Guidelines for Implementing Managed Lanes (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.

36 Design Elements Overview This chapter provides guidance and relevant discussion of a variety of elements that need to be considered in the devel- opment and design of a managed lane facility. While aspects of some of the topics in this chapter are also covered in other chapters (e.g., planning and policy decisions, appropriate traffic control devices), the material in this chapter focuses on the issues that a designer will consider in designing a man- aged lane facility. This chapter covers a variety of topics: • Design considerations for specific user groups. • Geometric design considerations. – Consistency. – Recommended values for specific design elements. – Issues related to conversion from HOV to other types of managed lanes. • Access considerations. – Managed lane placement. – Separation treatments. – Access point design and locations. • Operational impacts. In a number of sections within this chapter, cross-references are provided to other chapters for additional support, such as Chapter 2 (planning), Chapter 4 (traffic control devices), and Chapter 6 (operations and maintenance). Those chapters, while not specifically focused on design, have information and guidance that are relevant to the material found in this chapter. Additional external supporting information for the guidance in this chapter is derived from national sources such as the AASHTO Guide for High-Occupancy Vehicle Facilities (6), state design manuals such as those from California and Nevada (53, 54), and published research that is cited where discussed. User Groups A managed lane facility should be designed with its intended user group(s) in mind. This section will discuss potential user groups that should be considered. Design Vehicle/Eligibility One important consideration when identifying relevant user groups is determining whether those groups have a particular vehicle type that should be accommodated in the design. The following section describes characteristics of design vehicles and considerations for vehicle eligibility. Intended Users To determine the most appropriate design vehicle, one must first identify the intended user group(s) of the facility. A road agency (e.g., DOT or regional authority) can make this decision based on a variety of criteria. Examples include the following: • Selecting the vehicle occupancy or vehicle type that meets a potential management strategy goal such as providing for the maximum movement of people, vehicles, or goods in a corridor. • Providing free-flow operations. • Providing for the ability to accommodate future growth within a corridor. • Being consistent with the regional transportation plans and the policies adopted by the relevant agencies (55). • Where tolling is anticipated, determining restrictions needed for generation of sufficient revenue. Conversely, the design context may rule out certain classes of vehicles. For example, if a managed lane is being created from a left-side shoulder and is likely to have a narrow width C h a p t e r 3

37 and limited sight distance, then large commercial trucks will likely be excluded as a design vehicle. A private developer or a public–private partnership should carefully identify the likely users of the managed lanes, as well as their likely effects on revenue. In both cases, especially for a new facility, the selection of the facility’s user group(s) is related to policy decisions on what type of facility will be developed (e.g., HOV, HOT, bus rapid transit, toll only, truck only) and what purpose the facility is intended to serve; those policy decisions are made during the development of the con- cept of operations before the design process begins in earnest. For an existing facility, the parties responsible may decide that a change in intended users is appropriate. Increasing HOV eligibility from 2+ to 3+ on an existing facility will not neces- sarily change the design vehicle, but if an HOV facility changes to a HOT facility, the appropriate design vehicle may indeed change due in part to expanded vehicle mix or higher volume of design vehicles than previously experienced, which could require making design modifications to the existing facility to accommodate the expanded user group. When the intended users are identified, an informed deci- sion about the appropriate design vehicle can be made. A bus rapid transit or truck-only facility will have a classification of bus or large truck as its design vehicle, while a facility that does not allow large vehicles can accomplish its purpose with a passenger car or light truck as its design vehicle. A common situation is an HOV or HOT facility that allows bus transit vehicles; for that facility, consideration must be given to the accommodation of buses, particularly in the design of access points, acceleration and deceleration lanes, lane and shoulder widths, and horizontal curves. Accommodation for identification of different users within the managed lane traffic stream must be considered in the design of the facility. For example, correct identification of HOVs and SOVs determines which vehicles are assessed a toll for using a HOT facility. Similar considerations could be made for bus rapid transit facilities in which public transit buses travel at no charge but other buses might be charged a toll. Ease of intended user identification may stabilize user volume forecasts, reduce enforcement infrastructure, and otherwise provide better support for a particular managed lane strategy; additional discussion of vehicle identification is provided in Chapter 6. Design Characteristics The design vehicle should be used to establish the geometry of the facility’s design elements. For example, intersection radii, where appropriate, such as off-ramp connections with arterials, should be based on a bus or other large design vehicle, while alignment geometry is typically based on the stopping sight distance of a passenger car driver. The speed and braking char- acteristics of a bus should be considered for sustained grades because they may exceed those of a passenger car (6). If the facility will be open to truck traffic, the entire facility, includ- ing all access points and horizontal curvature, is designed for the semitrailer truck design vehicle. For example, it is gener- ally impractical to design long segments of the facility based on the semitrailer design vehicle and yet have select interchanges along the facility that do not accommodate that vehicle. It is possible that truck restrictions could be applied to selected loca- tions within a corridor, but it should be done on a case-by-case basis and should be a result of a conscious decision within the design process after considering design alternatives and their trade-offs. In general, design criteria for managed lane facilities that are part of general-purpose freeway corridors are the same as for the general-purpose lanes. These criteria apply to vertical and horizontal alignment, cross-slope, and lane and shoul- der widths in particular. When proposed design elements do not meet these criteria, approval for a design exception or deviation process is typically required, although details vary among jurisdictions (55). For stand-alone facilities that are in their own corridor (e.g., not adjoining a general-purpose freeway), the mini- mum criteria should be based on the intended design speed (i.e., typically freeway or arterial speeds similar to a general- purpose facility that would serve a user group with the same design vehicle). The benefit to designing a facility in a new controlled-access corridor rather than retrofitting a man- aged facility into an existing freeway corridor is that there are typically fewer restrictions within the available right- of-way and there is much more likelihood that the desired criteria can be accommodated. The intended purpose of a facility will have some effect on the selection of certain design criteria, particularly for the design of access as well as any toll collection accommo- dation that needs to be made. Primary considerations are cross-sectional elements such as lane and shoulder width, buffer width, median width, drainage, cross-slope, and traffic control device accommodation, particularly if the facility is separated by buffers, pylons, or barriers. Other important design elements include sight distance, horizon- tal curve design, and vertical clearance under toll gantries, sign bridges, and overpasses. More detailed discussions of specific design elements are provided in the following sec- tions of this chapter. Transit Considerations Transit vehicles (i.e., buses) have unique characteristics that should be considered if a managed lane facility is primarily intended for transit use. Buses operating in managed lanes can increase the productivity of the lanes in terms of person-trips

38 completed. When bus volumes are high, a bus-only lane might be desirable. The following section describes characteristics of transit vehicles for the design of managed lanes. Design Vehicle Considerations When a facility’s intended design vehicle includes buses, as described elsewhere in this chapter, several design consider- ations should be made in addition to those used for managed lanes that primarily serve passenger vehicles. The design of managed lanes influences (and can be influenced by) the types of transit service that exist and can be provided. Each type of managed lane as a minimum has access to general-purpose lanes at its terminals. Intermediate- access points are often provided if the facility is not open to access throughout its length. For restricted-access designs, direct (i.e., grade-separated) access connections for major transit services, such as bus terminals, park-and-ride lots, and city streets, may be provided to eliminate the need for buses to weave across the general-purpose lanes and there- fore reduce the potential for conflict with other freeway traffic. Design guidance in Nevada states that bus turning radii are perhaps the most significant design vehicle issue for managed lanes since each bus type has a different turning radius (54). Turning requirements are particularly critical for applications of direct-access ramps from transit facilities or local roads, or where turning at low speeds is required. For single-lane facilities, visibility is adversely affected in a single platoon of vehicles; sight distance restrictions caused by the mix of buses and lower-profile vehicles may reduce vehicle head- ways, and thus affect the lane’s operational performance measures. Since most managed lanes are located next to the median, the horizontal separation (or “shy” distance) to the median barrier, where full shoulders cannot otherwise be provided, may adversely affect sight distance around left- hand curves. Managed lane planning and design should reflect the char- acteristics and capabilities that are in service. Examples of design vehicles and performance characteristics are given in the AASHTO Green Book (27); NCHRP Report 414: HOV Systems Manual (3); Volume 2 of TCRP Report 90: Bus Rapid Transit (56); and TCRP Report 117: Design, Operation, and Safety of At-Grade Crossings of Exclusive Busways (57). Location of Transit Facilities Transit stations along managed lanes, or in their immedi- ate environs, help improve transit ridership and improve the productivity of the lanes in terms of passengers or passenger miles of travel carried. The stations can be located online (as along the I-110 Harbor Freeway in Los Angeles) or offline (as along freeways in Houston) or both (as along freeways in Minneapolis and Seattle). The buses using the managed lanes connect with various collection and distribution points and may operate point-to-point service between transit facilities or link service along a route. Preferential treatments may be provided to off-street terminals or via arterial bus lanes; refer to the AASHTO Guide for Geometric Design of Transit Facili- ties on Highways and Streets (28) for more information on arterial-based facilities. The design of online and offline bus stations along managed lanes should consider the following: 1. Locating stations adjacent to major activity concentrations, park-and-ride facilities, and interchanging bus lanes. 2. Maintaining the speed, reliability, and safety of the man- aged lanes by providing adequate spacing between stops, acceleration and deceleration lanes, and separate lanes for buses entering, exiting, and stopping at stations. 3. Providing adequate station platform capacity that conforms to Americans with Disabilities Act (ADA) requirements and prevents passenger overcrowding of platforms, which can be accomplished by: – Providing enough pedestrian access capacity to allow departing passengers to clear station platforms before the next bus arrives. – Providing attractive and convenient pedestrian con- nections to major activity concentrations adjacent to the managed lanes. – Utilizing barrier-free designs that meet ADA requirements. Online stations are located along or adjacent to managed lanes either within the freeway right-of-way or in an exclu- sive right-of-way outside of the nominal freeway right-of-way envelope. The two platform orientations (side and center) are shown in Figure 28. Both of these orientations allow managed lanes to bypass the platform area. In both cases, the right-of- way must be sufficiently wide enough to accommodate the station and pedestrian access to it. Center island platforms are used to minimize platform space since both directions share a common area for patrons that is removed as much as possible from freeway operations. They require only one set of vertical pedestrian access points. Because doors on buses are usually on the right-hand side, the center platforms require channelized crossovers on each end for bus drivers to negotiate. While it is possible to obtain bus designs with left-side doors, few operators currently operate this design because it complicates their fleet mix and requires dedicating specific buses to this facility. The crossover movement has to be taken at low speed and is less than desirable from a traffic operations perspective. Side platforms eliminate the need for bus crossovers. How- ever, they require two sets of vertical patron access points, they place patrons closer to high-speed traffic movements

39 (necessitating walls or screens), and they require more width. Both orientations can be found in freeway settings. Offline stations are sometimes provided to serve as col- lection and distribution points (e.g., transit centers, inter- modal centers, and/or parking facilities) that are located near the managed lanes. The stations can be incorporated into other land uses as part of parking facility design. Access to and from the managed lanes is either along local street connections or via direct ramps. The station may be some- what removed from the freeway corridor, or the station may be adjacent to the freeway right-of-way. Offline transit sta- tions often provide connections to a parking facility and rely on the local street network. Details on the design of transit stations, park-and-ride facilities, and other transit features not specifically related to the managed lanes can be found in other guidance documents (28, 56, 58, 59). Truck Considerations (Including Freight/ Truck-Only Facilities) Similar to transit vehicles, large trucks have unique char- acteristics that should be considered if a managed lane facil- ity is intended to serve primarily or only trucks and freight vehicles. The following section describes characteristics of trucks for the design of managed lanes. Design Vehicle Considerations If a managed lane facility’s primary design is to serve trucks, there are design considerations that should be made in addi- tion to those used for facilities that primarily serve passen- ger vehicles. Nevada’s design guidance states that the design of managed lane facilities must account for trucks from the outset (54). Nevada’s reasoning is that if large trucks are not accommodated in the initial design of managed lanes, trying to add those accommodations to an existing facility, particu- larly a physically constrained corridor, can adversely affect the project’s cost-effectiveness and ability to meet all design vehicle requirements. Highway designers must make design trade-off decisions routinely on projects as they are developed within the con- text of the surrounding environment, particularly for those facilities that are retrofitted into existing freeways, which may limit the type of vehicles that can be accommodated in the facility (60). Trucks generally require wider or longer geometric design features than do passenger vehicles. This need is attributed to trucks having longer wheelbases and greater minimum turning radii. Trucks are one of four general classes of design vehicles described in the Green Book (27). Within the truck class are the following seven design vehicles: Source: Fuhs (59). Figure 28. Platform orientations for online transit stations.

40 • Single-unit truck—SU. • Intermediate semitrailer—WB-40 (40-ft wheelbase). • Interstate semitrailer—WB-62 (62-ft wheelbase). • Interstate semitrailer—WB-67 (67-ft wheelbase). • Double-trailer combination—WB-67D (67-ft wheelbase). • Triple-trailer combination—WB-100T (100-ft wheelbase). • Turnpike-double combination—WB-109D (109-ft wheel- base). The Green Book states that the interstate semitrailer WB-67 should generally be the minimum-size design vehi- cle for consideration at freeway ramp terminals intersect- ing with frontage roads or arterials on routes that provide access for trucks. Key design characteristics of the WB-67 and other design vehicles can be found in Chapter 2 of the Green Book (27). Designers must consider horizontal and vertical clearance requirements for trucks. Horizontal clearance includes suffi- cient lane widths to account for possible offtracking on curves and shoulder widths to allow trucks to completely clear the lane in the event of a breakdown or emergency. Higher verti- cal clearance is necessary for trucks to safely pass under toll gantries, sign bridges, and overpasses. Access points require longer weaving areas and speed-change lanes for trucks than for passenger vehicles. Because the performance characteristics of trucks are dif- ferent from passenger vehicles, the designer should consider effects of grades on facilities with trucks as design vehicles. Prolonged grades should be avoided if possible because upgrades can lead to substantial speed differentials between trucks and passenger vehicles, and downgrades may require additional stopping sight distance for trucks. The location of the managed lane facility should also be considered. A facility that is in proximity to a hub for com- mercial or industrial access needs a design that accounts for increased truck traffic. While the truck percentage of total traffic at a typical facility might commonly be less than 10%, such a facility near a shipping port or a distribution center could have a truck percentage as high as 50%. The design of this type of facility needs to account for varia- tions in strength of pavement, lane width, turning radius, vertical clearance, and other characteristics to ensure that the facility meets the needs of the user group for which it is intended. Geometric Design Considerations This section describes key geometric design elements and related issues that should be considered in the design process for a managed lane facility. Some of the information presented here is applicable to freeway design in general, while other portions contain guidance specific to managed lanes. Design and Operational Consistency Whether a general-purpose facility or a managed lane facility, the geometric design should be based on principles that are appropriate for a freeway and are generally consis- tent with other freeways. This reinforces driver expectancy and promotes system connectivity when traveling from one facility to another. System Integration Considerations Managed lanes are often constructed as facilities that are dedicated lanes within a freeway or in some cases a freeway within a freeway. They are generally designed to appropriate state or national standards for the respective class of roadway. Managed lane design should match driver expectancy estab- lished by the presence of design elements and traffic control devices on nearby facilities. For example, if the adjacent road- way provides 50-mph ramps at major interchanges, a similar design should be considered for a managed lane ramp. Research (61) has identified ramps as one of the most critical geometric aspects to consider in making a managed lane facil- ity safe and functional with other facilities. In particular, the important aspects to consider in the geometric design of the ramps are ramp orientation (left versus right), type, and spac- ing. These guidelines typically vary by traffic level and appro- priate design speed, so understanding both the current and future traffic impact of the facility is important to ensure geo- metric adequacy both at the time of construction and in future years of operation. Providing a design that is flexible enough to accommodate all design vehicles is best because experience has shown that a number of bus- or HOV-only facilities have over time been converted to serve a wider mix of design vehicles than was envisioned in the original design. For example, some HOV access ramps have low design speeds and reverse curves because they were designed for experienced bus drivers, but they now serve all traffic. Access to and from managed facili- ties should be designed so that drivers can travel from origin to destination as seamlessly as possible. Details of particular types of access ramps are discussed later in this chapter. Other aspects of geometric interoperability include estab- lishing techniques for separation of managed lanes with adja- cent general-purpose lanes and consistent orientation and treatment for access, as well as considering within the design how enforcement, incident management, and maintenance of traffic control devices and tolling systems may be performed. For example, concurrent managed lanes are located adjacent to the median to best facilitate long-distance, high-speed opera- tion, and even though some details of design specifications may change as new information is gained, drivers traveling from one managed facility to another can anticipate and expect that all such facilities in a region are constructed in a similar

41 manner. Motorists benefit not only from having consistency between a managed lane and the adjacent infrastructure but also from having design consistency across all facilities within the region or area. Applying design consistency to a corridor or regional system involves a number of factors, including an appreciation of what will fit and how traffic will operate. Pref- erences in design practice can change over time, and managed lane systems can take many decades to implement. A perspec- tive from one design practitioner in California (67) offers a way of appreciating how consistency should be treated: The whole evolution of what it means to be consistent is a question. The express lane project currently in development has all four . . . design options in play. The temptation is to say this is pretty inconsistent. From a driver’s perspective what is going on? In one corridor this mix of designs may be inconsistent. But if we are sticking to the overall express lane concept, we are consistent with local standards of practice. The consistency becomes greater with a greater set of projects. But at a local level it might not look this way. There is this balance between what people consider is consistent and what the operational characteristics tell you will work or won’t work. What is a good standard? It is not uniformity throughout. Every design concept requires you to adjust to changes in traffic behavior. Local, Regional, and State Guidelines In many cases, the local and regional jurisdictions that gov- ern the design and construction of managed lane facilities do not have their own design guidelines, but rather they fol- low the design guidelines established by their respective state DOTs. However, the designer (e.g., government agency, private partner, design consultant) needs to know whether there are applicable manuals, standards, guidelines, and unpublished preferred design practice at the local or regional level that must be consulted and applied. If there is a local/regional design guidance policy that differs from a corresponding state policy, or if options exist, the designer needs to know which guidance or preference supersedes the other. A number of state DOTs have design guidelines for some types of managed lane facilities (e.g., HOV and HOT lanes). Among the states with the most extensive guidelines are Cali- fornia (53, 62), Nevada (54, 63), Texas (4, 64), and Washington (55, 65). These guidelines cover fundamental geometric ele- ments such as lane and shoulder width, horizontal and vertical alignment, and access design; some of them also provide guid- ance on specific features such as HOV bypass lanes on entrance ramps, enforcement treatments, and emergency/breakdown areas. It should be noted that the managed lane guideline doc- uments referenced here typically contain numerous references to their more general counterparts (e.g., Roadway Design Man- ual in Texas, Highway Design Manual in California). Those design manuals provide the broader design basis on which the more specialized managed lane guidance is supplemental. In addition, some states have their own manual on uniform traf- fic control devices that contains guidance and standards on the appropriate signs, markings, and other devices specific to man- aged lanes (see Chapter 4). The designer needs to be familiar with the appropriate guidance documents that apply to the state in which the managed lane facility will be located. Applicable policies change over time, so a set of guidelines that applied to a particular project may not completely apply to a subsequent project on the same corridor. In that sense, a measure of apparent inconsistency can arise from the driver’s perspective (e.g., if the required minimum buffer width changes from one project to the next), but the consistency in practice still applies since the most current applicable guide- lines are used at the time of each project. National Guidelines If no local, regional, or statewide guidance exists, the designer should consider using national guidance in designing managed lanes. In fact, the local/regional guidance may actually be to use national guidance. In either case, this guide and other national references should be useful to the designer: • AASHTO Guide for High-Occupancy Vehicle Facilities (6). • FHWA Manual on Uniform Traffic Control Devices (1). • FHWA Priced Managed Lane Guide (2). Design Variances and Flexible Design Philosophies Particularly for retrofit facilities, trade-offs and accommo- dations may need to be made in comparison to applica- ble guidelines for affected design elements (e.g., lane width, shoulder width, lateral clearance). In many cases, a waiver, design exception, or deviation process must be completed to construct a facility that incorporates such trade-offs, though details vary among jurisdictions (55). The designer needs to know what trade-offs to consider, and what agency protocols must be followed to approve design variances. An example from the Nevada DOT of a series of possible trade-offs for designing the cross section for a concurrent flow HOV lane is shown in Table 4. Note that this list is provided as an example and is not intended to supersede engineering judgment in specific design applications. As more states adopt policies of performance-based practi- cal design (PBPD) or implement similar programs, the design philosophy of managed lane projects in those states may be more focused on how the completed project meets a stated purpose and need and fulfills the goals that were defined for the project. Indeed, setting specific performance measures (operational, economic, etc.) is a common practice for man- aged lane facilities, so implementation of PBPD principles may be a complementary strategy to the design process used on previous projects. The designer should be familiar with

42 applicable PBPD-related policies and programs that are in place within the jurisdiction of the project. More informa- tion on PBPD principles and related resources can be found through FHWA (66). Design Speed The Green Book guidance for freeway design speed, or appropriate state or local standards if applicable, should be used to provide for a high level of service. For managed lane facilities that are part of a general-purpose freeway corridor, it is preferable to use a design speed that is comparable to the adjoining freeway. This is especially true if there is the pos- sibility of the facility being used by general-purpose traffic during the off-peak hours or at some time in the future. If there is no possibility that the facility will be used by general- purpose traffic, and if use of this facility is to be further lim- ited to a single vehicle type such as buses, then the specific physical dimensions and operating characteristics of that vehicle should be considered in design. For example, the dif- ference in driver eye height or braking characteristics may require a different roadway geometry than if the facility were to be used by all vehicle types. The Green Book recommends design speeds of 60 mph to 70 mph on most urban freeways (27). This information is provided to give a general idea of potential design speeds. The design speed for a specific facility should consider the antici- pated user groups, the application of transit facilities, gradients, and local conditions and requirements. According to AASHTO (6), desirable ramp design speeds should approximate low-volume running speed on the inter- secting highway. Using the middle range of ramp design speeds in the Green Book (27), for a 70-mph mainline design speed, minimum ramp design speed is 50 mph. For a 50-mph main- line design speed, minimum ramp design speed is 35 mph. For a direct-connect ramp, minimum design speed is 40 mph. A suggested (59) design speed for ramp connections for HOV lanes is approximately 0.7 times the mainline design speed, or nominally in the 35-mph to 50-mph speed range. This criterion is applicable to flyover ramps and connecting drop ramps with local streets. At-grade access locations may use this criterion if dedicated weave lanes are provided, or they may be designed at a higher speed based on the specific location and operating characteristics of the freeway through lanes. Some managed lane connections use grade-separated traffic intersections to connect directly with the local street network; such intersections require lower design speeds for turning maneuvers and good sight distance due to their median ori- entation. Adequate acceleration and deceleration lane lengths should be incorporated at these intersections for speed tran- sition. Lower ramp design speeds may also be appropriate where restrictive geometry or right-of-way exist for connec- tions; however, a negative trade-off is that travel time savings are reduced (59). Stopping sight distance is often based on a passenger car as the design vehicle and is a function of the driver perception– reaction time and the driver comfortable deceleration rate. It is generally considered that the additional stopping sight dis- tance a truck may need for slower deceleration rates is offset by the increased driver eye height; however, on downgrades, the added momentum of trucks compared to passenger cars makes it prudent to provide additional stopping sight dis- tance. In all cases, stopping sight distance should be based on a national guideline, such as the Green Book, or applicable state or local standards. These guidelines should be applied with appropriate engineering design knowledge and judgment. Cross Section and Alignment There are multiple sources for design specifications. In many cases, the specifications for general-purpose lanes on a particular facility type apply to managed lanes. The following Suggested Sequence Cross-Section Design Change First Reduce the 14-ft median (left) shoulder (for continuous enforcement) to 8 ft. Provide designated enforcement areas instead. Second Reduce median shoulder to typical minimum width as per Nevada DOT Standard Plans. Third Reduce median shoulder to 2 ft.* Fourth Reduce outside (right) shoulder to typical minimum width as per Nevada DOT Standard Plans. Fifth Reduce managed lane to 11 ft.* Sixth Reduce general-purpose lanes to 11 ft starting from left and moving to the right as needed. The outside lane should remain 12 ft.* Seventh Transition barrier shape at columns to vertical face. *Requires design exception. Source: Nevada DOT (54). Table 4. Suggested design sequence of trade-offs for concurrent flow HOV lanes.

43 information is provided for general reference only. The practi- tioner should determine what national, state, or local standard applies to a particular facility. Cross-Section Elements Key elements of the managed lane cross section, as well as their effects on managed lane operations, are described in the following list: • Number of lanes—Most managed lanes attempt to fit within available right-of-way, which is often quite con- strained. For a majority of projects, this retrofit means adding one lane in each direction. Very few corridors have been able to retrofit two-directional lanes. Planning studies will help confirm the number of through lanes, par- ticularly taking into account system-level demand inputs and outputs for any regional network. For managed lane facilities that are retrofitted into general-purpose freeway corridors, it may be difficult to accommodate one or two managed lanes in each direction and still provide the neces- sary shoulders and separation within the cross section. As a minimum, the same attention is needed for consideration of auxiliary lanes or treatment of lane drops downstream of any ramps that add high traffic volumes. The designer will have to consider the constraints of the corridor and then consider trade-offs to provide the best possible balance of managed lanes, shoulders, and separation in conjunction with the lanes and shoulders in the general-purpose facil- ity. Some state DOTs (53, 54, 55) have developed their own process or preferences for making these trade-offs based on managed lane experiences. • Width of lanes—The recommended lane width for a man- aged lane facility is 12 ft, just as with any other freeway lane. For a managed lane facility that includes a buffer that is 2 to 4 ft in width, some design guidelines allow up to 1 ft of lane width to be considered as part of this overall buffer dimen- sion where full buffers and lane widths will not fit. This is because the managed lane driver perceives this space as part of the overall width used for horizontal sight distance. A lateral position study (67) modeled different scenarios to investigate whether it is better to have a 12-ft lane and 1-ft buffer or an 11-ft lane and 2-ft buffer. The predicted lateral position showed that drivers position the vehicle closer to the center of the lane when the extra foot is given to the buf- fer rather than the managed lane. The practice of reducing the lane width by 1 ft (from 12 ft to 11 ft) and providing that foot of width to the buffer between the managed lanes and the general-purpose lanes or to the inside shoulder is often appropriate. Figure 29 shows an example of cross-section options, including 12-ft lane widths, for barrier-separated, two-way HOV lane facilities. The AASHTO Highway Safety Manual (23) provides information on safety trade-offs by lane width for freeways. • Width of shoulders—Ideally, managed lane shoulders should follow state DOT preferences, typically 10 to 12 ft, to provide ample clearance for disabled vehicles, enforce- ment activities, and emergency response. In practice, many retrofit applications have narrower shoulders so that the lane width is maintained and, to the extent possible, a shoulder width of at least 8 ft is maintained. For concurrent directional lanes, shoulders should be to the left, next to the median barrier. Shoulder space with a width between 4 and 8 ft should not be striped because drivers on a concurrent flow lane can misconstrue the space as wide enough to stop in, leaving their vehicles exposed to high-speed traffic that may not be able to negotiate around the exposed vehicles. Figure 29 shows an example of shoulder width options for barrier-separated, two-way HOV lane facilities. Where cross-section trade-offs result in a reduced-width inside shoulder, the operator of the facility can use strategies to mitigate related losses in sight distance that include spot widening, striping to create wider offsets in affected curves, and drainage treatments that reduce ponding. The Highway Safety Manual (23) and a field study on lateral position (67) indicate that maintenance of adequate shoulder width is desirable. The field study results demonstrated that vehicles traveling on a segment with a minimal shoulder on the left will shift closer to the right lane line, possibly due to limited sight distance and a driver’s concern to avoid the concrete barrier. Research findings indicate that a 1-ft reduction in shoulder width results in greater changes in lateral position when the shoulder width is near minimal values (e.g., 2 or 4 ft) as compared to when the shoulder width is near desirable (e.g., 8 ft or 12 ft). The study findings sup- port the idea that the buffer width should not exceed the shoulder width. In addition, it indicates that if extra space is available, it should be used for the shoulder rather than the buffer. • Width of median—Medians on managed lane facilities should be designed in the same manner as other freeway medians. When right-of-way is constrained, the median should be designed with sufficient width to accommodate the pedestals and bases for gantries and traffic control devices, the median barriers, and the enforcement activi- ties that will be required for the facility so that vehicles can travel in the leftmost lane or the left shoulder without their path being infringed upon by these or other required ele- ments in the median. • Width of buffer and buffer separation—A lateral posi- tion study (67) yielded the observation that managed lane drivers shifted away from the pylons placed in the buffer; however, a wider buffer can offset the impacts on lateral position.

44 In addition to the cross-section elements described in the previous list, there are additional design elements that are also considered in context as the design of the managed lane facility is developed. A selection of those elements and their effects on managed lane operations are listed below: • Horizontal alignment—The radius of horizontal curvature used in a particular roadway design is a function of design speed, rate of superelevation, and side friction with prac- tical limits due to right-of-way constraints (6). As design speed increases and the rate of superelevation decreases, the minimum radius of horizontal curvature required increases. Horizontal and vertical alignment should not be designed independently. A lateral position study (67) showed that horizontal alignment (tangent or curve) and direction of the horizontal curve (left or right) were influential on lateral position. The impact on lateral position was greater within the minimal values for shoulder, lane, and buffer widths. Drivers were closer to the left edgeline when on a curve to the left and farther from the left edgeline when on a curve to the right depending on assumed shoulder, lane, and buffer widths. • Vertical alignment—Basing the minimum lengths of crest vertical curves (and the rate of vertical curvature) on Source: Adapted from AASHTO (6), Figure 3-3. Figure 29. Examples of cross sections for barrier-separated, two-way HOV lane facilities.

45 stopping sight distance criteria is usually sufficient from the viewpoint of comfort and appearance. For sag curves, the use of the stopping sight distance criteria for establish- ing minimum rates of vertical curvature is recommended. Horizontal and vertical alignment should not be designed independently (6). • Superelevation—As with other facilities, the rate of super- elevation used in a particular roadway design will be a func- tion of design speed, radius of curvature, and side friction with practical limits based on driver comfort, safety, cli- mate, and local agency guidelines. As design speed increases and radius of curve decreases, the need for superelevation increases. In an urban environment, a maximum super- elevation rate (emax) of 4% to 6% is common practice, but the applicable state design manual should be consulted to confirm what is expected for a particular location. Super- elevations are typically projected for managed lane widen- ings. These projections can result in bridge structure conflicts where the resulting vertical clearance on the low end of the projection is compromised. In most cases the widening is modest enough that different structural approaches (such as box beams instead of “I” beams) are employed to hold the vertical clearance constant. • Horizontal clearance—For facilities within other freeways, the horizontal clearance may be affected by right-of-way constraints, but it should be sufficient to prevent the design vehicle from colliding with signs, guardrails, and other roadside apparatus. For stand-alone facilities, the guidelines in the AASHTO Roadside Design Guide (68) for freeways are applicable. • Vertical clearance—The appropriate vertical clearance depends on the design vehicle. The facility should have the necessary vertical clearance for the design vehicle to pass under toll gantries, sign bridges, and overpasses with- out striking them. It is recommended that for facilities within other freeways, the vertical clearance provided be the same as that provided on the general-purpose facility. The AASHTO HOV Guide (6) and the AASHTO Green Book (27) state that desirable clearance is 16 ft. Where this would be cost prohibitive in highly developed urban areas with existing structures, existing clearances may be allowed; the Green Book states that a minimum clearance of 14 ft may be used if there is an alternate freeway facil- ity with the 16-ft clearance. Allowance should be made for future resurfacing, and additional clearance should be provided to structures with lesser resistance to impacts (e.g., sign trusses, pedestrian overpasses, and cross-bracing of through-truss structures); vertical clearance for these structures should be 17 ft or, on urban routes with less than 16-ft clearance, 1 ft more than the minimum clear- ance for other structures. • Cross-slope and drainage—The cross-slope of a managed lane should generally follow the adjacent freeway, which is commonly 2%. However, adding managed lane facili- ties always expands the drainage run-off unless the prior space constituted a shoulder with impermeable surface. The added drainage caused by a longer crossfall presents a challenge to designers. Some states allow for extending the crossfall and adding specialized drainage treatments on the right (such as trench drains at ramp tapers), while others prefer to reverse the crossfall to the median barrier and break it in the middle of the managed lane or in the buffer area. In a few instances, supplemental drainage is placed in a wide buffer between parallel concurrent roadways, or barriers are employed with drainage provi- sions. Reversing cross-slope (i.e., creating a cross-slope break of greater than 4%) except along extremely wide buffer or barrier alignments is not desirable because it can affect driver expectations when crossing the buffer at designated access points and possibly degrade operational performance and safety. The AASHTO Green Book indi- cates that for intense rainfall areas, the cross-slope can be increased to 2.5%, resulting in a crossfall break of 5%. For a barrier-separated facility, drainage design under the barrier depends on the number of general-purpose lanes, cross-slope, and superelevation transitions to min- imize hydroplaning. Shoulders should have a cross-slope at least 1% steeper than the adjacent travel lanes. Paved shoulders typically have a cross-slope between 2% and 6%. The Green Book indicates that the algebraic differ- ence between the cross-slopes of a traveled lane and the adjacent shoulder should be below 8% at all points along the facility (27). Each decision about cross section needs to be considered in the context of the other design elements and its potential effects on safety, operations, and maintenance. Table 5 describes some of the advantages and disadvantages to changes in cross sec- tion that should be considered; the designer should have an appreciation for the specific effects of cross-section changes that apply to a particular facility during the design process, and developing a list similar to Table 5 could be beneficial in that process. Operational Effects of Cross Section for Weather Events, Special Events, Enforcement, and Maintenance In addition to the items mentioned in the previous sec- tion, the effects of cross section on operations during main- tenance activities—whether scheduled or due to special events, particularly weather events—must be considered. Additional discussion on some of these effects can be found

46 in their own respective sections of this and other chapters, but a selection of items are compiled here to provide examples of operations that can be affected by reduced lane or shoulder widths: • Any hardware or equipment placed in the median (where there is no full-width shoulder or built-in maintenance access) or overhead will require lane closures for access to complete maintenance, which equates to a loss of opera- tional time, particularly as the equipment ages (for more information, see the section later in this chapter on providing for associated equipment and devices). • Reduced lane widths next to narrow or non-existent shoulders could reduce or eliminate the potential for enforcement activities (see the section on enforcement pullouts elsewhere in this chapter). • Minimal or non-existent shoulders can negatively affect sight distance in curves, which can affect crash rate, crash frequency, and operational time of the managed lane, as well as operations in the adjacent general-purpose lane (see the discussion on characteristics of large design vehicles elsewhere in this chapter). • Narrow shoulders will also create a higher potential for clo- sure due to flooding, snow melt, and icing, particularly on barrier-separated or pylon-separated facilities that prevent moving the water or snow to a non-traveled portion of the right-of-way. Narrow shoulders limit their capability for storing snow removed from the travel lanes, which affects Design Element Pros Cons 12-ft travel lane • Matches typical guideline for general-purpose travel lanes on freeways • Ample for accommodation of buses and trucks • May be difficult to achieve in a retrofit without taking width from other elements in the managed lane or general-purpose facilities 12-ft shoulder • Matches typical guideline for general-purpose travel lanes on freeways • Provides refuge space for disabled vehicles • Provides space for enforcement activities • Provides space for temporary storage of snow removed from the travel lane • Provides option for throughput during incidents in the travel lane • May be difficult to achieve in a retrofit without taking width from other elements in the managed lane or general-purpose facilities 11-ft travel lane • Easier to provide in a retrofit than 12-ft lane, though there are safety trade-offs • May be acceptable if associated with a wider buffer • Narrower than typical freeway lane • Less-than-minimum width in some guidelines; not always allowed or recommended • Associated with higher number of crashes (23) 10- or 11-ft shoulder • Can accommodate passenger vehicles and most heavy vehicles • Easier to provide in a retrofit than 12-ft shoulder • Narrower than typical freeway shoulder • Reduces the available space to store and/or attend to disabled vehicles and to store snow • Restricts ability of enforcement officers to conduct activities outside of the travel lane 8- to 9-ft shoulder • Can accommodate most passenger vehicles • Easier to provide in a retrofit than 12-ft shoulder • Not suitable for heavy vehicles • Very restrictive for refuge and enforcement • Minimizes usefulness in snow storage and incident management Shoulder less than 8 ft • Provides a measure of lateral clearance for drivers compared to no shoulder • Not suitable for heavy vehicles • Not wide enough to store a passenger vehicle without encroaching on the travel lane • Not suitable for maintenance, enforcement, snow storage, or incident management • Can restrict sight distance in curves • Associated with higher number of crashes (23) 4-ft or wider buffer • Provides separation between managed lanes and general-purpose lanes • Provides additional accommodation for vehicles to shift laterally within the lane in horizontal curves • May be difficult to achieve in a retrofit without taking width from other elements in the managed lane or general-purpose facilities • Depending on width, may be seen as an additional travel lane for passenger vehicles or motorcycles without additional delineation or separation devices Buffer less than 4 ft • Provides a measure of separation between managed lanes and general-purpose lanes • Easier to provide in a retrofit than a wider buffer • Reduces distance between traffic streams that could be traveling at greatly different speeds • May not be allowed in some jurisdictions Source: Texas A&M Transportation Institute (TTI). Table 5. Pros and cons of managed lane facility cross-section trade-offs.

47 operating conditions on the facility (see sections on buffer separation, pylon separation, and emergency refuge areas, as well as Table 5). • In areas where managed lanes are in use all day with high volumes, there may not be an off-peak for maintenance and reconstruction. With narrow lanes and shoulders, maintenance on such facilities would require closures for sweeping, patching surfaces, applying pavement markings, and performing other scheduled maintenance activities. Managed Lane Orientation with Respect to General-Purpose Lanes The AASHTO Guide for High-Occupancy Vehicle Facilities (6) contains guidance on the placement or location of an HOV facility with respect to general-purpose lanes (or main lanes). Many of the principles described in the AASHTO HOV Guide may apply to other types of managed lane facilities and are dis- cussed here along with other guidance applicable to managed lanes in general. Location of Managed Lanes with Respect to Median Locating managed lanes next to the median or median bar- rier of a freeway is the most common practice, found in an estimated 95% of existing facilities (6). Managed lanes located next to the median have a variety of alternative cross sections, whether it is the type of separation used between the managed lanes and general-purpose lanes or the way in which traffic is assigned to the managed lanes themselves. Facilities constructed next to the median are within the free- way right-of-way and are separated from the general-purpose freeway lanes by either barriers or buffers (with and without pylons) or by broken- or solid-line pavement markings (see Chapter 4 for more details on separation pavement markings). Barrier-separated facilities may be two-way facilities (i.e., the facility operates in two directions all the time) or reversible facilities (i.e., the facility operates inbound toward the central business district or other major activity center in the morn- ing and outbound in the afternoon). Facilities not physically separated from the general-purpose traffic lanes (i.e., separated only by pavement markings or buffers, with or without pylons, with lanes operating in the same direction as general-purpose traffic) are described as concurrent flow lanes (6). Managed lanes adjacent to the median may be open for all or a portion of a day. Further discussion of separation treat- ments is provided elsewhere in this chapter. Part-Time Shoulder Use Part-time shoulder use (also called hard shoulder run- ning or dynamic shoulder use) is a specific type of managed lane strategy where a shoulder (usually, but not always, the right shoulder of the freeway) is temporarily used as a condi- tional travel lane, typically during peak travel times and/or in response to incidents or other conditions. Several components need to be considered when imple- menting part-time shoulder use: • Operating conditions must be established and clearly com- municated for determining when the shoulder will be opened and closed to traffic (e.g., time of day, level of ser- vice) and what kind of traffic will be allowed (e.g., buses only, passenger cars only). • The operations plan must have provisions for accommo- dating traffic at freeway access points (i.e., entrance and exit ramps) so that entering and exiting traffic can complete their access maneuvers. • The shoulder must have sufficient lateral (i.e., horizontal) clearance so that roadside apparatus (e.g., signs, guard- rails) are not struck by vehicles traveling on the shoulder. If the available lateral clearance is less than applicable state or local guidelines require, consideration must be given to approving variances or making other accommodations. • The shoulder must be sufficiently strong to withstand the impacts of the increased level of traffic, particularly if buses are considered as eligible users. • Emergency refuge areas, or pullouts, must be planned so that disabled vehicles and emergency responders have a place outside of the traveled way to stop. Traditionally, part-time shoulder use has been a strategy used in Europe (69), and there is an increasing number of applications in the United States. Minnesota and Virginia are two states with some of the earliest applications of part-time shoulder use in the United States, and guidance for imple- menting part-time shoulder use in other locations was under- way at the time these guidelines were written. Practitioners are encouraged to review the forthcoming Guidebook on Plan- ning and Evaluating Active Traffic Management Strategies from NCHRP Project 3-114 or FHWA’s Use of Freeway Shoulders for Travel (70) for more information on part-time shoulder use. Contraflow Lanes Another strategy is the use of contraflow lanes, whereby a freeway lane in the off-peak direction of travel is reconfigured to allow use by eligible vehicles traveling in the peak direction. The lane is physically separated by movable concrete barriers or pylons from the off-peak-direction general-purpose lanes. Existing freeway contraflow lanes use the off-peak median general-purpose lane, and they operate only during the peak periods. Some operate only during the morning peak period. During other times of the day, the lanes revert to normal use in the direction of travel of the general-purpose lanes. Some contraflow lanes are open to buses only, and others are open to buses and HOVs (6). When developing contraflow lanes,

48 it is important to consider the peak-hour directional split to ensure that sufficient capacity is available in the off-peak direction to allow one lane from the off-peak direction to be shifted to the peak direction without creating congestion in the off-peak direction. Separation Between Managed Lane and General-Purpose Lanes There are multiple ways to provide separation between managed lanes and the general-purpose lanes of a freeway. Each approach reflects different operational goals and needs. While all can be effective, the selection of separation type should take into account the design setting, business rules, design vehicles, and level of anticipated demand. Barriers, buffers, pylons, pavement markings, and separate facilities are all used, and some facilities may have a combination of separation types (e.g., a mixture of buffer separation and pavement marking separation where cross section is limited). Each option has its own potential advantages and disadvan- tages, and these should be considered early in the design of the managed lane facility to determine which treatment is best suited for that location. Each option is discussed in some detail herein, and Table 6 provides an overview of the opera- tional impacts. Barrier Separation Concrete barrier separation provides the most positive sepa- ration between managed lanes and general-purpose lanes (see Figure 30), which makes it the easiest to enforce among man- aged lane facilities. It also leads to the fewest opportunities for crashes between vehicles in the managed lanes and vehicles in the general-purpose lanes because access is physically limited. Barriers mitigate the potential negative effects of speed differ- ential between parallel traffic streams, so higher speeds can be sustained in the managed lanes regardless of conditions in the general-purpose lanes, and reliability is greater than for non- separated designs. Movable barriers also provide flexibility for reversible lanes, allowing lane balance to shift over time as the demand changes. Barrier separation is the most costly of separation treat- ments because of the need for duplicate shoulders as well as the additional width of the barrier. It is also the least flexible treatment, with changes to access points for the managed lanes being the most difficult change. Sufficient shoulder width must be provided within the managed lane envelope to accommodate an emergency situation, a disabled vehicle, or other special event. For this reason, a barrier-separated facil- ity should have sufficient shoulder width and preferably two travel lanes or a total clear width (i.e., envelope width) of at Separation Option Potential Advantages Potential Disadvantages Barrier (concrete) Facilitates reversible lane Physical separation from general- purpose lanes Easier to enforce compliance Users have feeling of confinement following a crash or incident No way out unless removable rail or gates are installed, which can be an issue in the event of a crash or other mishap Buffer Inexpensive relative to other separation options Vehicles have a way out Drivers can more comfortably position their vehicle within the managed lane with respect to the median barrier Possible operational or safety issues due to increased number of access points Extra right-of-way requirements Pylons (flexible delineators) Less expensive than concrete barrier Provides visible separation Easy to remove Provides a way out Frequent maintenance/replacement Safety concerns due to vehicles able to drive through delineators Possible flying hazard when hit by vehicles at speed No buffer (pavement markings only) Inexpensive relative to other separation options Vehicles have a way out Easy to remove Easy to install Safety and enforcement issues due to vehicles entering and exiting at numerous locations Separate roadway facility (alongside or outside right-of-way, or elevated/depressed around general- purpose lanes) Can be designed to desirable dimensions, requiring fewer design exceptions Easier enforcement Highest potential operational performance Lengthy construction time Must fit available right-of-way More difficult for emergency access Expensive Limited and expensive to access Source: Adapted from Sas et al. (71). • • • • • • • • • • • • •• • • • • • • • • • • • • • • • • Table 6. Operational impacts of separation options.

49 least 20 ft (71) so that the design can accommodate two side- by-side vehicles in case of crashes or other incidents. Buffer Separation A buffer is a physical space between managed lanes and general-purpose lanes operating in the same direction that provides separation without a barrier. A buffer is defined by pavement markings to provide guidance to the driver (see Figure 31), and access points are defined by changes in those markings. Buffers promote more efficient traffic flow where travel speed differentials in adjacent lanes can be substantial but where a barrier is not a practical or desired solution. Buffers are less costly than barriers; however, depending on the width of the buffer provided, they may require additional right-of-way. A 4-ft buffer is commonly recommended (6, 72), though wider buffers can sometimes provide additional benefits in terms of drainage, snow storage, capability to expand the number of lanes in the future, and visual separation. However, a buffer that is too wide may encourage drivers to use it as an additional travel lane. Conversely, 2-ft buffers have been used in some locales and can still provide a measure of separation where a wider buffer is not possible due to cross-section constraints (67). A consideration with buffer separation is the provision of access to and from the managed lane facility. If access is con- tinuous, then the pavement markings must reflect that, and the buffer needs to incorporate broken striping to facilitate ingress and egress. If access is limited to certain locations, the buffer will involve solid and dashed striping to provide at- grade access to and from the managed lanes. Consideration must be given to enforcement for a buffer-separated facility with limited access because, with no physical barrier, drivers may attempt to drive through the buffer to access the man- aged lane regardless of pavement markings. Similarly, crashes involving vehicles entering and leaving the managed lane should be considered when designing access openings; open- ings should be provided where they will be most beneficial to users for access to and from the managed lanes, but openings should not be provided in locations with high weaving vol- umes if they lead to an increase in crashes. Pylon Separation Pylons (also called flexible delineators or tubular markers) are a supplemental separation device used in a buffer to provide additional physical and visual separation (see Figure 32). Pylons can help deter managed lane violations and restrict access, but they do not have the same physical separation as barriers. There are differing opinions on the benefits, related to the cost, of the use of pylons versus the use of concrete barriers or other forms of separation. The true benefits of pylon access are difficult to assess due to the limited amount of data and difficulty in collecting the data to complete a comprehensive and conclusive evaluation. However, studies have shown that pylon maintenance and replacement can be costly, particu- larly when buffers are narrow (73). Some of the other benefits and disadvantages attributed to the use of pylons and con- crete barriers include the following: • Pylon benefits: – Incident management access—pylons are mountable/ passable. – Emergency vehicle access—pylons are mountable/ passable. – Lower initial cost. Source: Kay Fitzpatrick. Figure 30. Example of barrier-separated managed lane on I-25 in Denver, Colorado. Source: Darren Henderson. Figure 31. Example of buffer-separated managed lane on I-110 in Gardena, California.

50 – Sight distance improvement—no wall obstruction is present. • Pylon disadvantages: – Ability of motorists to travel from general-purpose to managed lanes (or vice versa) in attempts to increase speed or avoid toll readers or managed lane enforcement by driving through the pylons. – Maintenance cost for repair/replacement of pylons. – Exposure of maintenance staff and contractors to moving traffic during maintenance activities. – Traffic control cost to accommodate maintenance activi- ties (depending on buffer width). Another consideration for the use of pylons is that, com- pared to barrier-separated facilities, a crash that occurs on the facility is more likely to affect traffic in both the man- aged lanes and general-purpose lanes. Pylons also introduce a potential obstacle in snow-plowing and road-sweeping operations. Therefore, the decision to use pylons for man- aged lane separation should be made with consideration of the expected maintenance cost and in conjunction with other factors that influence maintenance and operations. When an agency determines that pylons may be a suitable device to provide lane separation for a managed lane facility, the agency needs to consider several aspects of these devices in order to implement them in the most efficient manner (74): • Curb-mounted vs. pavement-mounted pylons (curbs can offer more barrier-like features but can increase maintenance costs). • Longitudinal spacing (a minimum of 10 ft is recom- mended, though that length can be increased on tangents with unrestricted sight distance and where strict enforce- ment is regularly provided; spacing that is too large can encourage drivers to change lanes between pylons). • Wider buffers (particularly on curves, but avoiding buf- fer widths that place pylons 4 to 8 ft from the edge of the travel lane because those widths may encourage drivers to stop in the buffer with insufficient space to safely attend to their vehicles). • Pylon height [the 2009 MUTCD (1) specifies a minimum height of 28 in., but pylons that are too tall may be less durable when struck by vehicles]. • Running length (pylons should be used to restrict access where access is not desired, but the effective weaving area at access openings should be checked to ensure that the length of the pylons does not decrease the minimum required weaving distance given traffic volumes and speeds). • Color and retroreflectivity (the designer should consult the MUTCD for appropriate specifications, such as using pylons that are the same color as the pavement markings they supplement). In addition to the standards and guidelines provided by the MUTCD, some states and facility operators have their own requirements. There is not necessarily a definite value for mini- mum buffer width to use pylons; some projects have applied pylons in a 2-ft width, while other practitioners do not use pylons in buffers less than 4 ft in width (67). The California MUTCD describes characteristics of supplemental channel- izing devices in Chapter 3H (75) for use on roadways in that state. For more discussion on these and other traffic control devices, refer to Chapter 4. No Separation (Pavement Markings Only) For the most constrained rights-of-way and for projects that only operate during peak periods and revert to general- purpose operation at other times, the separation between managed and general-purpose lanes may be only a single- lane line pavement marking (see Figure 33). At a minimum, the separation needs to be a pavement marking that is wider than a typical lane line. While this treatment is the least costly option for separation, it is also the least restrictive and may be hardest to enforce when compared to other options. For tolling purposes, it may require more frequent toll readers to be installed. A single pavement marking may still distinguish an access-restricted design, but drivers may attempt to access the managed lane regardless of pavement markings. Reversible Lanes Reversible lanes provide a managed lane treatment that can benefit a corridor with high directional splits Source: Marcus Brewer. Figure 32. Example of pylons separating managed lanes from general-purpose lanes on the Katy Freeway (I-10) in Houston, Texas.

51 where significant peak-direction volumes can be collected and distributed to other roadways. With this treatment, barrier-separated travel lanes are assigned to one direc- tion of travel during a peak period (e.g., inbound traffic during the morning peak) and then reassigned to the oppo- site direction at a different time (e.g., outbound traffic during the evening peak). Reversible lanes often operate in the median and are separated from adjacent oncoming traffic by permanently placed barriers and channelized ramps. Reversible designs can accommodate single or mul- tiple travel lanes. A variety of cross sections are common to both. Positive control, usually in the form of a gate cushion capable of stopping freeway-speed vehicles, is essential to prevent wrong-way movements. Such vehicle-arresting bar- riers are a last resort to prevent wrong-way movements, and vehicles hitting these barriers are absorbed by a wire mesh so that extensive damage is avoided. This barrier should be pre- ceded by a series of breakaway gates to alert even impaired drivers of the wrong-way movement prior to striking the final barrier. Chapter 4 provides more details on gates and positive control in the section on traffic control devices for reversible lanes. An example of a reversible-lane facility serving HOVs is the I-35E/US-67 corridor south of downtown Dallas, Texas (see Figure 34). Access is limited to a few selected points. Several components need to be included for reversible- lane implementation: • The design vehicle will influence access and width (e.g., buses only, passenger cars only). • A variety of traffic control devices including barricades, gates, and dynamic signs must be used at access points to communicate direction of flow and prevent unauthorized or wrong-way entry. • The lanes must be separated by barriers from the adjacent lanes to minimize conflicts, and use of other types of sep- aration must take into account the necessity of avoiding head-on collisions. Contraflow Lanes Contraflow lanes may be designed in two ways: (a) mov- able barriers or (b) portable pylon separation. In the 1980s, all contraflow operations on freeways employed placement of pylons in pre-drilled holes in the pavement. Spacing for pylons was about 20 ft, dropping to 10 ft at crossover transitions where they terminated. Proprietary technology for moving concrete barriers subsequently replaced some pylon treatments in such projects as the Long Island Expressway project in Queens and the Gowanus Expressway project in Brooklyn. However, some projects have such narrow pavement widths that pylon placement is still practiced, including the SR-495 project in northern New Jersey. Both approaches require a heavy investment in deployment crews who must place barriers or pylons before and after each operation period. This pro- cess requires specialized equipment, provisions for storing the equipment (often in the highway median), and exten- sive consideration of the safest design of the facility. The movable barriers provide the freeway corridor with added Source: Darren Henderson. Figure 33. Example of pavement marking separation on Arizona Loop 202 in Maricopa County, Arizona. Source: TTI. Figure 34. Reversible lanes on I-35E/US-67 in Dallas, Texas.

52 flexibility to provide capacity in the direction of travel that needs it most. An example of such a facility is the I-30 corridor east of down- town Dallas, Texas (see Figure 35). The managed lanes operate as HOV lanes and are barrier separated to form their own corridor within the freeway. Access is limited to a few selected points. Several components need to be included in contraflow implementation: • There must be a way to store the barrier and barrier-moving equipment in the median and a safe way of implementing crossovers to allow peak-direction traffic to enter and exit the lane. The kind of traffic that will be allowed (e.g., buses only, passenger cars only) becomes a driving principle in the design. • Barricades or other treatments must be used at access points to prevent unauthorized or wrong-way entry during non-operating periods. • There must be sufficient difference in directional split to borrow a general-purpose lane without creating conges- tion at the times contraflow operation is taking place. • There must be a defined protocol for how and when to move the barrier to prevent disruptions in traffic and min- imize the potential for conflicts and crashes. Pullouts—Enforcement With the vehicle restrictions that are inherent to managed lane facilities, enforcement is necessary to ensure that only authorized vehicles are using the facility. Managed lane restric- tions are in addition to traditional traffic violations (e.g., speed- ing, reckless driving) that must also be addressed. Enforcement is best performed without the need for field presence such as systems that manage toll evaders, but some on-site enforcement presence is still a requirement to manage various types of viola- tions to operation rules. As a result, enforcement officers need accommodation to safely perform their field enforcement duties within the managed lane facility. A treatment that can provide these benefits, particularly in constrained rights-of-way, is an enforcement monitoring and pullout area. These are paved areas, typically in the median, that allow emergency vehicles to park within them, separating them from adjacent traffic. Some states provide guidance on the design and placement of enforcement pullouts; more details can be found elsewhere in this chapter. In general, though, pullouts should be consid- ered at regular intervals in facilities that do not have adequate shoulder width to store a vehicle outside of the travel lane. California (53) and Washington (55) guidelines recommend enforcement areas be located at intervals of 1 to 3 mi. For median-located managed lane facilities, the pullouts should also be located in the median. For part-time shoulder use facil- ities or stand-alone facilities, roadside pullouts would be more appropriate. Pullouts may also be provided on exit ramps if right-of-way is unavailable within the freeway corridor. When designing pullouts, the designer must include suffi- cient width to allow storage of the vehicles using the pullout, along with additional width to allow enforcement officers, emergency responders, and others to walk outside their vehicles without having to enter the adjacent travel lane. California (53) calls for a 23-ft (7.0-m) width for its median enforcement areas. Pullouts must also be designed with sufficient length to allow proper acceleration and deceleration for vehicles entering and exiting. California (53) and Washington (55) guidelines have lengths of 1300 ft for their enforcement pullouts (Washington allows a minimum of 1000 ft) in addition to the entry and exit taper lengths. Examples of bi-directional median enforcement areas found in California are shown in Figure 36 (for median widths of at least 23 ft) and Figure 37 (for median widths of less Figure 35. Contraflow lanes on I-30 in Dallas. Source: TxDOT (76).

53 Source: Caltrans (53). This figure is solely intended for use in the California Department of Transportation’s (“Caltrans”) High-Occupancy Vehicle Guidelines as examples of high-occupancy vehicle lanes used within California. It is neither intended as, nor does it establish, a legal standard for use in other environments. The figure is for the information and guidance of the officers and employees of Caltrans. The figure is not a substitute for engineering knowledge, experience, or judgment. The examples given herein are subject to amendment as conditions and experience may warrant. Copyright 2003 California Department of Transportation, all rights reserved. Figure 36. Bi-directional enforcement area for wide median. Source: Caltrans (53). This figure is solely intended for use in the California Department of Transportation’s (“Caltrans”) High-Occupancy Vehicle Guidelines as examples of high-occupancy vehicle lanes used within California. It is neither intended as, nor does it establish, a legal standard for use in other environments. The figure is for the information and guidance of the officers and employees of Caltrans. The figure is not a substitute for engineering knowledge, experience, or judgment. The examples given herein are subject to amendment as conditions and experience may warrant. Copyright 2003 California Department of Transportation, all rights reserved. Figure 37. Bi-directional enforcement area for narrow median.

54 than 23 ft). The design of a pullout may also serve to provide maintenance access to toll system and ITS components. The primary type of managed lanes rules infraction that enforcement officers confront is occupancy violations, which requires them to see inside a vehicle and to be able to count the number of occupants. Good lighting and a safe vantage point are needed to perform these enforcement functions. Enforcement areas should not be placed under bridges to ensure the safety of enforcement personnel (54). Figure 38 shows a recent design applied to improve visibility over the above design and to serve primarily as an enforce- ment monitoring platform to observe toll compliance when free HOVs are allowed. This design allows officers to see over the median barrier and is placed so that toll beacons are visible from the raised parking area. The design also includes barrier overlap to protect law enforcement vehicles from being struck. This design must provide enough space between the barriers for a vehicle to maneuver in and out and for the officer to enter and exit the vehicle. Pullouts—Refuge On corridors with narrow lanes and corridors where part- time shoulder use is permitted (i.e., where essentially the full roadway width is used as travel lanes), disabled vehicles can effectively render one or more travel lanes unusable and thereby reduce the effectiveness of the managed lanes. In these situations, provision needs to be made for disabled vehicles to be removed from the travel lanes. Figure 39 illustrates an emergency refuge area (ERA) in England, and Figure 40 shows an approach to an emergency refuge area from the perspective of the driver. Along the M42 motorway in England, ERAs are located approximately every 500 m and include emergency roadside telephones. The telephones are accessible to wheelchair users, located behind safety fencing, and feature text messaging and eight different languages of verbal assistance (77). Pullouts for enforcement are more common than refuge pullouts at many existing facilities because vehicle eligibility enforcement is a key feature in HOV and HOT facilities, though pullouts have been used as early as the 1970s (see Fig- ure 41). However, provision for disabled vehicles should be considered when sufficient shoulder width cannot be provided continuously along the length of the corridor. Pullouts may also be provided on exit ramps if right-of-way is unavailable within the freeway corridor, but ramp pullouts are less useful for refuge purposes because the greatest need may be for vehicles that cannot travel to the next ramp but still need to be removed from the travel lane. Geometric requirements for refuge pullouts have similarities to enforcement pullouts in that they need to be wide enough to store the disabled vehicle and still allow drivers or service providers to walk around the vehicle and assess its needs. Ref- uge pullouts also need to have sufficient length to allow storage of at least two vehicles (e.g., the disabled vehicle and a service vehicle or tow truck) in their full width plus accommodation to allow proper acceleration and deceleration for vehicles using the pullouts. With regard to the design of ERAs in England (77), the entrance taper is 82 ft (25 m), the parking length is 98 ft (30 m), and the exit taper is 148 ft (45 m). Issues Unique to HOV Lane Conversion into HOT Lane As needs change over time, an HOV lane operator may determine that a different kind of managed lane facility is preferable and convert that HOV lane into a facility that is more suitable. This section provides guidance on issues that Source: Adapted from Caltrans project plan sheet. Figure 38. Enforcement monitoring area with elevated platform for improved visibility.

55 Source: Google EarthTM. Figure 39. Example of emergency refuge area with hard shoulder in England. are unique to converting an existing HOV lane into a HOT lane. Some of the guidelines described herein may also be applicable for conversions to other types of managed lanes, but the designer should use strategies most appropriate for the specific type of managed lane being considered. General Design Considerations When pricing is added to HOV lanes, many other oper- ational changes may also occur that are reflected in design changes, for example: • Changing signing and pavement markings. • Changing or restricting access, and adding weave lanes or other means of accommodating changes in demand. • Enhancing design treatments that help enforcement. • Modifying utilities—providing power to tolling equip- ment and telecommunications to the back office. • Reconfiguring project limits and transitions/termini. • Providing maintenance access. When this is the case, trade-offs need to be assessed on an individual basis. Illustrations of key elements in the design of

56 assembly. The horizontal clearance, especially in the median for median-located facilities, should be such that anticipated users can drive past gantries without deviating from the travel way; similarly, the vertical clearance must be gener- ous enough that the truck or bus traffic will be able to safely pass underneath gantries or other toll collection hardware. Additional post-mounted and/or overhead signs associated with toll collection must be similarly accommodated. Many projects are moving toward multiple toll zones employing multiple readers. The gantry assembly for a toll reader may be as simple as a traffic signal mast arm that protrudes over the managed lane and leftmost general-purpose lane to capture and differentiate between tolled and non-tolled traffic carry- ing transponders, as shown in Figure 43. For a HOT facility that is converted from a single-lane HOV facility, some agencies provide a toll beacon mounted on the back side of the toll gantry to help enforcement per- sonnel determine who has not been charged and who has paid (Figure 44). A few projects employ separate self-declaration lanes for accomplishing this, using short sections with an additional lane that provides enough width for HOVs and tolled vehicles to separate as they pass under a gantry. An observation booth or enforcement area is typically located next to the declaration lanes so enforcement personnel can observe who is declar- ing themselves as a non-tolled vehicle and report offenders to downstream enforcement personnel (see Figure 45). Insert- ing declaration lanes, along with their lane addition and lane drop tapers, requires additional width in what is otherwise a more typical cross section. This may be accomplished through additional paved width or through restriping the lanes to use part of the adjacent shoulder or buffer; as with other lane-width decisions, trade-offs will have to be consid- ered to determine the best treatment for a particular facility. Accommodating Enforcement in Existing Cross Section More details on enforcement treatments were discussed previously in this chapter, but provision for enforcement is at least as important with tolled facilities as with HOV facili- ties. The cross section must be of sufficient width to allow for enforcement officers to complete their duties. Access Control and Separation If additional lanes are not required, the conversion of an existing general-purpose lane to an HOV or HOT lane is less complicated. The pavement is already in place, and it is likely that little or no additional widening or right-of-way acquisition will be necessary. However, in order to maintain premium traf- fic service levels and discourage toll violations, HOT lanes gen- Source: Beverly Kuhn. Figure 40. Driver’s view of approach to emergency refuge area in England. Source: Chuck Fuhs. Figure 41. Refuge pullout as part of interim design on Banfield Freeway in Oregon. a HOT lane are provided in Figure 42 as an example and are discussed in the following sections. Accommodating Toll Collection in Existing Cross Section For an HOV facility that is being converted to a HOT facil- ity, the appropriate elements needed for toll collection must be installed along the corridor. As with the consideration of additional travel lanes, the cross section must be sufficient to accommodate substructure and superstructure including system hardware without infringing on the travel lanes or increasing the potential for crashes. In particular, the design should accommodate the potential for both toll readers and cameras, which are increasingly located on the same gantry

57 Source: Adapted from Perez and Sciara (78), Figure 10. This material is based upon work by the Federal Highway Administration. Any opinions, findings, conclusions, or recommendations, and translations thereof, expressed in the FHWA publication are those of that publication’s authors and do not necessarily reflect the views of the Federal Highway Administration. Enforcement area located downstream of toll reader (police can visually inspect decals or occupancy from a stationary position, as in Figure 38) Figure 42. Key elements in a cross section of a HOT lane. erally require access control. Physical barriers are preferred for permanent HOT lane installations because they provide better access control and are more effective at reducing violations and maintaining premium traffic service. Since there are often high speed differentials between general-purpose lanes and managed lanes, physical barriers also help maintain safety by preventing potential violators from crossing the buffer into the managed lanes and disrupting traffic flows (78). Implementing a HOT lane conversion on a non-barrier-separated facility may be more difficult—issues related to weaving, safety, enforcement, toll zones, and toll rates are all more complicated with buffer- or non-separated HOV lanes. Specific issues in a buffer- or non- separated HOT lane project will depend on the configuration of the existing HOV lane. Tolling will complicate issues since it pertains to specific traffic movement and behavior related to a driver’s positioning for advantage in a facility without barriers (such as weaving to avoid a toll, or entering and exiting the lane frequently to reduce toll charges). If the proposed HOT lane project is in a region that is new to the HOT concept, it could be advisable to pilot the concept on an existing HOV facility that is barrier separated (79). Additional details about separation of managed lanes were provided previously in this chapter. Similar to HOV lanes, there are two general approaches to providing access to other types of managed lanes: restricted at-grade access and grade-separated access. If the new facility will require a different type of access than currently exists on

58 the HOV lanes, the design will have to accommodate both the new ramps and the connections to them. In particular, if a facility changes from at-grade access to grade-separated access, the design must have sufficient cross section to include the transitions and speed-change lanes associated with the new grade-separated ramps. Access Considerations Access into and out of a managed lane facility is one of the more critical elements in the design of the facility. Drivers must be able to make their way safely into and out of the facility in order to derive any usefulness from the facility. This section contains discussion on some of the considerations related to the design of managed lane access. Consideration of Limited Access Versus Continuous Access Unless a facility is designed to allow access at only the beginning and end, provision for intermediate access is typi- cally provided. This is accomplished through either continu- ous access (in which eligible vehicles may enter the facility at any location along the facility) or a limited- or restricted- access approach (in which selected locations are designated for ingress, egress, or both, usually at 3- to 5-mi spacing). In discussing the differences between the limited and con- tinuous access, California’s Traffic Operations Policy Directive 11-02 (35) states that consideration should be given to both access types when planning managed lanes. The choice of access type is based on a general evaluation of the performance and management benefits for the entire freeway as well as the capital costs of building and operating the managed lanes. A summary of design, cost, and performance considerations for the two types of access designs is provided in Table 7. The directive adds that various research and engineering studies on managed lane facilities have found that the highway features that can have the greatest effect on performance, including safety and throughput, are: • Frequency, location, type, and design of intermediate-access openings on limited-access facilities. • Shoulder widths. • Traffic control and safety devices that provide positive guid- ance (usually related to access points and driver decision- making, such as overhead signing, striping, and lighting). Experience with existing facilities suggests that when con- sidering whether to use limited or continuous access on a managed lane facility, operational effects should be the high- Source: Chuck Fuhs. Figure 43. Toll reader gantry on SR-167 in Seattle, Washington. Source: Chuck Fuhs. Figure 44. Toll beacons located on back side of gantry on I-15 in San Diego, California. Source: Chuck Fuhs. Figure 45. Toll declaration lanes at an electronic toll location on SR-91 in Orange County, California.

59 est priority (67), and such effects should be identified through a formal operational analysis (35). Safety, economic, and enforcement effects are also valid considerations, but they often exist in conjunction with operations. For example, if a proposed access point or series of points causes bottlenecks or does not allow managed lane drivers to use their desired general-purpose entrance and exit ramps, then the access is not serving its purpose and the effects of other factors may not be fully appreciated. This suggests a level of analysis that looks not only at the entire corridor’s travel patterns (specifi- cally including the desired origins and destinations for man- aged lanes users) but also at the immediate area surrounding each proposed access point or driver decision point. The Highway Capacity Manual (HCM) 2010 (80) contains analysis methodologies for freeway weaving segments (in Chap- ter 12) and freeway merge and diverge segments (in Chapter 13) to predict operational characteristics such as the rate of lane changing, average speed of weaving and non-weaving vehicles, level of service, and capacity. The HCM makes the disclaimer that a limitation of the methodologies is that it does not specifi- cally address special lanes, such as HOV lanes, within the weav- ing segment without modifications by the analyst. However, the principles used in the HCM methodologies have applicability for managed lane facilities, particularly for two-sided weaving segments such as those that might be found between a general- purpose ramp and a managed lane access opening. In addition, Criterion Limited Access Continuous Access Cost Detailed operational analysis and an iterative design process are needed for best placement of access points. This access may require more roadway width to accommodate buffer and access opening. Additional pavement markings and overhead signing are required. Investment in monitoring for congestion and/or crashes due to weaving near access points may be needed. Lower cost is incurred for design, analysis, construction, operation, and maintenance. Adjustments require fewer engineering resources. Mobility, Safety, and Performance Access points can become initial source of unstable flow and queuing in the managed lane, which can trigger the onset of congestion among all lanes. Left-side access openings intensify weaving in the form of concentrated flows and consecutive lane changing across all freeway lanes, which may present difficulties for all drivers during periods of congestion. Drivers are unable to access the managed lane when desired; this could induce violation of the buffer striping, which may be unexpected by drivers in the managed lane. Limited access is a potential strategy to restrict lane changing where demand has produced or may produce a performance deficiency. Longer-distance trips are accommodated by discouraging short-term use of lane. Smoother flow and higher speeds can result from limited merging. Limited access can provide greater separation to accommodate lane closure activities in the lane or adjacent lanes. Access to some general-purpose ramps is not as convenient. Users must focus on potential for vehicles to enter or exit the managed lane at any point; this may reduce speeds. Flexibility allows last-minute lane changing to reach freeway exit ramps. There is no concentrated weaving; lane changing occurs along entire corridor when gaps appear. Users can readily access all general-purpose ramps. Drivers face less complex decision making. The facility is easily utilized during off-peak hours (for part- time facilities). There is less separation to accommodate lane closures. Drivers will not worry about violating barrier striping when managed lane is closed for construction, maintenance, or incidents. Enforcement There is a potential for lower toll evasion and occupancy violation. Enforcement is simpler. Express lane toll collection is simplified due to need for fewer readers. Greater investment in enforcement activity, systems, and zones is needed to produce the lower violation rates expected with limited-access designs. There is a potential for higher toll evasion and occupancy violation. The expected cost for express lane toll collection is greater due to need for additional readers. Note: This summary does not apply to limited-access designs in which managed lane access is provided only via direct ramps to a local or other state highway or freeway. Source: Compiled by TTI. • • • • • • • • • • • • • • • • • • • • • • • • • • • • Table 7. Summary of design, cost, and performance considerations for limited- and continuous-access facilities.

60 some operators of managed lane facilities may have policy direc- tives that specify analysis tools to be used, and those applicable policies should be considered in the planning process as well. A Minnesota study (81) investigated design, operational, and safety differences between two managed lane corridors: I-394, with a restricted-access design, and I-35W, with an open- access design. The authors stated that it is difficult to compare the two design philosophies because they were devised to serve the needs of the two distinct roadways. They concluded that I-394 operated very well with the closed-access design mainly because the majority of the demand originated from three distinct service interchanges, and the remaining ramps had comparatively little demand. Conversely, the interchange density on I-35W was much higher, with entrance ramps very closely spaced and with the majority of those ramps carry- ing large demands of HOT-eligible vehicles. The researchers stated that it would have been very difficult to follow a closed- access design on I-35W, and their research results led them to conclude that it would have made little difference in terms of mobility and safety. Comparisons of shockwave characteristics of four access zones were highlighted, and, although the vol- umes at each access zone were different, the shockwave lengths observed were comparable, signaling no difference in terms of safety between the two design philosophies. Practitioners who work with existing continuous-access facilities recognize operational benefits of open access, and open access has a simpler analysis process, which identifies potential locations for restrictions instead of justifying locations for access. Practitioners in a recent survey also indicated that based on their anecdotal feedback from drivers, those who travel on open-access facilities also prefer greater access, and their experience was that effects on safety can be minimal (67). However, for managed lane facilities that are revenue funded or are based on public–private partnerships, the economic effects of continuous access must be considered. If access is not con- trolled, there is a potential for lost revenue, which could affect a project’s financial feasibility. Also, in areas with very high con- gestion, it may be desirable to limit access to the managed lane to prevent congestion from the general-purpose lanes affecting the operations of the managed lane, or conversely, reduce the effect of managed lane traffic weaving across all the general- purpose lanes to enter the managed lanes or exit the freeway. It is also important to have appropriate context when making decisions on operational effects of access. Managed lane and general-purpose lane operations are intertwined, so decisions about managed lanes should not consider those lanes as if operating in a vacuum. In addition, the context of time must also be considered; an access point that operates satisfactorily on Day 1 or in Year 5 may not maintain that level of operation at the end of the expected design life, so every decision maker involved in the process needs to have the same expectation of the time horizon being considered. A managed lane access analysis should not attempt to address major free- way operational shortcomings, particularly for forecast con- ditions, which may overwhelm any access location selection. Continuous access and restricted access do not have to be mutually exclusive, in that a managed lane operator does not have to choose either one or the other for the entirety of a managed lane facility. Within a given facility, restrictions on access can be applied to certain parts of the facility while continuous access is provided in other parts of the same facility. Practitioners should use the principles discussed in the following sections to determine where access restrictions should be located. Continuous-Access Considerations Continuous access allows eligible vehicles to enter and leave the lane at any point. No additional weave, acceleration, or deceleration lane is provided, and no specific ingress/egress locations are designated. Instead, vehicles move into and out of the managed lane at any point in the same way they would change lanes in the general-purpose lanes. The striping used to separate the general-purpose and the managed lanes, along with signing and pavement markings, should indicate that access can occur at any point. Continuous access can be applied in projects with or without buffer separation between the managed lane and the general-purpose lanes. Recent guidance from California (35) on the conversion of restricted-access facilities to continuous access indicates that the conversion may be allowed if it is funded by the project sponsor requesting the change. A traffic study is required for any conversion project. If a new or conversion project is on a route where express lanes (i.e., managed lanes that utilize con- gestion pricing) are planned within the next 5 years, and there is an intent to operate the express lane with continuous access, the California guidance requires joint consultation among the project sponsor, the state DOT, and the state police to iden- tify strategies to limit violations. Among the strategies to be considered in the multiagency consultation are frequent toll readers and visible manual enforcement. Frequency of Restricted-Access Points For facilities that do not provide continuous access, defined access points are provided at regular intervals or logical loca- tions; restricted-access points are not intended to serve every general-purpose entrance and exit ramp. A survey of managed lane practitioners (67) found that spacing for access openings on existing facilities tend to be between 1 and 3 mi, though the reasons for providing access of a particular type or in a specific location vary, ranging from policy decisions to safety or operational considerations. Each agency determined which factors were most important to a given facility, and the factors

61 chosen varied from one facility to another. Similar consider- ations of applicable policies and safety or performance goals must be made when planning for new or revised access points. Chapter 10 of the AASHTO Green Book (27) provides guid- ance on spacing between freeway access points. While this is intended for general-purpose access, the principles can be applicable to managed lanes. To provide sufficient weaving length and adequate space for signing, a reasonable distance should be provided between successive ramp terminals. Figure 10-68 of the Green Book presents recommended minimum ramp terminal spacing for four various ramp-pair combinations that are applicable to interchange classifications; the exit-entrance (EX-EN) and entrance-exit (EN-EX) com- binations are particularly applicable to managed lane access. The stated minimum distance for EX-EN pairs on freeways is 500 ft. For EN-EX pairs (i.e., weaving sections) on system- to-service interchanges and service-to-service interchanges, the minimum distances are 2000 ft and 1600 ft, respectively. EN-EN and EX-EX pairs on freeways have a minimum dis- tance of 1000 ft between successive ramp terminals. As an alter- native to fixed dimensions for ramp and interchange spacing, NCHRP Report 687 (82) provides performance-based guid- ance on ramp spacing that considers geometric design, traffic operations, signing, and safety. California’s HOV guidelines (53) state that when it is operationally possible, ingress and egress locations are based on the following criteria: • To serve every freeway-to-freeway connection. • To serve high-volume ramps. • To serve ramps with high numbers of carpools. • When adjacent to park-and-ride facilities. • When requested by transit districts. • To assist in the modification of local commute patterns (may be at local request). • To help balance and optimize interchange operational level of service within a local jurisdiction, within a corridor, or within a region. • To support and encourage ridesharing programs (HOV demand/usage). Depending on the purpose of the facility, decision makers may add to or reprioritize the criteria in this list to determine where access should be provided. Additional guidance from California (35) states that exist- ing interchange spacing is the primary consideration for deter- mining the location of access openings. An equally important consideration is the existing and expected location of main- line operational bottlenecks and geometric constraints that produce recurrent congestion and queuing along the general- purpose lanes. Access openings should be located and designed such that they will perform at LOS C or D. They should not produce adverse impacts to managed lane and general-purpose lane performance, nor should they be placed where recurrent general-purpose lane congestion is expected. This guidance avoids the potential for undesirable conditions that result in operational and safety deficiencies. If the mainline queuing at a proposed access location is limited to a small portion of the overall peak period, then a weave-lane or merge-lane configu- ration might need to be evaluated and provided if it will elimi- nate or minimize adverse impacts. The location of managed lane access points should avoid the creation of short weaving distances between upstream and downstream right-side ramps and left-side managed lane access. It may be difficult to find appropriate locations for man- aged lane access on older freeways with frequent on- and off- ramps. This is one reason that managed lanes on freeways in the San Francisco Bay Area operate with more open access, so that weaves are not constrained. Ample distance between man- aged lane access and general-purpose access will minimize the likelihood of congestion and related safety implications involv- ing weaving vehicles. Recommendations for spacing between general-purpose ramps and managed lane access in previous studies and current guidelines (4, 35, 53, 54, 59, 83, 84, 85) have suggested that cross-facility weaving areas should provide between 400 and 1000 ft per lane change, depending on antici- pated traffic volumes and other conditions. A distance of 1000 ft is the common recommendation among the more recent ref- erences, as shown in Figure 46. These distances are typically applied to passenger vehicles, so the weaving distance provided for buses should be at least as long. Source: Adapted from Kuhn et al. (4). Desirable – 1000 ft [305 m] per lane change Figure 46. Example of termination of managed lane with 1000 ft per lane change.

62 California guidance (35) mandates that the type and loca- tion of proposed access openings shall be determined by an operational analysis. It is expected that an iterative process would be used in this analysis. For example, an access open- ing using the simplest design and minimum lengths might be evaluated first. If the analysis supports this concept, then no further analysis of that location is necessary. Otherwise, the process would continue until an appropriate concept is identified, or all reasonable and feasible concepts are exhausted. The iterative process may require consideration of the following modifications or features (not necessarily in this order): • Increase in weaving lengths. • Addition of alternative types of access. • Relocation of the access opening. • Addition of auxiliary lanes connecting ramps on the general-purpose lanes. • Addition of drop ramps or direct-connector (flyover) ramps. Proposed access openings that are estimated to operate below the performance thresholds or that propose less-than- minimum lengths or spacing receive additional scrutiny. Approval will be considered when the need for the opening is justified by traffic data and the safety analysis and if traffic impact mitigation is incorporated. Approval may also require specific system monitoring to identify and correct potential performance deficiencies. Nevada DOT (54) and the FHWA Priced Managed Lane Guide (2) also require or recommend, respectively, an operational analysis to determine and/or justify the loca- tion and type of access openings in a limited-access facil- ity. Key items that should be considered in an operational analysis include level of service and avoidance of general- purpose weave turbulence such as bottlenecks near major interchanges. This guidance suggests a level of analysis that looks not only at the entire corridor but also at the immedi- ate area surrounding each proposed access point or driver decision point. The Nevada DOT manual (54) states that, generally, an access opening should be provided before and after system- to-system interchanges and other major interchanges. Nevada DOT has the following desirable minimums for frequency of access openings for different types of managed lanes: • For HOV lanes, provide a minimum of 2 mi between access openings. • For express lanes without toll, provide a minimum of 4 mi between access openings. • For priced managed lanes (HOT lanes, express toll lanes, etc.), provide a minimum of 2 mi between access openings. The FHWA guide (2) says that, in all cases, access openings should be located and designed in a way that will not produce adverse impacts to the managed lanes and the parallel highway lanes. However, when forecast conditions are evaluated, the managed lane impacts should not attempt to address and be responsible for mitigating much larger operational issues ema- nating from growth in general-purpose traffic; otherwise, any near-term resolution may be difficult to reach. The locations of at-grade access openings need to be closely coordinated with highway entrance and exit ramps and allow adequate room for motorists to complete weaving movements when mov- ing between the general-purpose and managed lanes and an entrance or exit ramp. Treatment for Beginning a Managed Lane Entering a managed lane facility should require a deliberate movement. A design configuration that requires vehicles to change lanes to avoid entering the managed lane could be sus- ceptible to violations. The only exceptions may be associated with reversible and contraflow lanes where a mixed-flow lane drop condition is common. Since a majority of managed lanes are located on the left side, next to the leftmost lane on most free- ways, the beginning of these lanes should not typically change designation and cause general traffic to drive into a downstream restricted condition. It is desirable to add managed lanes to the overall roadway cross section via a left-side exit, and signing should reflect this. Figure 47 provides an example. Entrances to a managed lane facility are to be designed as lane changes to prevent motorists from entering the facility unintentionally. The entrance point to any managed lane facility that does not have continuous access should begin no earlier than a Source: Fitzpatrick et al. (83). Figure 47. Example of entrance to a managed lane.

63 distance equivalent to 1000 ft per lane change required to enter the managed lane from the nearest entrance ramp. The design should also include accommodation for structures for advance signing upstream of the managed lane(s) and sign- ing at or downstream of the beginning of the facility (see the section of this chapter on operational impacts on design for more information). Intermediate-Access Treatments— Weave Zones and Weave Lanes An access point at an intermediate location may accom- modate both those moving into the managed lane and those leaving the managed lane. In some situations, however, only ingress or only egress may be allowed. A common design that accommodates both maneuvers is a weave lane, which is a short-distance lane added to the cross section to provide space for weaving into and out of the managed lane facility sepa- rate from the travel lanes. A schematic for a buffer-separated option with a weave lane is shown in Figure 48. A weave zone is a design that provides for weaving maneuvers into and out of the managed lane. A weave zone may or may not have a weave lane; in designs without a weave lane, the entry and exit maneuvers take place similar to lane changes between other travel lanes on the freeway. Figure 49 shows an example of a weave zone without a weave lane, which is a design that can be more easily implemented within a constrained right-of-way. However, a weave lane can be provided if it is deemed benefi- cial and sufficient right-of-way is available, particularly when high volumes of access maneuvers are observed or anticipated. An opening or merge area of 1300 to 2000 ft has been recom- mended in other guidelines (4, 53, 54), and a total of 4000 ft is recommended for the entire length of the access area when a weave lane is provided (86, 83). Increasing the length of access points and providing weave lanes are two common treatments that are available for facil- ities with restricted access that have increased demand or anticipate increased demand, particularly if the facility is being considered for expansion from a single lane to mul- tiple lanes. Providing or converting to continuous access is also an option, but each option should be considered in the context of an operational analysis to determine which treatment is best for a particular facility (or specific location within that facility). Source: Fitzpatrick et al. (83). Figure 49. Example of buffer-separated intermediate weave zone access without a weave lane. Source: Fitzpatrick et al. (83). Figure 48. Example of buffer-separated intermediate weave zone access with a weave lane.

64 The weave zone should be longer than the minimum shown in Figures 48 and 49 if the design vehicle is a bus or if a high vol- ume of such vehicles is anticipated. If a merge area is sufficiently long, it may be used as a passing zone for managed lane facilities with just one lane (83); if passing behavior is not desired, then openings longer than 2000 ft are discouraged (83). An illustra- tion of California’s most recent guidance (35) for access open- ings of 2000 ft in length, including dimensions for lane-change maneuvers to adjacent entrance and exit ramps, is shown in Figure 50. The first example is for a weave zone, the second is for a weave lane, and the third is for an auxiliary or merge lane, which is discussed in more detail in the next section. Experience with the use of weave lanes in Los Angeles has highlighted the problem of general-purpose traffic using the weave lane for queue jumping where it is provided in a typi- cally congested area. In such circumstances, the use of sepa- rated ingress and egress merge lanes should be considered to separate the two respective movements and create a break in the weave lane to discourage queue jumping. Intermediate-Access Treatments— Auxiliary Lanes If the buffer is sufficiently wide, managed lane access can take the form of an auxiliary speed-change lane, where a driver desiring to enter into the managed lane facility enters an at- grade ramp, travels through the buffer, and then merges into the managed lane. The same scenario would apply for a driver leav- ing the managed lane facility to enter the general-purpose lanes, as shown in Figure 51 and the third example in Figure 50. The design must have sufficient length for the driver to adjust speed and merge into the desired facility, and there must be sufficient spacing between ramps to accommodate weaving, similar to the design of an auxiliary lane for a freeway entrance ramp. Source: Caltrans, © 2014, all rights reserved. Figure 50. Access types with minimum recommended opening lengths and weaving distances.

65 A shortcoming noted on some projects is drivers attempting to use such ramps to both enter and exit; pavement markings should clearly define whether the ramp is intended for entrance or exit, and application of pylons or concrete barriers may be necessary to prevent unintended movements. Intermediate-Access Treatments— Direct Access Similar to traditional freeway entrance/exit ramps, direct (grade-separated) access ramps carry managed lane traffic directly to and from major interchanges with other freeways and managed lanes, and with the street network (i.e., they require no cross-facility weaving to enter the managed lane facility). Two typical designs for direct-connect ramps are flyovers that are either directional ramps for reversible flow (Figures 52 and 53) or common bi-directional structures (Figures 54 and 55). Ramps to reversible-lane facilities are typically directional and reversible as well. The design of a direct-connect access ramp should follow the guidance of a traditional freeway entrance/exit ramp as described in the AASHTO Green Book (27) and applicable state-level guidance. Design speeds are critical, similar to any high-speed ramp. Most flyovers connecting managed lanes on different freeways are designed with a 50-mph design speed or better. Most bi-directional ramps are separated by a bar- rier between opposing flows. In particular, the designer should ensure that there is provision for sufficient speed-change/ merge/diverge length at the ramp termini, as well as provision for sufficient sight distance around the separation barrier and for attenuation at the gore area of the ramp. Direct ramps may also be located on the left side of the facility, and this setup is more common for managed lanes than for entrance and exit ramps for general-purpose lanes. Two possible purposes for left-side ramps are to provide access from one managed lane facility to another or to pro- vide access to transit facilities. Left-side ramps are typically discouraged for general-purpose lanes, and while they may be appropriate for certain situations in managed lanes to Source: TTI. Figure 52. Example of direct-access managed lane ramps over general-purpose lanes in Houston, Texas. Source: TTI. Figure 53. Example of direct-access managed lane ramp to park-and-ride facility in Houston, Texas. Figure 51. Exit from managed lane using an auxiliary lane on the Katy Freeway (I-10) in Houston, Texas. Source: Marcus Brewer.

66 facilitate access, the design of the roadway must provide suf- ficient cross section for the lanes needed to accommodate the anticipated volumes for each movement. The ramp must be long enough to minimize the potential for queue spillback from the ramp into the managed lane, and sufficient signing must be provided in advance of and at the access point to inform drivers which direction is the mainline since the ramp is located on the left instead of the right. Intermediate-Access Treatments— High-Volume Direct Access For high-volume access movements, particularly between managed lane facilities, it may be appropriate to use a ramp that is equivalent to a freeway-to-freeway connector. In loca- tions where high managed lane volumes are anticipated for connecting traffic between two managed lane facilities or with major transit or activity centers, high-speed flyover ramps are justified. For concurrent flow facilities, ramps typically share a common structure to serve both directions on a common alignment and bridge structure. Flyover ramps are designed to the same geometric and design speed conditions as any other higher-speed freeway-to-freeway connector. To connect man- aged lanes located to the left of general-purpose lanes, ramps are most commonly oriented to the left of the mainline man- aged lane roadway and connect two freeway lanes left to left, sharing a common structure (see Figure 56). Two-way flyover ramps contain barrier separation between opposing flow (see Figures 57 and 58). The cross section on the ramp is the same as a two-way barrier-separated facility. Treatment for Ending a Managed Lane Terminating a managed lane facility requires proper accom- modation of all vehicles (both in the managed and general- purpose lanes) approaching the point of termination. Two methods are commonly used: continuing the managed lane(s) as general-purpose lane(s) or merging the managed lane(s) into the general-purpose facility. The recommended method is the former; multiple guidance documents recommend that a managed lane continue as a general-purpose lane when termi- nated (4, 53, 54). If managed lane use is high and right-of-way is constrained, the design should carry managed lane volumes into a general-purpose lane while dropping a general-purpose lane farther downstream on the right (see Figure 59). Both origin and termination treatment locations need to consider proximity to existing or planned right-side freeway ramps. Experience suggests that for single-lane treatments, provid- ing 1000 ft of minimum lane weaving is desirable, with 500 to 800 ft being adequate in some instances. Determination of the actual weaving distance should be based on a weaving analy- sis. Locating a terminus should consider grade, curvature, and specific traffic conditions. Tangent locations are encouraged where good sight distance can be provided. Each design setting is unique in terms of the traffic mix, lane demand, roadway geometrics, and related factors. The termination treatment location and weave or merge section should also consider traf- fic impacts under differing peak and off-peak travel conditions. If the managed lane volumes do not exceed 1000 vehicles per hour, a merge area of approximately 2500 ft in length may be acceptable, but effects on the general-purpose lanes should be checked. Also note that the merge tapers in design are desirably 115:1 with a minimum of 60:1, and diverge tapers are desirably 50:1 with a minimum of 20:1 (see Figure 60). Some facilities will need to have an interim terminus until they are extended in a subsequent phase. Design of those termini should follow the same guidelines as for permanent Source: Chuck Fuhs. Figure 54. Ground-level view of direct-access managed lane ramp in Houston, Texas. Source: Chuck Fuhs. Figure 55. Example of direct-access managed lane ramp to local street in Houston, Texas.

67 Source: Nevada DOT, Planning Division, Safety Engineering Section (54). Figure 57. Aerial view of two-way interchange flyover ramps. Source: Nevada DOT, Planning Division, Safety Engineering Section (54). Figure 58. Two-way interchange flyover ramp. Source: Nevada DOT, Planning Division, Safety Engineering Section (54). Figure 56. Example layout for a two-way flyover ramp.

68 treatments, and the cross section of the facility must provide sufficient width to accommodate the selected treatment. The design should also include accommodation for struc- tures for advance signing upstream of the terminus of the facility and needed signing at or downstream of the end of the facility (see the following section on operational impacts on design for more information). Operational Impacts on Design The design must allow the facility to serve the purpose for which it is intended, so certain aspects of operation must be considered in the design. This section contains guidance on selected operational issues unique to managed lane facilities that have an effect on the design of the facility. Capacity The geometric features of a managed lane can strongly influence capacity, as well as the pricing and operating rules. The number of lanes, type of buffer, and access design are all characteristics that impact capacity. It is important to consider these features when designing a managed lane, and a capacity analysis may be useful to examine how dif- ferent design alternatives may perform. Some projects such as US-290 in Houston (see Figure 61) have specifically been designed with enough width so that a new access feature, auxiliary lane, or other capacity enhancement can be added in the future without major redesign. Such design flexibil- ity becomes more important in highly dynamic corridors where changes in operation and demand are anticipated into the future. Based on the research conducted for NCHRP Project 03-96, a draft version of a chapter for the Highway Capacity Manual provides guidance to consider how different geometric fea- tures influence capacity (87). The draft chapter is divided into sections that correspond to facility type, specifying these concepts: • Continuous access. • One- and two-lane, buffer separation. • One- and two-lane, barrier separation. Desirable – 1000 ft [305 m] per lane change Source: Adapted from Kuhn et al. (4). Figure 59. Terminating an HOV lane as a general-purpose lane. Figure 60. Terminating an HOV lane into a general-purpose lane. Source: Kuhn et al. (4). Source: Chuck Fuhs. Figure 61. Managed lanes with available width for future enhancements on US-290 in Houston, Texas.

69 In the draft chapter (87), speed-flow diagrams were calcu- lated and provided for each facility type. The speed-flow dia- grams for continuous-access and one-lane, buffer-separated facilities have an additional set of curves that account for con- gestion in the adjacent general-purpose lanes. These curves show a frictional effect. Two-lane facilities that are buffer sepa- rated and all barrier-separated facilities are not influenced by congestion in the general-purpose lanes and thus do not show any friction in the diagram. Adjustments for cross-weaving movements were also sug- gested in the HCM draft chapter from NCHRP Project 03-96 (87). The adjustments are recommended to be applied in instances where intermittent access to the managed lanes is pro- vided. Prior research has indicated that capacity in the general- purpose lanes is negatively correlated with the amount of cross-weaving traffic entering the managed lane and the num- ber of lanes that have to be crossed by vehicles. Capacity can also be reduced when the managed lane access point and the freeway entrance or exit ramp are relatively close to one another. The amount that capacity is reduced in the general-purpose lanes, due to the placement of intermediate-access segments, can be represented through a set of capacity reduction factors. The capacity reduction factors can be estimated as a function of the number of general-purpose lanes, the amount of cross- weaving flow, and the distance between the freeway on-ramp and the start of the managed lane access segment. The follow- ing equation is used to calculate a capacity reduction factor [from Table 26 in NCHRP Web-Only Document 191 (87)]: CRF % 8.957 2.52 ln CW 0.001453 0.2967 GPLs CW MinL( ) ( )= − + × − × + × − Where: CRF = capacity reduction factor CW = cross-weave flow measured in passenger cars per hour. LCW-Min = length from the gore of the freeway entrance ramp to the beginning of the managed lane access segment. GPLs = number of general-purpose lanes, ranging from two to four lanes. Examples of various capacity reduction factor estimates (developed using the above equation) can be found in Table 8. These estimates are derived based on common combinations of roadway geometric features and cross-weave flow conditions. Tolling Systems This section contains a brief discussion of selected design topics related to tolling systems. For more information on toll- ing systems, please refer to the section on toll collection sys- tem development, deployment, and phasing considerations in Chapter 5. Electronic toll collection requires a decision on where the toll collection occurs, and such locations are almost always in a constrained median environment. A significant investment in gantries for signing and tolling is required, so the place- ment of tolling installations needs to strategically assess how the facility can be best managed with a minimum of installa- tions. Toll installations need to be accessible for maintenance, both from the median and often from the right side where power and communication are provided. A variety of differ- ent vendors offer electronic tolling systems, but all use some form of transponders to read electronic tags in users’ vehicles and/or cameras to read license plates and send bills to those without transponders. Regardless of the choice of tolling system, the design will need to accommodate the necessary infrastructure, which typically involves large structures, such as gantries, as well as hardware such as cameras and electronic readers, roadside controllers, communications equipment, and power. The use of variable pricing on priced managed lanes requires additional 4 GP Lanes LCW-Min (ft) Cross-Weave Flow (vph) 100 200 300 400 500 1500 1.6% 3.3% 4.5% 5.7% 6.5% 2000 1.0% 2.2% 3.7% 4.7% 5.3% 2500 0.2% 1.6% 3.3% 4.1% 4.7% 3 GP Lanes LCW-Min (ft) Cross-Weave Flow (vph) 100 200 300 400 500 1500 1.2% 3.1% 3.9% 5.5% 6.1% 2000 0.8% 2.0% 3.5% 4.3% 4.7% 2500 0.2% 1.4% 2.6% 4.1% 4.3% 2 GP Lanes LCW-Min (ft) Cross-Weave Flow (vph) 100 200 300 400 500 1500 1.0% 2.7% 3.7% 5.1% 5.3% 2000 0.6% 2.0% 3.1% 4.3% 4.9% 2500 0.0% 1.2% 2.2% 3.3% 4.1% GP = general purpose. Table 8. Capacity reduction factor estimates.

70 infrastructure and communications abilities to communicate the current toll to drivers. Additional discussion of appropriate infrastructure can be found in Chapter 6 in the section on toll collection system operations. If pricing is being used to maintain a specified operational threshold, a variable toll system needs to either be based on a schedule that reflects typical peak-demand curves or be dynamic and receive real-time traffic input to calculate the toll rate. This real-time traffic information is obtained using loop detectors or other devices capable of detecting characteristics of the traffic stream in the managed lane(s), such as traffic vol- ume and speed. Dynamic signs are commonly used to display toll rates for downstream destinations. The toll that customers see when making a choice to use or remain in the lane should not change from the time they see the rate to the time they actually enter (or remain in) the facility. The tolling system design opens a customer transaction at the first access point but does not process the completed trip transaction until the vehicle passes one or more downstream tolling gantries and the transaction is closed (2). If a single-lane HOT facility has declaration lanes to sepa- rate HOVs from tolled vehicles, the cross section must have two lanes at the tolling gantry and must also provide the necessary lane addition and lane drop tapers on either side of the gantry. An alternative to this method is the use of transponder technol- ogy that enables the driver to declare status as an HOV or tolled user through the transponder or by changing the account status prior to the trip, allowing all traffic in the HOT facility to use the same lane at the gantry. In this case, enforcement usually occurs using an indicator beacon visible to law enforcement personnel on a gantry or other structure indicating the status (HOV or SOV) that the vehicle has chosen to declare. Priced managed lane toll zones should be equipped with all necessary infrastructure to identify vehicles, initiate toll transac- tions, identify and photograph license plates of potential vio- lators (or pay-by-plate/pay-by-mail customers), and inform enforcement personnel of account status through strategically placed transaction status indicator beacons or through displays inside the enforcement vehicle. In a typical toll zone configura- tion, a vertical post with counterbalanced cantilevered horizon- tal arms will serve as the toll gantry. In the dual-gantry system shown in Figure 62, a minimum vertical clearance of 18 ft is provided between the automated vehicle identification (AVI) antenna and rear-plate-facing license plate image camera on the mast arm. A transaction status indicator beacon is mounted where it can be seen by toll enforcement personnel. Many toll zones will also have a designated area for adjacent enforcement personnel monitoring. The availability and placement of these observation locations will generally be in the vicinity of the toll reader and beacons. Sufficient lighting must be present to sup- port license plate recognition and image capture; many current camera systems include their own required lighting for either infrared or color images, but additional lighting must be pro- vided separately if not included as part of the camera system. Variations on the dual-gantry system shown in Figure 62 include single-gantry systems, systems that capture images of only front license plates (instead of rear plates or both), use pavement loops instead of overhead vehicle detectors, and adjust the position of the status indicator beacon if right-of- way is constrained or if there is an adjacent rail corridor. Figure 62. Example managed lane toll zone design. Source: Perez et al. (2), Figure 6-18. This material is based upon work by the Federal Highway Administration. Any opinions, findings, conclusions, or recommendations, and translations thereof, expressed in the FHWA publication are those of that publication’s author(s) and do not necessarily reflect the views of the Federal Highway Administration.

71 All of the priced managed lane toll zone components need appropriate access for preventive maintenance and other in- field needs. To provide this access, all components should be housed collectively in hardened and protected utility cabinets with sufficient controls to prevent tampering, promote safety for maintenance personnel, and provide easy access. These cabinets should be placed a sufficient distance from the travel way, preferably beyond the clear zone, to provide safer access for maintenance and to minimize the fixed-object hazard to drivers. Sufficient conduits to the gantries are installed under the general-purpose lanes (2). Appropriate access to the gan- tries also needs to be considered for maintenance or replace- ment of gantry-mounted hardware and similar activities. Figure 62 shows one example of hardware configuration. Other configurations may place utility cabinets or closets within the median where sufficient width is available; cabi- nets can be mounted on the gantries, or they can be installed at ground level. This median-oriented configuration allows the conduit to also be installed within the median, elimi- nating the need to close freeway lanes for conduit-related maintenance. For median facilities, cantilevered signs from the median can be used to visually separate managed lane signs from others. One solution is to use separate sign structures staggered lon- gitudinally so that each sign sequence on the left (managed) or right (general-purpose) is perceived as a separate sequence, separated laterally and longitudinally. This type of sign struc- ture strategy can be costly, but if addressed early in the design process, proper footings can be installed. For more details on signs to be used on managed lanes, see Chapter 4. While not strictly a geometric design consideration, the design of the pavement (e.g., driving surface and subbases) should be such that it can accommodate necessary conduits for wiring (e.g., power, communications) below the surface of the roadway. Figure 63 shows an example from Colorado (88) of design provisions for conduits and a structure that might be used in a facility represented in Figure 62. Considerations for maintenance of those conduits and the adjacent roadway are discussed in Chapter 6. Enforcement Systems With the vehicle restrictions that are inherent to managed lane facilities, enforcement is necessary to ensure that only authorized vehicles are using the facility. Managed lane restric- tions are in addition to traditional traffic violations (e.g., speeding, reckless driving) that must also be addressed. As a result, enforcement officers need accommodation to safely perform their field enforcement duties within the managed lane facility. Enforcement activities that include writing cita- tions and interacting with drivers are more easily accommo- dated on facilities with full-width shoulders; however, for facilities with narrower shoulders, and for enforcement activ- ities such as observation of vehicle eligibility or traditional traffic monitoring, additional treatments (e.g., enforcement monitoring and pullout areas) may be appropriate. Some states have developed design guidelines for those treatments, which are summarized in this section. The California HOV guidelines (53) call for the follow- ing enforcement area configurations, listed in order of preference: 1. Continuous paved median 14 ft or wider in both direc- tions for the length of the HOV facility should be built. If space is available, additional enforcement areas may be built in conjunction with the median. 2. When 14-ft continuous paved median shoulders are not possible, paved bi-directional enforcement areas spaced 2 to 3 mi apart should be built. A separation in the median barrier should be provided for motorcycle officers to patrol the HOV facility in both directions of travel. 3. Where median width is limited, some combination of 1 and 2 should be included. 4. Paved directional enforcement areas should be spaced 2 to 3 mi apart and staggered to accommodate both direc- tions when space limitations do not allow any of the above outlined considerations. 5. Where space is limited, directional enforcement areas should be located wherever right-of-way is available. Guidance in Washington’s Design Manual (55) states that enforcement of an inside HOV lane can be done with a mini- mum 10-ft inside shoulder. For continuous lengths of bar- rier exceeding 2 mi, a 12-ft shoulder is recommended for the whole length of the barrier. For inside shoulders less than 10 ft, enforcement and observation areas should be located at 1- to 2-mi intervals or based on the recommendations of the state police; these areas can also serve as refuge areas for dis- abled vehicles. Washington’s manual recommends observation points approximately 1300 ft before enforcement areas, though they may also be located just downstream of toll gantries within sight of the toll beacon. Observation areas can be designed to serve both patrol cars and motorcycles or motorcycles only. The designer should coordinate with the appropriate enforcement agency(ies) during the design stage to provide effective place- ment and utilization of the observation points. Median open- ings give motorcycle officers the added advantage of being able to quickly respond to emergencies in the opposing direction. The ideal observation point places the motorcycle officer at least 18 in. in the median so the officer can look down into a vehicle and see over the barrier. The enforcement area should be located on the right side for queue bypasses and downstream from the stop bar so the officer can be an effective deterrent. While these examples are based on enforcement of HOV lanes, principles of HOV enforcement can also be applicable to other types of managed lanes. The designer must consider

Source: Colorado DOT (88). Figure 63. Example design provisions for ITS and electrical conduits.

73 the specific needs of a particular facility when determining what elements to include in an enforcement design. Litera- ture on this topic, as it pertains to the design implications for implementing managed lanes, can be found in the tables in the eligibility validation and enforcement sections within the operations and maintenance chapter (Chapter 6). The sec- tion of this chapter that focuses on pullouts provides more guidance on the design of enforcement areas. Incident Management Incident management is a critical element for consider- ation in design of managed lanes. Often, right-of-way con- siderations limit the corridor envelope, making incident management more challenging. For projects where full shoul- ders can be developed both inside and outside of the travel lane, incident management is similar to any other freeway setting. However, consideration should be given to incident management regardless of the cross section involved. Incident management can be broadly divided into three major categories: • Crash or disabled vehicle. • Enforcement. • Natural or man-made disaster. A response to a crash or disabled vehicle on a managed lane is similar to a response to an incident on a general-purpose lane. If, however, the cross section is limited, it can add com- plexity to a response plan; several approaches are applied in the design that can be considered to account for unique char- acteristics of a limited cross section. These include: • Consideration of access openings or gates in the barrier (i.e., offsets) that allow for easier response on barrier- separated lanes. • Continuous illumination or spot illumination at access points. • Emergency pullouts or shoulders periodically provided if continuous shoulders cannot be provided. • Comprehensive camera surveillance to the traffic opera- tion center. • Parking or monitoring locations for dedicated tow services for very restricted lane designs and reversible and contra- flow treatments. To better address incidents for the entire freeway and managed lane, dynamic lane control through active traffic management can be employed to enable the managed lane to discontinue operation, if needed, to allow an incident to be resolved. In the case where a substantial amount of shoulder space was used to develop the managed lane, this design can be particularly appropriate. In essence, under conditions when there is no inci- dent, the managed lane provides the potential capacity that the shoulder represents if used as a travel lane, meaning that the additional capacity allowed by the use of the shoulder is usu- ally available, and, when it is not (i.e., when an incident occurs), conditions revert to what they would be without the managed lane being developed. Unless the managed lane is barrier sepa- rated, responding to a crash or other incident in the managed lane is similar to responding to an incident in the inside lane of a general-purpose facility. In this scenario, an operation is either better than or no worse than what would be experienced without the managed lane. This design creates an incentive to implement a managed lane even with a restricted cross section. Dynamic lane-use control may be possible in the general- purpose lanes. In this way, the facility operator can close the affected general-purpose lane(s) and open the managed lane(s) to all traffic. It should also be noted that the type of barrier or buffer plays a role in the effectiveness of this strat- egy. If a fixed barrier (e.g., a concrete barrier) is used, access to the managed lane to provide incident response can be an issue that is not present with buffer or pylon separation. This factor should be taken into account when selecting the sepa- ration treatment and design. Enforcement should also be considered in incident manage- ment. The business rules for the facility will need to address enforcement decisions. For instance, if only specialized vehi- cles such as buses or emergency vehicles are allowed to travel with no charge and all other vehicles are charged the same toll regardless of occupancy or vehicle type, enforcement is very simple and can be handled primarily through video enforce- ment coupled with double-white-line or other barrier-crossing violations enforced in the same way as other traffic viola- tions. If discounts or non-tolled usage based on vehicle type or vehicle occupancy are offered, enforcement becomes more complex, often substantially more complex. In these cases, an enforcement area to allow officers to pull over suspected viola- tors will likely need to be developed at periodic locations in the corridor. These areas should be designed to the extent possible to take advantage of areas where additional right-of-way may be available or easily obtained. Guidance on design of enforce- ment areas was provided earlier in this chapter. Managed lanes can provide a significant improvement to a corridor’s ability to handle emergency operations. Poten- tial scenarios should be developed in advance, and operating strategies should then be developed, as discussed in Chapter 6, Operations and Maintenance. As an example, in a coastal area subject to tropical storms, both sides of an expressway may be devoted to general evacuation in one direction under contraflow operation. However, in these cases, it is necessary to provide emergency personnel with a route into the affected area during the evacuation to respond to problems that can develop. In these situations, the managed lane may be able to

74 operate in the opposite direction of evacuation flow to allow responders to enter the evacuation area. Incident strategies will be based on the needs of the facility and the community the facility serves. Specific strategies and procedures will need to be developed that take into account the incidents that could occur on a particular corridor. Drainage and Hydraulic Needs Drainage considerations for managed lanes are similar, if not identical, to drainage considerations for freeways in gen- eral. For managed lanes located to the left of general-purpose lanes, drainage is a particular consideration if the opposing directions of travel are separated by a median barrier. In those conditions, the shoulder should also be designed to handle gutter spread and drainage. In many guidelines, that shoulder width is a minimum of 4 ft, and guidelines are also typically provided for the design and spacing of the drainage inlets, but there is a wide variety of practices in use by agencies across the country, and drainage preferences in one jurisdiction may not be acceptable in another. The designer must take into account the drainage design requirements in effect in the jurisdiction (e.g., a state DOT, county, or city) that apply to a particular managed lane facility.