, and maintenance and construction experts. Through a combination of data analysis and review, and collaborative field exercises to view the project site, the RSA team seeks to make low-cost safety improvement recommendations that are consistent with the scope and nature of the project. These may often represent insights that go beyond published standards, but relate to unique risks or conditions that may pose particular problems to drivers. Some agencies apply RSAs to certain project types (typically HSIP or safety-focused 3R or reconstruction projects); others are seeking to incorporate RSAs on all project types involving existing roads. The revised geometric design process should acknowledge this independent sub- process, and encourage its application at appropriate times as follows: • For 3R projects (in which state of good repair is the fundamental problem) the RSA may best be conducted at the beginning of the project—after Step 4 and during Step 5. Indeed, the RSA may be a primary source of ideas for including very low-cost safety improvements as part of the project.

116 a performance-Based highway Geometric Design process • For reconstruction projects that may include safety performance, operational performance, or both, the RSA may best be performed following Step 8, with the focus being a review of the preferred plan. This will necessarily involve a virtual review using design visualization of the three-dimensional geometry. • For roads on new alignment, the value of an RSA is questionable. If performed, it should occur during Step 9, with the focus of the review being a virtual drive-through of the roadway. The RSA may include review of signing and traffic control placement (noting visual conflicts with overhead structures, tree limbs, and plantings) and a second look at roadside barriers and attenuation to ensure they are both necessary and properly placed. 5.8.1.3 Cost Estimate Validation Process The CEVP was developed by the Washington State DOT (Washington State Department of Transportation nd). It addressed a problem associated with major highway projects of great complexity having long development times. The many risks associated with moving through project development to construction often resulted in unforeseen increases in construction cost that emerged between the time the final decision was made and actual construction plans were completed for bidding. CEVP uses a Monte Carlo simulation process in which many aspects of the project are assessed as to their risks of change over time and resultant potential effects on constructability and cost. The CEVP allows characterization of a probability that a project will not cost more than a given amount. It also provides an analytical method for testing potential changes, and for proactively managing the project as it moves through final engineering to address the risks of change. The revised geometric design process does not mitigate or reduce the need for or value of the CEVP. Indeed, an optimization process is reliant on the quality of estimates of cost, and to the extent that these are uncertain in the middle of design development, the risk is that the solution will end up being suboptimal. The CEVP should be performed early in Step 9 and updated depending on the length of time necessary to complete final engineering. 5.8.1.4 Alternative Technical Concepts The ATC process is a fundamental process within the alternative project delivery design-build (DB) process. A typical DB project involves three to five competitive teams who are typically provided pro- posed contract terms and an owner’s preferred design, which typically reflects a level of engi- neering consistent with that described in Step 8. It includes the basic roadway design in three dimensions, right-of-way, environmental mitigation and related features, and basic structural solutions. The terms for selecting the winning firm are spelled out in the request for proposals. These may be lowest constructed cost, lowest cost with credit for reduced time of construction, or (in states where legislation allows this) a selection in which factors relating to schedule, quality, additional features, and cost may all be used to score a proposal. Selection of the winning team is through a carefully defined process that may include a scoring process in which schedule, budget, and quality (clearly defined) are evaluated. The process where allowed by state procurement laws is often referred to as best value. Contractor teams compete in a process to complete the design detail using their own ideas; this does not necessarily require all the engineering details that agencies typically seek for tradi- tional plans. An important feature of DB is the inclusion of an ATC. Owning agencies furnish the 30 percent plans and ground rules or boundaries for ATCs, but then invite the competitive teams to submit their own unique ideas for changing the design to accomplish a stated goal or objective of the owner in a manner demonstrably better than that provided by the owner’s plan.

performance-Based highway Design process 117 ATCs can include substantive changes to the original concept. For example, they may include reconfiguration of an interchange, realignment (horizontal, vertical, or both) to reduce construc- tion costs, reallocation of widths in the cross section, or other major geometric changes. ATCs also may involve alternative pavement designs, alternative structure designs, and different traffic control plans than originally envisioned during construction. The ATC process is where the potential value over traditional design/bid/build exists. The FHWA and AASHTO acknowledge that the current project development process often produces less than optimal design solutions. ATCs represent the opportunity for a fresh set of eyes to review the project and propose innovative ideas to meet owner objectives. As such they may continue to be part of project devel- opment. The revised geometric design process outlined above, if carried out as intended, would automatically address many of the perceived shortcomings of current project delivery that have led to the attraction of ATCs. Many ATCs emanate from projects that did not undergo rigorous alternative development. Many ATCs also involve design exception ideas in which the ATC pro- poser demonstrates performance values at lower costs by changing a design dimension or design control. A well-performed design study would tend to negate many of these sources of ATCs. Some ATCs involve materials and specifications, construction means and methods, or other elements that may have limited or no influence on the geometric design of the road. For major projects, ATCs often focus on differing ways of managing traffic or detours during construction to minimize costly phasing and shorten construction time. These would be unaffected by a new design process. For agencies employing ATCs, it is critical that the stated design controls and values devel- oped during project development are retained in the value scoring process and in the guidelines around what is considered acceptable. Project procurement (DB or conventional) is often man- aged by different staff and offices within DOTs. As projects are handed off the institutional memories and values associated with decisions risks being lost. This is one major reason for the emphasis on good and complete design documentation. 5.9 Step 9—Transition to Preliminary and Final Engineering The final step in the geometric design process follows the selection of the preferred plan. This includes the full design and detailing of all elements of the roadway, including below-grade features such as closed drainage systems and utility relocations. The scale of the drawings pro- duced is typically increased and multiple notes and other specifications added to the drawings to produce bid-ready and construction-ready documents. Note that this last step of the design process—the detailing of the basic selected functional geometric plan—typically requires 70 percent or more of the engineering effort and design schedule. It is not necessary and indeed wasteful to perform detailed design work and calcu- lations on more than one alternative, when the element in question either has no bearing on selection of the preferred plan, or the geometric details of the preferred plan are needed to enable the calculations. The preliminary and final engineering phases of work are performed on the selected and approved geometric design. The focus of these phases is to add design detail regarding road elements separate from the basic roadway. This effort includes surveys and plans for right-of-way acquisitions and easement. Engineering design also include subsurface elements (utility locations or relocations and stormwater management), detailed drainage and erosion control, lighting,

118 a performance-Based highway Geometric Design process signing and ITS elements, traffic control devices, roadside barriers, sound walls, striping, and landscaping. The functional design (type and location) of guardrail and other roadside barriers should be advanced into Step 7. The presence of barriers has a known and significant influence on safety performance. Exercising the HSM or other safety models thus requires at least preliminary design of barriers. Barriers also are significant maintenance elements. The exercise of cost models for roadway maintenance will undoubtedly require location-specific information on guardrail and other barriers during the alternatives evaluation and selection steps. Step 9 engineering design also will include preparation of detailed plans for maintenance of traffic during construction. Bridge design work in earlier steps is typically limited to sketch plans or type, size, and location plans; with detailed bridge calculations and final design occurring in final engineering. Specifications for construction also are prepared in the final engineering phase of work. It is commonplace, particularly in urban contexts, for there to be revisions to road geometry as the level of detail increases and conflicts or unforeseen conditions arise. In most cases these changes will be minor in terms of the effect on the operational and safety performance intended by the efforts through Step 6. However, circumstances sometimes arise that force major changes in the plan concept. Such circumstances may include preliminary cost estimates that greatly exceed the budget requiring redesign, subsurface conditions that require different structural solutions, and difficulties in acquiring right-of-way that result in redesign to avoid the acquisition. The majority of the effort involved in developing plans for construction is typically included in these phases of work; generally considered to be 70 percent of the overall engineering design effort. What is unique about this 70 percent effort is that it is solely focused on providing the detail sufficient for bidding and construction. Once the basic road design is set per Step 6, the focus changes away from transportation performance to the implementation tasks of procurement and construction. For major projects an important subprocess would be the rerunning and confirmation of the analyses that documented the operational and safety performance of the final plan. This should occur at the 90 percent to 95 percent stage of engineering (typically associated with quantity estimates, specifications development, and final quality assurance reviews). 5.9.1 Technology Applications—Building Information Modeling Geometric design and the assembly of final engineering plans has traditionally been a paper- focused, two-dimensional process in which calculations and data describing the road are translated to drawings that are dimensional, with written notes and specifications supporting the plan sheets. In recent years, advances in infrastructure technology development have produced what is referred to as building information modeling (BIM). The U.S. National BIM Standard Project Committee has the following definition: Building Information Modeling (BIM) is a digital representation of physical and functional character- istics of a facility. A BIM is a shared knowledge resource for information about a facility forming a reliable basis for decisions during its life cycle; defined as existing from earliest conception to demolition. (Zegeer and Neuman 1993). BIM treats roadway design data in five dimensions—spatial (width, height, and depth), time, and cost. BIM is the representation of a design as combinations of objects, with each object defined by data describing its geometry, relationship to other objects, and basic attributes. BIM design tools enable designers to extract different views from a building model (e.g., a roadway design) for production of design drawings and other uses. BIM software also defines objects as parameters in relation to other objects. If a related object is revised, dependent objects will automatically be

performance-Based highway Design process 119 revised. Each building model element can carry attributes for selecting and ordering the objects automatically, providing construction quantities for cost estimates, and material tracking and ordering. BIM is now routinely applied to major roadway design projects. When used to its fullest advantages, BIM allows a multidisciplinary team of roadway designers to: • Design virtually within a 3D world with all disciplines working simultaneously together; • Develop design iterations or scenarios that can be quickly and easily accomplished; • Generate prompt or automated quantity takeoffs from the BIM components; • Automatically detect conflicts or interferences among objects (e.g., utility and pile or founda- tion), thereby avoiding potential onsite errors or conflicts; • Facilitate design and constructability reviews via virtual model walk-throughs; • Accumulate core business data (e.g., a storm drain pipe may include size and material infor- mation) for use in assembling cost and schedule information; • Publish 3D PDF and DWF files for consumption by a wide audience; • Visualize construction during different phases for better visual communication and coordi- nation; and • Simulate construction from BIM components. The ability to visualize a design and design impacts in three dimensions is difficult for many designers working with two-dimensional plans. For example, roadways must be able to drain adequately along the entire length. Design criteria for cross-slope and minimum grade are intended to assure drainage. However, designers are often faced with situations in which a roadway profile may be in crest or sag curve, and also in horizontal curve or curve transition, in which the pavement is being rotated to develop the necessary superelevation. This unique combination of geometry may result in an unintended consequence of creating road segments that are flat longitudinally and also have no effective cross slope. A 3D BIM evaluation of the pavement con- tours can readily display this condition and assist the designer in adjusting the horizontal and/or vertical geometry accordingly. BIM enables the project designers to get real-time feedback of geometric-based elements and functionality during the assessment of the existing facility and the design of proposed alterna- tives. Overlaying the geometric model of existing conditions with design standards and geographic information system (GIS) statistical data allows the engineer to visually demonstrate locations of geometric conflicts or locations where criteria may not fit the context. During design, pro- posed alternatives are quickly evaluated against the design criteria, environmental acceptability, right-of-way availability, and construction impacts to maximize the cost/benefit ratio of making improvements to certain geometric and roadside features. The geometric model overlaid with GIS data on population density and zoning also can be used in the planning and sequencing of construction. Throughout construction, maintaining safe access for various homes, businesses, and other operational considerations is critical. In addition, ensuring construction personnel and equipment can safely navigate the work zone should be considered. The geometric model can be overlaid with seasonal construction schedules to evaluate sight distances and reduce accessibility or obstructions through the various phases of construction. BIM also allows for the direct output and transferal of machine control grading data, accelerating construction operations. BIM applications to geometric design include the following: • Verification of pavement elevations to assure drainage (Figure 39); • Review, confirmation, and adjustment to alignment for continuous sight lines to the roadway from driver eye locations (Figure 40);

120 a performance-Based highway Geometric Design process Figure 39. BIM output to evaluate vertical alignment design for drainage. Figure 40. BIM output to evaluate sight lines along curved roadway with barrier.

performance-Based highway Design process 121 • Placement of navigational signing and confirmation of adequate sight lines (conflicts with overhead structures and retaining walls on curved alignment); • Placement of roadside objects to avoid conflicts (signs, light poles, and traffic signals); • Placement of foundations for mast arms and major sign structures to avoid conflicts with subsurface utilities; and • Adjusting alignment and/or cross section to facilitate working space for construction equipment. A logical next step in a fully integrated, data-driven process is the direct linkage of geometric design data into the full suite of operational, safety, and maintenance models. This is the original vision of FHWA’s IHSDM, but taken to the ultimate conclusion. The revised geometric design process transitions the historic two-dimensional, paper-based process into a five-dimensional, virtual database development and evaluation process. Those responsible for roadway design are not merely highway engineers, but managers of large scale, robust databases to be used in all aspects of the highway project development process. 5.10 Step 10—Agency Operations and Maintenance Database Assembly The final step in the design development process is to input core project data into O&M databases. Agencies now employ asset management databases and systems to make strategic decisions about future investments; to plan, budget, and execute key maintenance activities; and to continually monitor the status of their infrastructure. The databases and algorithms imbedded in asset management systems rely on carefully defined performance measures. The recommended geometric design process includes objective analysis of O&M costs. Advances in technology, including real-time infrastructure monitoring conditions, predictive modeling of condition deterioration and failure, and asset management systems, are being implemented and gaining acceptance in the industry. An essential aspect of construction or implementation will be the inclusion of ITS technology in projects and the collection and evaluation of data obtained from the technology. The real-time monitoring and prediction of infrastructure condition to support programming and project decision making is comparable to operational and safety performance monitoring and decision making. The quality and currency of such data will in the future provide agencies the capability of optimal decision making at both the project and program level for both 3R and reconstruction projects. 5.11 Step 11—Continuous Monitoring and Feedback to Agency Processes and Database The final step in the process is the full utilization of data flowing to the agency from all sources on each roadway within their system. Real-time traffic operational-performance data, high- quality safety performance data, O&M data, and continuous self-monitoring of infrastructure condition become central to the design process. Such data are used to continually review and refine the geometric design processes, guidance, and decision rules. The geometric design process envisioned becomes dynamic in response to data and changes in performance that are observed over time. Consider the following trends: • As multidisciplinary approaches to addressing highway fatalities and serious injuries have been applied in many states over the past 15 years, marked decreases in fatalities have been observed. The effectiveness of any design strategy going forward is less now given the overall frequency of serious crashes than it was previously, which is good, but it heightens the need

122 a performance-Based highway Geometric Design process to constantly revise safety-effective design policies and solutions, assuming the positive trends continue. • The advent of connected vehicles, driverless cars, and other new technology is unknown. It is believed to have substantial safety benefits. As the vehicle fleet changes to include this technology, the benefits will eventually become evident in a further lowering of observed safety risks. To the extent that such technology plays a meaningful role in highway safety, the effectiveness of geometric design solutions to improve safety will lessen. This technology also offers the potential to improve traffic flow (by allowing shorter headways under high-volume conditions); which also would lessen the effectiveness of geometric design solutions aimed at improving traffic operations. • Widespread use of ITS solutions to monitor and manage traffic on all road types is an emerging trend. To the extent that ITS technology replaces the need for roadway infrastructure, this trend may demonstrate the ability to forego either certain design features, or be less constrained in the application of geometric controls (e.g., active traffic management on freeway). Real-time monitoring of traffic and providing speed data in advance of congestion may replace the need for the driver to see a downstream queue, relying instead on the warning of lower speeds. The roles of technology and other non-geometric measures to address both safety and opera- tional problems and needs is an important emerging trend. Continuous feedback of data, research applications, and increased knowledge will assure that agencies are always applying cost-effective solutions. Such solutions may differ over time from those used in past years, but that should not matter—indeed, it should be welcomed. Over time, agencies may find themselves investing less in physical infrastructure and more in operational solutions, which the geometric design process should facilitate.