Signalized Intersections

Description of Strategies

Objectives

The main goal of the objectives in this guide is the improvement in the safety of signalized intersections and their approaches. Safety improvement measures include geometric design modifications, changes to traffic control devices, enforcement, and education. Exhibit V-1 lists the objectives and the related strategies for improving safety at signalized intersections.

• Reduce frequency and severity of intersection conflicts through traffic control and operational improvements—Improvements to the method of assigning ROW at signalized intersections can reduce the potential for conflicts. This can be accomplished by modifying signal phasing, providing additional traffic control devices and pavement markings, and restricting turn movements. Improvements to traffic control can also benefit traffic operations and reduce emergency response time.

• Reduce frequency and severity of intersection conflicts through geometric improvements— Reducing the frequency and severity of vehicle-vehicle conflicts at intersections can reduce the frequency and severity of intersection crashes. This can be accomplished by incorporating geometric design solutions that separate through and turning movements at the intersection, restrict or eliminate turning maneuvers, and close or relocate intersections.

• Improve sight distance at signalized intersections—Provision of clear sight triangles in each quadrant of an intersection can minimize the possibility of crashes related to sight obstructions.

• Improve driver awareness of intersections and signal control—Some intersection-related collisions occur because one or more drivers approaching an intersection are unaware of the intersection until it is too late to avoid a collision. Improved signing and delineation and installation of lighting can help warn drivers of the presence of the intersection. In some situations, where other measures have not been effective, rumble strips may be used to get the driver's attention.

• Improve driver compliance with traffic control devices—Many accidents are caused or aggravated by drivers' noncompliance with traffic control devices or traffic laws at intersections. Both public education and enforcement have been shown to be effective in reducing traffic-law violations and consequently improving safety at intersections. Automated enforcement of traffic signals and speed limits is an increasingly common and cost-effective approach to improving driver compliance with traffic laws. At certain high-speed intersection approaches, implementing speed-reduction measures may provide an approaching driver with additional time to make safer intersection-related decisions.

• Improve access management near signalized intersections—Navigation, braking, and decision-making on intersection approaches creates additional workload on the driver. The presence of driveway access at or near a signalized intersection may confuse drivers using the intersection and create additional vehicle-vehicle conflicts. Measures to restrict driveways and to preclude cross median turning movements in close proximity to signalized intersections can effectively reduce or eliminate serious multivehicle conflicts.

• Improve safety through other infrastructure treatments—Other improvements can be made to the intersection to decrease frequency and severity of crashes at signalized intersections. These include improving pavement conditions, coordinating operation of signals near railroad crossings, and moving signal hardware out of the clear zone.

Ultimately, the driver is the target of all objectives, but specifically of those objectives relating to public education and traffic law-enforcement.

This section discusses each of the strategies listed in Exhibit V-1. The order in which the strategies are listed does not imply a priority with which they should be considered.

Most of the strategies are low-cost, short-term treatments to improve safety at signalized intersections, consistent with the focus of the entire AASHTO SHSP. For each of these, a detailed discussion of the attributes, effectiveness, and other key factors describing the strategy is presented below. Several higher-cost, longer-term strategies that have been proven to be effective in improving safety at signalized intersections are also presented in this section, but in less detail. While application of these is outside the implementation framework envisioned by the SHSP, their inclusion in this guide serves to complete the picture of proven, tried, and experimental strategies to improve safety at signalized intersections.

EXHIBIT V-1
Emphasis Area Objectives and Strategies

Objectives	Strategies
17.2 A Reduce frequency and severity of intersection conflicts through traffic control and operational improvements	17.2 A1 Employ multiphase signal operation (P, T) 17.2 A2 Optimize clearance intervals (P) 17.2 A3 Restrict or eliminate turning maneuvers (including right turns on red) (T) 17.2 A4 Employ signal coordination (P) 17.2 A5 Employ emergency vehicle preemption (P) 17.2 A6 Improve operation of pedestrian and bicycle facilities at signalized intersections (P, T) 17.2 A7 Remove unwarranted signal (P)
17.2 B Reduce frequency and severity of intersection conflicts through geometric improvements	17.2 B1 Provide/improve left-turn channelization (P) 17.2 B2 Provide/improve right-turn channelization (P) 17.2 B3 Improve geometry of pedestrian and bicycle facilities (P, T) 17.2 B4 Revise geometry of complex intersections (P, T) 17.2 B5 Construct special solutions (T)
17.2 C Improve sight distance at signalized intersections	17.2 C1 Clear sight triangles (T) 17.2 C2 Redesign intersection approaches (P)
17.2 D Improve driver awareness of intersections and signal control	17.2 D1 Improve visibility of intersections on approach(es) (T) 17.2 D2 Improve visibility of signals and signs at intersections (T)
17.2 E Improve driver compliance with traffic control devices	17.2 E1 Provide public information and education (T) 17.2 E2 Provide targeted conventional enforcement of traffic laws (T) 17.2 E3 Implement automated enforcement of red-light running (cameras) (P) 17.2 E4 Implement automated enforcement of approach speeds (cameras) (T) 17.2 E5 Control speed on approaches (E)
17.2 F Improve access management near signalized intersections	17.2 F1 Restrict access to properties using driveway closures or turn restrictions (T) 17.2 F2 Restrict cross-median access near intersections (T)
17.2 G Improve driver compliance with traffic control devices	17.2 G1 Provide public information and education (T) 17.2 G2 Provide targeted conventional enforcement of traffic laws (T) 17.2 G3 Implement automated enforcement of red-light running (cameras) (P) 17.2 G4 Implement automated enforcement of approach speeds (cameras) (T) 17.2 G5 Control speed on approaches (E)
P = proven; T = tried; E= experimental. A fuller explanation of P, T, and E appears in Section V. Several strategies have substrategies with differing ratings.

Types of Strategies

The strategies in this guide were identified from a number of sources, including recent literature, contact with state and local agencies throughout the United States, and federal programs. Some of the strategies are widely used, while others are used at a state or local level in limited areas. Some have been subjected to well-designed evaluations to prove their effectiveness. On the other hand, it was found that many strategies, including some that are widely used, have not been adequately evaluated.

The implication of the widely varying experience with these strategies, as well as the range of knowledge about their effectiveness, is that the reader should be prepared to exercise caution in many cases before adopting a particular strategy for implementation. To help the reader, the strategies have been classified into three types, each identified by letter symbol throughout the guide:

• Proven (P): Those strategies that have been used in one or more locations and for which properly designed evaluations have been conducted which show them to be effective. These strategies may be employed with a good degree of confidence, with the understanding that any application can lead to results that vary significantly from those found in previous evaluations. The attributes of the strategies that are provided will help the user make judgments about which ones may be the most appropriate for their particular situation(s).

• Tried (T): Those strategies that have been implemented in a number of locations, and may even be accepted as standards or standard approaches, but for which there have not been found valid evaluations. These strategies, while in frequent, or even general, use, should be applied with caution, carefully considering the attributes cited in the guide, and relating them to the specific conditions for which they are being considered. Implementation can proceed with some degree of assurance that there is not likely to be a negative impact on safety, and very likely to be a positive one. It is intended that as the experiences of implementation of these strategies continues under the AASHTO SHSP initiative, appropriate evaluations will be conducted. As more reliable effectiveness information is accumulated to provide better estimating power for the user, any given strategy labeled “tried” can be upgraded to a “proven” one.

• Experimental (E): Those strategies representing ideas that have been suggested, with at least one agency considering them sufficiently promising to try them as an experiment in at least one location. These strategies should be considered only after the others have proven not to be appropriate or feasible. Even where they are considered, their implementation should initially occur using a very controlled and limited pilot study that includes a properly designed evaluation component. Only after careful testing and evaluations show the strategy to be effective should broader implementation be considered. It is intended that as the experiences of such pilot tests are accumulated from various state and local agencies, the aggregate experience can be used to further detail the attributes of this type of strategy, so that it can be upgraded to a “proven” one or identified as being ineffective and not worthy of further consideration.

Related Strategies for Creating a Truly Comprehensive Approach

The strategies listed above in Exhibit V-1 and described in detail in the remainder of Section V are considered unique to this emphasis area. However, to create a truly comprehensive approach to the highway safety problems associated with signalized intersections, it is recommended that additional strategies be included as candidates in any program planning process. These additional strategies are of five types:

• Public Information and Education Programs (PI&E): Many highway safety programs can be effectively enhanced with a properly designed PI&E campaign. The primary experience with PI&E campaigns in highway safety is to reach an audience across an entire jurisdiction or a significant part of it. However, it may be desirable to focus a PI&E campaign on a location-specific problem, such as an individual intersection or corridor with a history of severe crashes related to red-light running. While this is a relatively untried approach compared with areawide campaigns, use of roadside signs and other experimental methods may be tried on a pilot basis.

  Within this guide, PI&E campaigns, where application is deemed appropriate, are usually used in support of some other strategy. In such a case, the description for that strategy will suggest this possibility (in the exhibits, see the attribute area for each strategy entitled “Associated Needs”). In some cases, where PI&E campaigns are deemed unique for the emphasis area, the strategy is explained in detail. As additional guides are completed for the AASHTO plan, they may address the details regarding PI&E strategy design and implementation.

• Enforcement of Traffic Laws: Well-designed and well-operated law enforcement programs can have a significant effect on highway safety. It is well established, for instance, that an effective way to reduce crashes and their severity is to have jurisdictionwide programs that enforce an effective law against driving under the influence of alcohol (DUI) or driving without seatbelts. When that law is vigorously enforced with well-trained officers, the frequency and severity of highway crashes can be significantly reduced. This should be an important element in any comprehensive highway safety program.

  Enforcement programs, by the nature of how they must be performed, are conducted at specific locations. The effect (e.g., lower speeds, greater use of seat belts, and reduced impaired driving) may occur at or near the specific location where the enforcement is applied. This effect can often be enhanced by coordinating the effort with an appropriate PI&E program. However, in many cases (e.g., speeding and seatbelt usage) the impact is areawide or jurisdiction-wide. The effect can be either positive (i.e., the desired reductions occur over a greater part of the system) or negative (i.e., the problem moves to another location as road users move to new routes where enforcement is not applied). Where it is not clear how the enforcement effort may impact behavior or where it is desired to try an innovative and untried method, a pilot program is recommended.

  Within this guide, where the application of enforcement programs is deemed appropriate, it is often in support of some other strategy. Many of those strategies may be targeted at either a whole system or a specific location. In such cases, the description for that strategy will suggest this possibility (in the exhibits, see the attribute area for each strategy entitled “Associated Needs”). In some cases, where an enforcement program is deemed unique for the emphasis area, the strategy will be explained in detail. As additional guides are completed for the AASHTO plan, they may address the details regarding the design and implementation of enforcement strategies.

• Strategies to Improve Emergency Medical and Trauma System Services: Treatment of injured parties at highway crashes can have a significant impact on the level of later treatment and length of time in which an individual undergoes treatment. This is especially true when it comes to timely and appropriate treatment of severely injured persons. Thus, a basic part of a highway safety infrastructure is a well-based and comprehensive emergency care program. While the types of strategies that are included here are often thought of as simply support services, they can be critical to the success of a comprehensive highway safety program. Therefore, for this emphasis area, an effort should be made to determine if there are improvements that can be made in how emergency medical services interact with signalized intersections, especially for programs that are focused upon location-specific (e.g., corridors) or area-specific (e.g., rural areas) issues. As additional guides are completed for the AASHTO plan, they may address the details regarding the design and implementation of emergency medical systems strategies.

• Strategies Directed at Improving the Safety Management System: There should be in place a sound organizational structure, as well as infrastructure of laws, policies, etc., to monitor, control, direct, and administer a comprehensive approach to highway safety. It is important that a comprehensive program not be limited to one jurisdiction, such as a state Department of Transportation (DOT). Local agencies often have jurisdiction over the majority of the road system and are responsible for its related safety problems. They know better than others do what the problems are. As additional guides are completed for the AASHTO plan, the guides may address the details regarding the design and implementation of strategies for improving safety management systems.

• Strategies Detailed in Other Emphasis Area Guides: Several of these objectives and many of the corresponding strategies are applicable to unsignalized intersections as well as signalized ones. The discussion in this guide of these overlapping strategies is based upon the Unsignalized Intersection guide. Strategies that overlap between these two guides are discussed briefly in this section, and the Unsignalized Intersection guide should be consulted for more details. In addition, there are many treatments for signalized intersections that would improve safety for pedestrians, bicyclists, and older drivers. The pedestrian and older driver guides should be consulted for additional information. Any program targeted at the safety problem covered in this guide on signalized intersections should be created with consideration given to potentially appropriate strategies in these other guides.

Objective 17.2 A—Reduce Frequency and Severity of Intersection Conflicts through Traffic Control and Operational Improvements

Virtually all traffic signal timing and phasing schemes are established with the primary objective being the efficient movement of traffic. Certain timing, phasing, and control strategies can produce safety benefits with only marginal adverse effects on delay or capacity. Low-cost improvements to signalized intersections that can be implemented in a short time period include revising the signal phasing and/or operational controls at the intersection to explicitly address safety concerns. Signalization improvements may include adding phases, lengthening clearance intervals, eliminating or restricting higher-risk movements, and coordinating signals. A review of crash history at a specific signalized intersection can provide insight into the most appropriate strategy for improving safety at the intersection. See the presentation and discussion of the Model Implementation Process, Step 1, for further details. In particular, guidelines linking crash types to candidate improvement strategies are useful (See Appendix 10).

Strategy 17.2 A1: Employ Multiphase Signal Operation (Combination of Tried and Proven Strategies)

General Description

This strategy includes using protected left-turn phases and split phases.

A two-phase signal is the simplest method for operating a traffic signal, but multiple phases may be employed to improve intersection safety. Left turns are widely recognized as the highest-risk movements at signalized intersections. Protected left-turn phases (i.e., the provision for a specific phase for a turning movement) significantly improve the safety for left-turn maneuvers by removing conflicts with the left turn.

Split phases, which provide individual phases for opposing approaches may also increase the overall delay experienced at an intersection. However, this strategy may improve intersection safety, as it allows conflicting movements to proceed through the intersection independently, on separate phases.

Implementation of improvements to signal phasing may necessitate the replacement of older electromechanical signal controllers. Even if not necessary, replacing the controller should be considered as it may be more cost-effective to implement the changes at the same time as replacing the controller.

Use Protected Left Turns

The safety problems that left-turning vehicles encounter arise from three sources of conflict:

• Opposing through traffic,

• Through traffic in the same direction, and

• Crossing vehicular and pedestrian traffic.

These conflict types often produce angle, sideswipe same direction, and rear-end crashes. There are several treatments that could alleviate operational and safety impacts of—and on—left-turn traffic. Protected left-turn phases are warranted based on such factors as turning volumes, delay, visibility, opposing vehicle speed, distance to travel through the intersection, and safety experience of the intersections. Agency policies on the specific thresholds of each of these factors vary in the United States. There are several geometric and operational characteristics of intersections that should be analyzed when considering which type of left-turn signal phasing to use to accommodate left turns (turning volumes, opposing through volumes, pedestrian crossing volumes, approach speeds, sight distance, number of lanes, delay, type and nature of channelization, and crash experience).

There are various options available for controlling left turns with signals: permitted, protected only, and protected/permitted (including both lead-protected/permitted and lagprotected/permitted). Several Web sites are available that provide additional information on signal phasing:

• http://www.webs1.uidaho.edu/niattproject/

• http://www.dot.state.mn.us/metro/trafficeng/dsg_crse/chap21.html#_Toc429824696

The use of “protected/permitted” phasing represents a compromise between fully protected phasing and permitted-only phasing. This operational strategy has several advantages, the most important being the reduction in delay for left-turning vehicles achieved by permitting left turns while the opposing through movement has a green indication. Other benefits include less green time needed for protected left turns (and hence more time for other highpriority movements) and the potential for improved arterial progression. The safety performance of protected/permitted left-turn phases is not as good as that of protected-only phases, due to the increased exposure of left-turning and opposing through vehicles to conflicts with each other during the permitted phase. Dual or triple left-turn lanes should only operate with protected turn phases.

In terms of explicit concern for safety, protected-only phasing may be the best option. A study by Shebeeb (1995) showed that the left-turn signal phases that provide the greatest operational benefit to left-turning vehicles, with respect to stopped delay, increase the crash risk for left-turning vehicles the most. Additional guidance on choosing a type of left-turn phasing is summarized in NCHRP Synthesis 225: Left-Turn Treatments at Intersections (Pline, 1996).

The choice of lead versus lag phasing for protected left-turn phases depends on intersection capacity and the presence of, or desire for, coordinated system timing. Providing the leftturn arrow before the conflicting through movement receives a green indication (“lead” left turn) minimizes the conflicts between left-turning and through vehicles. With a “lag” leftturn phase, however, left-turning vehicles are given the opportunity to turn during the permissive portion of the cycle, which may allow clearing all or part of the left-turn queue, resulting in a shorter lag phase or eliminating the need for it during that specific cycle. A study of intersections in Kentucky found a higher average number of crashes per approach for protected/permitted phasing schemes having lag left turns (2.07 crashes per 1,000 leftturning and opposing vehicles) than for those having lead left turns (1.27 crashes). (Stamatiadis et al., 1997). On the other hand, a study of intersections in Arizona concluded that the potential for left-turn head-on crashes is not high enough to be a main factor in determining whether to use lead or lag left-turn phasing (Box and Basha, 2003).

EXHIBIT V-2
Strategy Attributes for Use of a Protected Left-Turn Signal Phase (P)

Attribute	Description
Technical Attributes
Target	The strategy is targeted at reducing the frequency of angle collisions resulting from conflicts associated with left-turn maneuvers at signalized intersections involving leftturning vehicles and opposing through vehicles. A properly timed protected left-turn phase can also help reduce rear-end and sideswipe crashes between left-turning vehicles and the through vehicles behind them.
Expected Effectiveness	Various studies have proven that installing protected left-turn phases improves leftturn safety due to the decrease in potential conflicts between left-turning and opposing through vehicles. The isolation of left-turning traffic usually reduces rear-end, angle, and sideswipe crashes, as well as improves the flow of through traffic. A protected/permitted left-turn phase has not been shown to provide the higher degree of safety of a protected-only phase, but it is safer than permitted-only phasing. Given the wide range of conditions at intersections used for studying the effectiveness of leftturn phases, a consensus on the extent of this effectiveness has not been reached. Consideration may be given to adding a protected left-turn phase when left-turn lanes are constructed. FHWA's Signalized Intersections: Informational Guide (to be published in 2004) provides a summary of studies of the effectiveness of adding leftturn lanes and protected left-turn phases and concludes that providing both a left-turn phase and left-turn lane appears to provide the most safety benefit. California reported a 35-percent average reduction in total crashes when left-turn lanes were constructed and left-turn phases were implemented, as opposed to a 15-percent reduction when left-turn lanes were installed without a separate left-turn phase (Neuman, 1985). Overlapping the adjacent right-turn phase with the protected left-turn phase will allow for improved operation of the right-turn movements at the intersection.
Keys to Success	This strategy applies only where a separate left-turn lane exists (see below for discussion of split phasing, which may apply where separate left-turn lanes are not present.) The overall length of the turn lane is a key element in the design of the lane. A lane that does not provide enough deceleration length and storage space for leftturning traffic may cause the turn queue to back up into the adjacent through lane. This can contribute to rear-end and sideswipe crashes, as well as adversely affect delay for through vehicles. NCHRP Synthesis 225: Left-Turn Treatments at Intersections (Pline, 1996) summarizes recent guidance on determining left-turn phasing (protected, protected/permitted, leading, lagging, etc). This information is based on both traffic volume data and crash histories. Appropriate protected left-turn signal indications should be used to communicate the signal phasing to drivers. Several experimental signal displays are being used across the country (see Appendix 1).
Potential Difficulties	A separate phase for the left-turn movement may reduce delay for the vehicles turning left but could result in more overall intersection delay, because other movements will lose green time or gain more red time and because the total signal cycle length may increase. The length of signal phase and cycle length should be compatible with the length of the left-turn lane. Turn lanes that are too short may be blocked by through-vehicle queues, making the lane inaccessible and also negating the effectiveness of a lead left-turn phase. Provision of a left-turn lane on an intersection approach may involve restricting left turns in and out of driveways on the intersection approach. The strategies in this guide in the access management objective should be consulted (see Objective 17.2 F).
Appropriate Measures and Data	Key process measures include the number of intersections for which protected leftturn phases are implemented and the number of conflicts eliminated by the improvements. Crash frequency and severity by type of crash are key safety effectiveness measures. It is especially important to identify crashes related to left turns (angle, rear-ends). Crash frequency and severity data are needed to evaluate such improvements. Traffic volume data are needed to represent exposure, especially the volumes of left-turn movements of interest and the opposing through volumes. Delay data are needed to determine the operational impacts of a change.
Associated Needs	There is no need for special public information and education programs, as most drivers understand the operation of a protected left turn. However, signs may be temporarily erected on the approach to an intersection for which a phasing plan has been significantly altered to help frequent users of the intersection be aware of the change and not violate expectancy of the familiar driver.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	The signalization policies of many agencies are primarily driven by traffic operational and delay/capacity concerns. Highway agencies should review their traffic engineering and design policies regarding the use of, or warrants for, protected left-turn phases to ensure that appropriate safety-based action is being taken on routine projects. Highway agencies and other agencies should ensure that their policies for new or reconstructed intersections incorporate provisions for protected left-turn lanes and signal phases, where applicable. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas. Where alternatives may involve restricting access, it will be important to involve those potentially affected in the early stages of planning.
Issues Affecting Implementation Time	Implementing this strategy may range from a few months to 3 or 4 years. Protected-only phasing can be implemented only where separate left-turn channelization exists. Where the intersection geometry already exists, the cost can be very small (engineering and technician time to install the phasing scheme). In some cases, upgrading of the existing signal equipment, including the controller, may be necessary. Even where no such channelization exists, it may be possible to re-stripe an approach to provide it. At other locations, lengthening the left-turn lane, widening the roadway, acquiring additional ROW, or redesigning the horizontal and vertical alignment may be needed in conjunction with changes in signal operation policies. The latter types of projects require time for design and construction.
Costs Involved	Costs may be highly variable and may depend on the condition and flexibility of the existing traffic signals and controller. If the existing traffic signal only requires a minor modification to allow for a protected left-turn phase, then the cost would be low. If a completely new traffic signal is needed to accommodate the protected left-turn phase, then the cost could be higher. In addition to the costs of the equipment needed for the signal, expenditures are needed for advance warning signing and signs and markings needed at the intersection (such as a “Stop Here On Red” sign). Similarly, costs would be higher if additional dedicated left-turn lanes are required; these costs may include right-of-way, pavement, pavement markings, and lane use signs.
Training and Other Personnel Needs	None identified.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with other strategies for improving safety at signalized and unsignalized intersections. Most notably, strategies concerning addition of left-turn lanes would be compatible
Other Key Attributes to a Particular Strategy	None identified.

Use Split Phases

Certain geometric configurations, such as left-turn travel paths that overlap with the opposing left-turn path, may require the use of split phasing at an intersection. Split phasing allows opposing movements on the same roadway to proceed through the intersection at different times and is a way to address several geometric situations that pose safety problems for vehicles on opposite approaches (see Exhibit V-3). These include the following:

• Skewed intersections,

• Intersections with a large deflection angle for the through movement,

• Wide medians,

• Intersections too small to allow simultaneous left turns (limited ROW),

• Intersections with lanes shared by left-turn and through movements (i.e., without separate left-turn lanes),

• Intersections with significantly unbalanced opposing left-turn volumes, and

• Intersections on a divided highway with different profiles.

EXHIBIT V-3
Split Phasing on One Intersection Approach
Source: Federal Highway Administration, in press.

EXHIBIT V-4
Strategy Attributes for Use of a Protected Left-Turn Signal Phase (P)

Attribute	Description
Technical Attributes
Target	This strategy targets crashes that occur related to opposing movements proceeding on the same phase through an intersection with complex geometry or lane assignment. Crash types related to this situation include sideswipe between opposing left turns, rear end, head on, and angle.
Expected Effectiveness	Though studies have not conclusively proven that implementation of split phases reduces fatalities and severe injuries at signalized intersections, the elimination of conflicts can logically be expected to reduce crashes. Using split phases to separate opposing traffic can be expected to greatly reduce the sideswipe, rear-end, and angle conflicts and the collisions associated with the geometric situation that contributes to the conflicts between the opposing vehicles. The effectiveness in reducing crashes involving left-turning vehicles should be similar to that of adding a protected-only leftturn phase. With no movements conflicting with vehicles on a given approach, angle, head-on and sideswipe-opposite-direction crashes should be eliminated. Rear-end and sideswipe-same-direction crashes may not be completely eliminated, but some of these crashes may be related to congestion or other factors rather than conflicts with vehicles moving in opposing directions.
Keys to Success	A key to success is balancing the safety benefits of split phases with the operational disadvantages, such as increased lost time and intersection delay. Care should be taken to examine other potential strategies that could provide the same safety benefit, but with less operational cost. Such strategies might include restricting turning maneuvers (Strategy 17.2 A3), improving left-turn channelization (Strategy 17.2 B1), and revising geometry of complex intersections (Strategy 17.2 B4).
Potential Difficulties	The use of split phasing will generally result in less efficient intersection operations, depending on the intersection characteristics. Increasing the number of phases usually requires a longer signal cycle and increases lost time, resulting in a longer overall intersection delay. The delay on an approach could be increased to a point where queues will exceed available storage lengths. This should be a factor to consider in any change of phasing and timing. Adverse effects on arterial progression may also result from implementation of this strategy. (See Strategy 17.2 A4). Driver error is a potential problem associated with this strategy, specifically when first implemented. Changes in signal phasing may violate the expectancy of drivers familiar with the intersection. Since drivers understand the operation of traffic signals in general, this should not be a significant problem.
Appropriate Measures and Data	Process measures include the number of intersections for which a split signal phase is implemented and the number of conflicts affected by the improvement. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to targeted movements at the intersection should be analyzed separately. Traffic volume data are needed to represent exposure. Delay data are needed to estimate the operational impacts of a change.
Associated Needs	Except for temporary warning of significant phasing changes, public information should not be needed when implementing this strategy. Drivers are familiar with the operation of traffic signals, and the effect of split phasing on driver expectancy is not anticipated to be a serious issue.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding signal phasing to ensure that appropriate action is being taken when split phases may provide a safety benefit. Any highway agency with jurisdiction over signalized intersections can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas.
Issues Affecting Implementation Time	Implementation of split signal phasing could vary from a few days to a few months, depending upon the condition and flexibility of the existing traffic signal. Should anticipated queuing exceed available storage lengths, much longer time would be required for reconstruction of the approach(es) in need of additional storage space.
Costs Involved	Costs are variable and may depend upon the condition and flexibility of the existing traffic signal. If the existing traffic signal only requires a minor modification to allow for split phasing, then the cost would be low. If a completely new traffic signal is needed, the cost would of course be higher, due to signal design, timing, and equipment costs. Reconstruction of storage lanes may also result in major costs.
Training and Other Personnel Needs	The safety benefits of split phasing should be addressed in agency training on intersection safety and traffic signals.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with other approaches for improving safety at signalized intersections.
Other Key Attributes to a Particular Strategy	None identified.

Strategy 17.2 A2: Optimize Clearance Intervals (P)

General Description

The clearance interval is the portion of a signal cycle between the end of a green phase and the beginning of the next green phase for a conflicting movement. Clearance times provide safe, orderly transitions in ROW assignment between conflicting streams of traffic. The clearance interval can include both yellow and all-red time between conflicting green phases.

There is no standardized method for determining clearance intervals. Clearance intervals are a function of operating speed, the width of the intersection area, lengths of vehicles, and driver operational parameters such as reaction, braking, and decision-making time. ITE has developed an equation for determining the length of the yellow change interval. Many agencies use rule-of-thumb methods as well. See Appendix 2 for more information on establishing clearance intervals.

Clearance intervals that are too short in duration can contribute to rear-end crashes related to drivers stopping abruptly and right-angle crashes resulting from signal violations. One study showed clearance intervals shorter than those calculated using the ITE equation have higher rear-end and right-angle crash rates than intersections with timings that exceed the ITE value (Zador et al., 1985). In the extreme, a too-short interval can result in drivers operating at the legal speed limit being forced to violate the red phase. A study by Retting et al. (2000) noted that signal intervals that are considered too short are associated with vehicle conflicts and red-light running.

Increasing clearance intervals may improve safety at signalized intersections where the existing yellow (or yellow plus red) change intervals do not allow drivers adequate time to react to the reassignment of ROW. Longer clearance intervals may also be effective at intersections with significant physical size, to allow drivers to clear the intersection before the opposing traffic enters. See Appendix 2 for more information on establishing clearance intervals. A detailed discussion on yellow and all-red intervals is provided in Making Intersections Safer: A Toolbox of Engineering Countermeasures to Reduce Red-Light Running (McGee, 2003; available online at http://www.ite.org/library/redlight/MakingInt_Safer.pdf).

EXHIBIT V-5
Strategy Attributes for Optimizing Clearance Intervals (P)

Attribute	Description
Technical Attributes
Target	The target of this strategy is crashes related to clearance interval lengths that are too short for a particular intersection. These crashes include angle crashes between vehicles continuing through the intersection after one phase has ended (possibly due to being in the dilemma zone as the clearance interval started) and the vehicles entering the intersection on the following phase. Rear-end crashes may also be a symptom of short clearance intervals. A vehicle stopping at a signal may be rearended by a vehicle following it when the following driver expected to be able to proceed through the intersection during a longer clearance interval.
Expected Effectiveness	This strategy is proven effective in reducing multivehicle crashes at signalized intersections. A study of signalized intersections in one city in New York found a 9-percent reduction in multivehicle and a 12-percent reduction in injury crashes at intersections where the duration of the change intervals was lengthened to meet ITE recommendations. The crash risk for rear-end and angle crashes did not change significantly. The same study showed a 37-percent reduction in crashes involving pedestrians or bicyclists. The authors explained that pedestrian- and bicycle-related crashes may be more affected by changes in clearance interval timing because many pedestrians and bicyclists will enter the intersection during the change interval before they are given a walk signal. (Retting et al., 2000).
Keys to Success	A clearance interval should not be so long as to encourage disrespect in drivers for the interval, thereby contributing to red-light running and even more severe crashes, nor so short as to violate driver expectancy regarding the length of the interval, resulting in abrupt stops and possible rear-end crashes.
Potential Difficulties	The Retting et al. (2000) study cited above suggests that drivers do not generally assume that longer change intervals at one or more locations will mean that they will be used at all signalized intersections. Therefore expectancy problems related to this are not likely to be experienced at intersections having shorter change intervals. Further research may be needed, however, to provide more evidence that the effect of lengthening a change interval does not create general expectancies among drivers. As clearance intervals are increased, there will usually be an attendant increase in the cycle length and delay. Thus, an intersection may become safer, but the resulting level of delay increases, which may raise objections from the traveling public. Moreover, increased cycle lengths and delay may have adverse operational effects on one or more approaches (e.g., left-turn-lane overflow or blockage, loss of progression, queue collision with adjacent intersections). Longer cycle lengths may also lessen the effectiveness of a signal progression scheme for a route or corridor. Any of the above difficulties may create a degradation in safety away from the intersection, thus potentially negating some of the benefits of improved clearance times.
Appropriate Measures and Data	A key measure of the implementation process is the number of signalized intersections for which clearance intervals are optimized. Crash frequency and severity by type of crash are also key safety effectiveness measures. It is especially useful to separately analyze crashes by movement or type. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to a clearance interval (right angle and rear end) should be analyzed. Traffic volume data are needed to represent exposure. Delay data are needed to assess operational impacts.
Associated Needs	Except for temporary warning of timing changes, there is no need for special public information and education programs relating to signal clearance intervals. Public information and education campaigns for red-light running should encompass this issue.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway and other agencies should ensure that their signal design policies provide guidance and allow some flexibility in clearance interval length. Tort liability is an issue to be considered when selecting change intervals. Agencies responsible for traffic signals have paid large settlements in cases where clearance intervals did not meet recommended values. Examples of this are discussed in the ITE Traffic Safety Toolbox (Institute of Transportation Engineers, 1999b). Highway agencies should review their traffic engineering policies regarding clearance intervals to ensure that appropriate action is being taken on projects. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas. Very long clearance intervals may be required in rural areas or at intersections with very high-speed traffic.
Issues Affecting Implementation Time	Implementation time is low for changing the length of a clearance interval. Engineering studies, development of retiming plans, and field implementation are required.
Costs Involved	Costs for changing the length of a clearance interval will be low. The design of the new signal timing and the reprogramming of the signal should be the only costs.
Training and Other Personnel Needs	Effective use of clearance intervals, including the length of the red clearance interval, should be addressed in highway agency training concerning intersection design and operation. Agency engineering staff should be aware of the legal implications of dilemma zones and clearance intervals in their jurisdiction.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Optimizing the length of the clearance interval is compatible with other strategies to improve signalized intersection safety. Note that some strategies to improve safety may increase required clearance intervals. One such strategy is widening an approach to add left-turn lanes. An alternative to optimization of clearance interval is implementation of measures to reduce speeds on one or more approaches.
Other Key Attributes to a Particular Strategy	None identified.

Information on Current Knowledge Regarding Agencies or Organizations That Are Implementing This Strategy

Agencies are utilizing technology to automatically extend signal intervals in order to aid drivers approaching a signalized intersection in determining whether to stop or proceed through the intersection. The “dilemma zone” for an intersection is a specific road segment, prior to the intersection, in which the driver will be able neither to stop safely before entering the intersection nor to proceed through the intersection without violating the red indication. The ITE Traffic Engineering Handbook (Pline, 1999) contains a more detailed description of the dilemma zone, including zone boundaries based on approach speed. Agencies are implementing systems to protect drivers in the dilemma zone by extending the green or red interval. Refer to Appendix 3 for additional details. European countries have also implemented treatments for protecting drivers in the dilemma zone (see Appendix 4).

Strategy 17.2 A3—Restrict or Eliminate Turning Maneuvers (Including Right Turns on Red) (T)

General Description

This strategy includes restricting or eliminating left- or right-turning maneuvers using channelization or signing and prohibiting right turns on red (RTOR).

Safety at some signalized intersections can be enhanced by restricting or prohibiting turning maneuvers, particularly left turns. This strategy can be applied during certain periods of the day (such as peak traffic periods) or by prohibiting particular turning movements altogether. This strategy may be appropriate where a turning movement is considered to be “high risk” and other strategies (such as left-turn channelization or retiming of signals) are impractical or not possible to implement.

Crashes related to turning maneuvers include angle, rear-end, pedestrian, and sideswipe (involving opposing left turns) type crashes. If any of these crash types are an issue at an intersection, restriction or elimination of the turning maneuver may be the best way to improve the safety of the turn.

Restrict or Eliminate Turning Maneuvers Using Channelization or Signing. Turn restrictions and prohibitions can be implemented by channelization or signing. Raised concrete channelization or flexible delineators can be used to physically prevent drivers from making restricted maneuvers. Turning prohibitions or restrictions implemented with signing alone will most likely require some periodic enforcement. The cost of enforcement should be considered when discussing methods for restricting or prohibiting turns.

Prohibit Right Turns on Red. Prohibition of RTOR can help reduce crashes related to limited sight distance and pedestrians that involve right-turning vehicles. This strategy can also help reduce the frequency and severity of crashes between vehicles turning right on red and vehicles approaching from the left on the cross street or turning left from the opposing approach. Prohibition of RTOR may also be a safety-effective strategy where weaving or other conflicts are evident downstream of the right turn. This strategy can be implemented with signing, although enforcement is often needed to realize the potential benefits of the new regulation. Prohibition of RTOR at specific intersections can be implemented during certain times of the day (such as when pedestrians are more likely to be present). Also, supplemental sign plaques prohibiting RTOR, when pedestrians are present, have been used to help protect pedestrians.

EXHIBIT V-6
Strategy Attributes for Prohibiting or Eliminating Turning Movements (Including RTOR) (T)

Attribute	Description
Technical Attributes
Target	The target of this strategy is crashes related to turning maneuvers including angle, rearend, pedestrian, and sideswipe (involving opposing left turns) type crashes. For RTOR, the target of this strategy is right-turning vehicles that are involved in rear-end or angle crashes with cross-street vehicles approaching from the left or vehicles turning left from the opposing approach, as well as crashes involving pedestrians. See the pedestrian guide for additional discussion on strategies for improving pedestrian safety.
Expected Effectiveness	Though there are no studies proving that prohibition of turning movements reduces fatal and severe crashes at signalized intersections, prohibition of left-turning movements—if enforced—would be expected to eliminate crashes involving left turns over the time period of the prohibition based on the assumption that no drivers will violate the restriction. Note, however, that a complete assessment of the effect of a turn restriction or prohibition on safety requires consideration of the impacts on alternative routes to which the traffic that desires to make the affected turn is diverted. Also, the benefit of restricting turn movements may be reduced by an increase in accidents related to formation of queues (such as rear-end collisions) at alternative turn locations. No data on the effectiveness of prohibiting RTOR are available, but it is expected that prohibition of RTOR will eliminate crashes related to vehicles making that turn during the time period the restriction is in effect, assuming that no drivers violate the restriction. Crashes related to right turns that occur on green would not be affected by prohibiting RTOR. Fleck and Lee (2002) report in an ITE Journal article that RTOR collisions are not always reported as such; rather, they are often coded as violation of pedestrian ROW, driving under the influence, or other types of violations. Therefore, it is important to carefully analyze crash histories, especially those involving pedestrians, to determine the problem's nature. Retting et al. (2002) report that prohibition of RTOR during certain hours of the day is effective in reducing RTOR without stops. However, prohibition of RTOR when pedestrians are present is a much less effective strategy. These results are based on a study of intersections in Arlington, Virginia, but it could be expected that similar urbanized areas would experience the same results. A reduction in drivers turning right on red without stopping could lead to a lower number of pedestrian crashes.
Keys to Success	A key to success of the prohibition of left turns is the provision for safe and adequate alternative locations to make the left turn in close proximity to the intersection where the prohibition is placed. As noted above, a careful traffic engineering study should be made to ensure that the safety and operational problems calling for the prohibition are not merely relocated elsewhere. It will be important to include stakeholders in the planning and implementation of this strategy. Law enforcement agencies in the jurisdiction should be partners in the effort. If access to properties may be negatively affected, representatives of those involved should be included in the process. Affected transit agencies should also be involved. With respect to RTOR prohibition, a key to success is to establish that prohibition of RTOR is justified due to an existing pattern of right-turn collisions. RTOR prohibitions should be provided only in areas where the restriction could be beneficial, such as urbanized areas with high pedestrian volumes, at intersections with concentrations of children (e.g., enroute and adjacent to schools, parks, playgrounds), or where experience has shown a high number of crashes involving vehicles attempting to turn right on red. Otherwise, installation of a RTOR prohibition is unlikely to provide substantial safety benefits, while possibly contributing to driver disrespect for the prohibition. Enforcement of the prohibition is also important to the success of the strategy.
Potential Difficulties	Prohibition of left turns at a major intersection may be difficult to justify, unless the leftturn volumes are very low. Refer to Strategy 17.2 F2 for discussion on restricting median left turns. If at all possible, it is generally preferred to more safely accommodate the turning movement at the point where the driver desires to turn than to displace the turn activity to an alternative location. Restriction and prohibition of turning maneuvers are discussed in more detail in the unsignalized intersection guide. However, issues in implementing turn prohibitions become more complex in higher-volume suburban and urban signalized intersections. Drivers familiar with the intersection might fail to notice the prohibition of RTOR when the restriction is first put into place. This is expected to be a common occurrence where other intersections within the jurisdiction permit RTOR operation. Additional signing or public information and educational materials may help alleviate this.
Appropriate Measures and Data	Process measures include the number of intersections for which a prohibition has been implemented, the percentage of intersections at which there is a turn problem for which a prohibition has been implemented, and the number of conflicts affected by the improvements. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to targeted turning movements at the intersection should be analyzed separately. Traffic volume data are needed to represent exposure. It will be important that the analysis include all intersections potentially affected by the restriction, including those to which turning traffic will be diverted.
Associated Needs	There is a need to inform the public about the change in regulations at the intersection and about the safety benefits of the prohibition. Informing the public of a RTOR prohibition takes on added significance when other intersections within the jurisdiction permit RTOR operation.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding RTOR and policies for restricting turns and prohibiting RTOR. When planning turn restrictions, it is important to include public transport agencies, due to the potential effects on bus transit. Either rerouting a bus route or allowing buses to make the turns that other vehicles are prohibiting from making are options, should this be an issue.
Issues Affecting Implementation Time	Implementation of the turn restriction or prohibition could vary from a few days to a few months, depending upon the extent of public information and education provided.
Costs Involved	Costs may be variable—turn restrictions can be implemented with low-cost signing, but enforcement of the regulation and PI&E campaigns regarding the new regulation will increase costs.
Training and Other Personnel Needs	Turn restrictions, including RTOR, should be incorporated into agency training on intersection operations and safety.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with other strategies for improving safety at signalized and unsignalized intersections. Refer in particular to the pedestrian guide for a range of strategies aimed at pedestrian safety, many of which can work in concert with implementation of turn restriction, specifically with a RTOR prohibition.
Other Key Attributes to a Particular Strategy	RTOR reduces delay for right-turning vehicles and, in fact, was encouraged by the federal government in the 1970s as an energy conservation measure. Previous research has shown that RTOR movements result in a 5-percent reduction in fuel consumption on urban streets. RTOR from an auxiliary lane has a negligible impact on delays if the average gapacceptance is less than 15 sec per vehicle. If the cross flow does not exist or is light, then multiple RTOR can be performed at a rate of one vehicle per 4.7 seconds. This could result in a significant reduction in delays. If 10 percent of the approaching flow turns right, then the RTOR has little influence on right turn delay. If 40 percent of the approaching flow turns right, then the RTOR movements may reduce delays significantly. However, RTOR is not likely to reduce delays significantly if the saturation ratio of the cross flow is greater than 0.6 sec and the delays without RTOR are less than 30 sec per vehicle.

Strategy 17.2 A4: Employ Signal Coordination (P)

General Description

Signal coordination has long been recognized as having beneficial effects on the quality of traffic flow along a street or arterial. Good signal coordination can also generate measurable safety benefits, primarily in two ways.

Coordinated signals produce platoons of vehicles that can proceed without stopping at multiple signalized intersections. Reducing the number and frequency of required stops and maintaining constant speeds for all vehicles reduce rear-end conflicts. In addition, signal coordination can improve the operation of turning movements. Drivers may have difficulty making permitted turning maneuvers at signalized intersections (e.g., permitted left turns, RTOR after stop) because of lack of gaps in through traffic. Crashes may occur when drivers become impatient and accept a gap that is smaller than needed to complete a safe maneuver. Such crashes could be reduced if longer gaps were made available. Increased platooning can create more gaps of increased length for permitted vehicle movements at intersections and result in improved intersection operation. Also, platooning will contribute to consistent vehicle speeds along a corridor, which will help decrease rearend type crashes.

Corridors with coordinated signals that experience a higher level of rear-end and angle crashes should be reviewed to determine if the timing should be revised or if the signals should be optimized again.

EXHIBIT V-7
Strategy Attributes for Signal Coordination (P)

Attribute	Description
Technical Attributes
Target	The target of this strategy is crashes involving major-street left-turning and minorstreet right-turning vehicles where adequate safe gaps in opposing traffic are not available. These crash types are generally angle and rear-end crashes. Major road rear-end crashes associated with speed changes can also be reduced by retiming signals to promote platooning.
Expected Effectiveness	Studies have proven the effectiveness of signal coordination in improving safety. The ITE Traffic Safety Toolbox (Institute of Transportation Engineers, 1999b) cites two studies of coordinated signals with intersection crash frequencies that dropped by 25 and 38 percent. One of the studies showed an improvement in crash rates for midblock sections as well. Signal coordination can also contribute to a decrease in red-light running. A study on the effectiveness of traffic signal coordination (Rakha et al., 2000) concluded that there is a small but significant improvement in crash rates on intersection approaches after signal coordination. Crashes along the study corridor in Arizona decreased 6.7 percent.
Keys to Success	A key to success is the appropriate spacing of the signals. Signals within a half mile of each other should be coordinated, but signal systems that operate on different cycle lengths do not need to be coordinated. The grouping of the signals to be coordinated is a very important aspect of design of a progressive system. Factors that should be considered include geographic boundaries (see discussion below), volume/capacity ratios, and characteristics of traffic flow (random vs. platoon arrivals).
Potential Difficulties	Spacing of traffic signals is an important factor. As with all signals, coordinated signals too close together can present problems related to drivers focusing on a downstream signal and not noticing the signal they are approaching or proceeding through a green signal and not being able to stop for a queue at an immediate downstream signal. Dispersion of platoons can occur if signals are spaced too far apart, resulting in inefficient use of the signal coordination and loss of any operational benefit. Operations on cross streets may be negatively impacted. Achieving a coordinated system along a corridor may be complicated by signal requirements associated with crossing facilities, any of which may also require signal coordination. Need for long signal cycles associated with multiphase operation and long clearance intervals will dictate the cycle length on which progression will be based. Such a cycle length may produce additional delays on crossing facilities. Furthermore, if there are unsignalized access points that serve substantial entering and exiting traffic volumes along the segments between intersections, this may disrupt the platoon effect of signal coordination. Care should be taken to address effects on gaps produced at unsignalized intersections once coordination is implemented. Site-specific measures may be necessary if adverse operational effects occur or are expected (see the guide on crashes at unsignalized intersections). Coordinating signals for an extended length of highway can involve multiple governmental jurisdictions. There is often disagreement over the benefits or desirability of signal progression, as well as practical issues of developing and maintaining a coordinated signal system. Agreement among the many governmental stakeholders must be achieved in such cases. Along corridors heavily used by fire, ambulance, and other emergency services, implementation of signal preemption for emergency vehicles may be considered (see Strategy 17.2 A5). On some corridors heavily served by bus transit, transit operators are provided in-vehicle traffic signal override capability to enable bus operators to maintain schedule and enhance service. Other corridors may include at-grade rail crossings. In such situations, preemption by emergency vehicles, transit operators, or arrival of trains will break up a platoon and negate the effectiveness of a coordination scheme.
Appropriate Measures and Data	Key process measures include the number of conventional signalized intersections, or length of corridor, for which coordination is implemented. Crash frequency and severity by type of crash are key safety effectiveness measures. It is especially useful to separately analyze crashes by movement or type. Traffic conflicts may be used as a surrogate measure. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to gap acceptance (right-angle crashes involving a vehicle turning left off the main road or permitted RTOR from minor street) and to driver unawareness of signals or signal indications (rear-end crashes) should be analyzed. Traffic volume data are needed to represent exposure. Operational measures are also needed to assess the impact on the subject street and cross streets. Number or percent of vehicles stopped, average speed of progression, and other measures can be derived from use of operational analyses using CORSIM or other traffic operational models. Operational improvement along a corridor may attract vehicles from parallel corridors, potentially offsetting the benefits of coordination in the corridor. However, if the effect on operations in parallel corridors is also evaluated, a more complete understanding of the system benefits would result.
Associated Needs	None identified. However, informing the public that signals are coordinated for safety and operational benefits can be a positive. Some jurisdictions post the speed at which the progression is established, thereby encouraging drivers to maintain that speed.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering policies regarding signal coordination to ensure that appropriate action is being taken on projects. Nearly any highway agency can participate in implementing this strategy, which is applicable mainly to urban and suburban areas where signals are typically more closely spaced. Agreement among jurisdictions as to the need for coordination, appropriate operational parameters, and responsibility for signal system upgrades and maintenance need to be addressed.
Issues Affecting Implementation Time	Implementation time for signal coordination is short to moderate. Installation of signals that may otherwise be unwarranted will increase implementation time, due to additional approvals required. The type of signal system to be installed or upgraded will also affect implementation time.
Costs Involved	Costs involved will be low to medium. If a new system is required to control the coordination, costs will be higher and will include design of the system and purchase and installation of new equipment. If existing signals in a coordinated system are spaced far enough apart that platoons begin to disperse, additional intervening signals may prove beneficial in keeping platoons together (refer to MUTCD Signal Warrant 6). This will also increase costs.
Training and Other Personnel Needs	Traffic signal coordination should be addressed in highway agency training concerning intersection operation.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Traffic signal coordination is compatible with most other strategies to improve signalized intersection safety. Strategy 17.2 A7 discusses removing a signal that is no longer warranted. Consideration may be given to retaining an unwarranted signal to use in a coordinated system.
Other Key Attributes to a Particular Strategy	Traffic signal coordination is generally implemented to improve traffic operations along a major route or in a network and not solely for safety reasons. Other factors to consider include distance between intersections, volume/capacity ratio, and other traffic characteristics.

Strategy 17.2 A5: Employ Emergency Vehicle Preemption (P)

General Description

Signal preemption allows emergency vehicles to disrupt a normal signal cycle in order to proceed through the intersection more quickly and under safer conditions. The preemption systems can extend the green on an emergency vehicle's approach or replace the phases and timing for the whole cycle. The Manual on Uniform Traffic Control Devices [MUTCD (Federal Highway Administration, 2000, 2003)] discusses signal preemption, standards for the phases during preemption, and priorities for different vehicle types that might have preemption capabilities.

Providing for emergency vehicle preemption capability at a signal or along a corridor can be a highly effective strategy in two ways. Any type of crash could occur as emergency vehicles try to navigate through intersections and as other vehicles try to maneuver out of the path of the emergency vehicles. In addition, a signal preemption system can decrease emergency vehicle response times therefore decreasing the wait to receive emergency medical attention. Preemption is especially useful where emergency vehicles are likely to have to travel some distance along a corridor. Also, preemption can provide both a safety and operational benefit at signalized intersections on high-speed roadways where emergency vehicles need to enter the intersection from the minor road.

Technologies for detecting emergency vehicles are described briefly in Appendix 5. Many of these systems have applications in transit-vehicle priority as well as signal preemption for emergency vehicles. Some jurisdictions use confirmation lights to inform drivers that emergency vehicles are preempting the signal or signs that inform drivers that a police pursuit is in progress.

EXHIBIT V-8
Strategy Attributes for Employing Emergency Vehicle Signal Preemption (P)

Attribute	Description
Technical Attributes
Target	The target of this strategy is signalized intersections where normal traffic operations impede emergency vehicles and where traffic conditions create a potential for conflicts between emergency and nonemergency vehicles. These conflicts could lead to almost any type of crash, due to the potential for erratic maneuvers of vehicles moving out of the paths of emergency vehicles.
Expected Effectiveness	Installation of signal preemption systems for emergency vehicles has been shown to decrease response times. A review of signal preemption system deployments in the United States shows decreases in response times between 14 and 50 percent for systems in several cities (Collura et al, 2001). In addition, the study reports a 70-percent decrease in crashes with emergency vehicles in St. Paul, Minnesota, after the system was deployed (though the extent to which emergency vehicle priority was implemented in the city is unclear).
Keys to Success	A key to success is ensuring that the preemption system works when needed by providing clear sight lines between emergency vehicles and detectors. Also, it is important to ensure that vehicles from a variety of jurisdictions will be able to participate in the signal preemption program. The focus of the treatment should be on fire and EMS. Some police agencies have found that since officers respond to incidents from many directions, the preemption system is not as effective for their needs. Another key to success is the coordination of implementation across jurisdictions, including compatibility of equipment and technology, as well as operational policies.
Potential Difficulties	Preempted signals that stop vehicles for too long may encourage disrespect in drivers for the red signal, and they may decide to proceed even though the signal is red. Preemption of signals by emergency vehicles will temporarily disrupt traffic flow. Congestion may occur, or worsen, before traffic returns to normal operation. One study of signal preemption systems in the Washington, DC, metropolitan area demonstrated that once a signal was preempted, coordinated systems took anywhere from half a minute to 7 minutes to recover to base-time coordination. During these peak periods in more congested areas, vehicles experienced significant delays. Agency traffic personnel indicated that signal preemption impacts increase as the length of the peak period increases (Collura et al., 2001). Light-based detectors need a clear line of sight to the emitter on the vehicles; this line could become blocked by roadway geometry, vehicles, foliage, or precipitation. Systems from different vendors may not interact well together. Also, other alarms, such as from nearby buildings, may be erroneously activated by a sound-based system.
Appropriate Measures and Data	Key process measures include the number of intersection approaches for which signal preemption systems are implemented and the number and percent of emergency response vehicles that are equipped. A key operational measure of effectiveness is response time of emergency vehicles proceeding through the intersections where signal preemption is implemented. Other operational measures include delay, conflicts en route, and time to return to normal operation along affected streets. Frequency and severity of crashes involving emergency vehicles by type of crash are also key safety effectiveness measures. Traffic volume data are needed to represent exposure. These data should be collected before and after installation of the system.
Associated Needs	It is extremely important to coordinate with all surrounding jurisdictions to maximize use of the preemption system chosen.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway and other agencies should ensure that their policies for traffic signals include use of signal preemption systems. A successful program requires the coordinated and cooperative involvement of agencies from engineering, enforcement, emergency medical services, etc. throughout the area. Implementation of a preemption system should be considered as part of programs to upgrade corridor or jurisdictional traffic signal and control systems. Highway agencies should review their traffic engineering policies regarding use of emergency vehicle signal preemption to ensure that appropriate action is being taken on individual projects. Nearly any agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas. In some cases, multijurisdictional programs will be desirable to create an effective system.
Issues Affecting Implementation Time	Implementation time will vary from short to medium based upon the number of intersections and number of agencies involved in the preemption system.
Costs Involved	Costs for installation of a signal preemption system will vary from medium to high, based upon the number of signalized intersections at which preemption will be installed and the number of emergency vehicles to be outfitted with the technology. The number of detectors and the intricacy of the preemption system could increase costs.
Training and Other Personnel Needs	Appropriate signal phasing and timing for periods of preemption control should be addressed in highway agency training concerning traffic signal operations and signal preemption.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Signal preemption is compatible with most other strategies to improve signalized intersection safety.
Other Key Attributes to a Particular Strategy	None identified.

Information on Agencies or Organizations Currently Implementing This Strategy

The Oregon DOT uses an on-line explanation and form for localities desiring to install signal preemption devices along state highways; see http://www.odot.state.or.us/traffic/signalpre.htm.

Strategy 17.2 A6: Improve Operation of Pedestrian and Bicycle Facilities at Signalized Intersections (Combination of Tried and Proven Strategies)

Nearly one-third of all pedestrian-related crashes occurs at or within 50 feet of an intersection. Of these, 30 percent involve a turning vehicle, whereas another 22 percent involve a pedestrian either running across the intersection or darting in front of a vehicle whose view was blocked just prior to the impact. Another 16 percent of these intersectionrelated crashes occur because of driver violation (e.g., failure to yield the ROW).

The companion guide for crashes involving pedestrians comprehensively addresses pedestrian safety. The following discussion summarizes key issues relative to pedestrian safety at signalized intersections.

Traffic control improvements that can be made to an intersection to increase pedestrian safety include the following:

• Pedestrian signs, signals, and markings,

• Crossing guards for school children,

• Lights in crosswalks in school zones,

• Pedestrian-only phase or pedestrian-lead phase during signal operation,

• Prohibition of RTOR,

• Public information or signs that educate pedestrians regarding use of push buttons (specifically, that they will not receive the walk signal immediately), and

• Technology to show a push button is working (such as a button that lights up, similar to an elevator).

Providing pedestrian push buttons may facilitate safe pedestrian roadway crossings at signalized intersections (vs. midblock crossings), where pedestrian conflicts with motor vehicles can be managed through use of pedestrian crossing signals and/or exclusive pedestrian-only phases during the signal operation. However, pedestrian push buttons at an intersection are often obscured by roadside furniture or other items. Providing visible signs alerting pedestrians to the presence of push buttons and anticipated wait time for the crossing signal may increase the use of existing pedestrian push buttons.

Several strategies employed in Europe to improve pedestrian safety at signalized intersections are described in Appendix 4.

The AASHTO Guide for the Development of Bicycle Facilities (American Association of State Highway and Transportation Officials, 1999) should be consulted for information on bicycle safety. Traffic control improvements that can be made to an intersection to increase safety for bicyclists include the following:

• “Bicyclist Dismount” signs at intersections, and

• Stop and “Bicyclist Dismount“ signs at intersections with bike trails.

Additional details are provided in the guides for crashes at unsignalized intersections and for pedestrian crashes.

Strategy 17.2 A7: Remove Unwarranted Signal (P)

General Description

Traffic signals can remedy many safety and operational problems at intersections. However, signals often can adversely affect intersections. It is possible that a signal may no longer be warranted due to changes in traffic conditions. Problems created by an unwarranted signal, such as excessive delay, increased rerouting of traffic to less-appropriate roads and intersections, higher crash rates, and disobedience of the traffic signal can be addressed by removing the signal if doing so would not create worse problems.

Signalized intersections generally experience crashes of different types than unsignalized intersections but not necessarily a lower total crash rate. Converting the intersection to unsignalized may not improve the total crash rate, but it may improve crash severity for some crash types.

Studies should be performed when considering removing a signal, just as installation of a signal is studied. This study should identify the appropriate replacement traffic control devices and any sight distance restrictions that may not have been an issue while under signalized control.

Once the new traffic control has been installed, the signal heads should be set to flash or should be covered for a minimum of 90 days to draw driver, pedestrian, and bicyclist attention to the change in control. After this period, the signal can be removed if the data collected during the study period support removal of the signal. The poles and cables may remain in place, however, for up to a year while additional analysis continues.

EXHIBIT V-9
Strategy Attributes for Removing Unwarranted Traffic Signals (P)

Attribute	Description
Technical Attributes
Target	This strategy is targeted at signalized intersections where traffic volumes and safety record do not warrant a traffic signal. Signalized intersections tend to have higher rear-end crash rates than unsignalized intersections, and conversion to two-way stopcontrol or all-way stop-control may reduce the rear-end crash rate.
Expected Effectiveness	Removal of an unwarranted signal will eliminate excessive delay and disobedience of the signal indicators at the targeted intersections if these conditions exist because the signal is no longer needed. Signal removal should also decrease the use of inappropriate routes (e.g., residential streets) used by drivers in an attempt to avoid the traffic control signals and decrease the frequency of collisions (especially rear-end collisions). Two studies have examined the effectiveness of removing traffic signals. Kay et al. (1975) found a decrease in annual average crash frequency of greater than one crash per year when intersections are converted to all-way stop control. Where signals were replaced by two-way stop control intersections, right-angle crashes increased, but rear-end crashes decreased by approximately the same amount. The ITE Traffic Engineering Handbook (Pline, 1999) cites two studies that present conflicting results of safety analyses of removed signals: Kay et al. (1975) found that frequency for all crash types did not change after signals were removed and that rates for right-angle crashes increased and rates for rear-end crashes decreased. Another study showed that rates of right-angle and rear-end crashes both decreased. Since there is conflicting evidence on the safety benefits of signal removal, each intersection for which signal removal is considered should be analyzed separately, and other conditions at the intersections (specifically, geometry, sight distance and traffic conditions) should be carefully considered as well.
Keys to Success	Keys to success include determining the appropriate traffic control to be used after the removal of the signal and removing any sight-distance restrictions through the intersection. Pedestrian and bicycle movements through the intersections should be considered when determining traffic control, geometric changes, and signing improvements that will be made when the signal is removed. Intersection sight distance may not be required where signals are present. If a signal is to be removed, care should be taken to ensure that adequate intersection sight distance is provided and that necessary improvements to sight distance, such as clearing sight triangles, should accompany the signal removal. Keeping the public informed about the traffic control removal study will also lead to the success of this strategy.
Potential Difficulties	Right-angle crashes may increase after the signal is removed. Removal of the traffic signals could delay the flow of pedestrians and bicyclists through the intersection. Confusion to drivers regarding ROW and disorderly movement through the intersection may also result if sufficient PI&E is not provided regarding the change in traffic control.
Appropriate Measures and Data	A key measure of the implementation process is the number of intersections where signals are removed. Another measure is the volume of conflicting flows that are affected by removal. Crash frequency and severity by type of crash are also key safety effectiveness measures. It is especially useful to separately analyze crashes by movement or type. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes that may be related to the traffic control (right angle, rear end) should be analyzed. Traffic volume data are needed to represent exposure.
Associated Needs	The public should be informed when a study is underway for the removal of a traffic signal at each intersection. PI&E is a key element of a project to remove a signal, to help ensure that driver expectancy is not violated.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding the removal of traffic signals to ensure that appropriate action is being taken. Policy guidance regarding the removal of traffic signals is discussed in the MUTCD. The MUTCD should be consulted if agency policy has not incorporated the information from the MUTCD. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas.
Issues Affecting Implementation Time	Implementation time can vary, depending upon the extent and nature of the public involvement.
Costs Involved	Since implementation of this strategy requires the removal of traffic signals and replacing them with signs, its cost would be low. Costs would be attributed to the equipment needed for signal removal and temporary traffic control while implementing the new traffic control method (usually signs).
Training and Other Personnel Needs	Traffic signal warrants should be addressed in highway agency training regarding traffic control devices, along with guidelines for recognizing the appropriateness of removing a signal.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Removal of traffic signals is typically done when studies show that traffic patterns have changed significantly. This strategy is not usually associated with any other strategies.
Other Key Attributes to a Particular Strategy	None identified.

Objective 17.2 B—Reduce Frequency and Severity of Intersection Conflicts through Geometric Improvements

Geometric improvements can provide both operational and safety benefits at signalized intersections. Improvements to turning movements, through channelization or even physically preventing turns, can result in reductions in certain types of crashes. Geometric changes can also improve safety for pedestrians and bicyclists. Higher-cost, longer-term improvements, such as redesign of the intersection, can also improve safety and are briefly discussed in this section.

Strategy 17.2 B1: Provide or Improve Left-Turn Channelization (Combination of Tried and Proven Strategies)

General Description

This strategy includes the following:

• Providing left-turn lanes,

• Lengthening left-turn lanes,

• Providing positive offset for left-turn lanes,

• Providing positive guidance with channelization, and

• Delineating turn path.

Many intersection safety problems can be traced to difficulties in accommodating left-turning vehicles. A key strategy for minimizing collisions related to left-turning vehicles (angle, rearend, sideswipe) is to provide exclusive left-turn lanes, particularly on high-volume and highspeed major-road approaches. Left-turn lanes allow separation of left-turn and through-traffic streams, thus reducing the potential for rear-end collisions. Because they provide a sheltered location for drivers to wait for a gap in opposing traffic, left-turn lanes may encourage drivers to be more selective in choosing a gap to complete the left-turn maneuver. This may reduce the potential for collisions between left-turn and opposing through vehicles. Provision of a leftturn lane also provides additional flexibility in designing a phasing plan.

Installation, lengthening, and offsetting of left-turn lanes are discussed in further detail in the guide for crashes at unsignalized intersections.

Install Left-Turn Lane. Left-turn lanes are a proven treatment for addressing safety problems associated with left-turning vehicles. By removing left-turning vehicles from the throughtraffic stream, conflicts with through vehicles traveling in the same direction can be reduced (and even eliminated, depending on the signal timing and phasing scheme [see Strategy 17.2 A1]). Drivers wait in the turn lane until there is a gap in opposing traffic through which they can turn, which helps reduce the conflicts with the opposing through traffic.

The design of the left-turn lane is crucial to its effectiveness as either a safety or operational improvement strategy. In providing left-turn lanes, vehicles in opposing left-turn lanes may block the respective driver's view of approaching vehicles in the through lanes. This potential problem can be resolved by offsetting the left-turn lanes (see below).

See Appendix 6 for further considerations for installing left-turn lanes.

Improve Left-Turn Lane Geometry. Safety improvements can also be made to approaches that already incorporate separate left-turn lanes. Three treatments are discussed below: lengthening of the left-turn lane, redesigning to provide positive visual offset, and delineating the turning path.

Lengthen Left-Turn Lane. The length of a left-turn lane consists of three components: entering taper, deceleration length, and storage length. The left-turn lane length should allow for the removal of slow or decelerating vehicles from through traffic, thus reducing the potential for rear-end collisions. A turn lane long enough to accommodate deceleration can have safety benefits for higher-speed intersections such as are typically found in rural highways. The turn lane should be of adequate length to store vehicles waiting to turn left without the queue overflowing into the adjacent through lane. If a left-turn queue extends into the adjacent through lane, through vehicles will be forced to stop or, if there are multiple through lanes, change lanes. These maneuvers can lead to rear-end and sideswipe crashes. Also, if access to a left-turn lane is blocked by a queue of through vehicles at a signal, the left-turners may drive into the opposing lane to reach the left-turn lane. This could lead to head-on crashes.

Design criteria for selecting an appropriate left-turn lane length are presented in the AASHTO Policy on Geometric Design for Highways and Streets (American Association of State Highway and Transportation Officials, 2001), the TRB Highway Capacity Manual (2000), NCHRP Report 279 (Neuman, 1985), and the policies of individual highway agencies. NCHRP Report 457 (Bonneson and Fontaine, 2001) also includes guidance on determining left-turn length depending on volume and traffic control. A detailed analysis of traffic conditions should be performed to ensure that a left-turn lane is of a length appropriate for the given traffic.

Provide Positive Offset for Left-Turn Lanes. A potential for conflict exists when vehicles in opposing turn lanes on the major road block the drivers' views of approaching traffic. A leftturning driver's view of opposing through traffic may be blocked by left-turning vehicles on the opposite approach. When left-turning traffic has a permissive green signal phase, this can lead to collisions between vehicles turning left from the major road and through vehicles on the opposing major-road approach. To reduce the potential for crashes of this type, the left-turn lanes can be offset by moving them laterally, so that vehicles in opposing lanes no longer obstruct the opposing driver (See Exhibit V-10). This helps improve safety and operations of the left-turn movement by improving driver acceptance of gaps in opposing through traffic. This is especially true for older drivers who have difficulty judging gaps in front of oncoming vehicles. Note that the effectiveness of this strategy is greatest where signal operations include permissive signal phasing or permissive/protected phasing for left-turning movements. (See Strategy 17.2 A1).

AASHTO's Policy recommends that medians wider than 18 feet should have offset left-turn lanes. One method for laterally shifting left-turning vehicles is to narrow the turn lane width using pavement markings. This is accomplished by painting a wider stripe at the right side of the left-turn lane, which causes left-turning vehicles to position themselves closer to the median. Wider lane lines were implemented at six intersections in Nebraska with good results. The width of these lines ranged from 0.5 feet to 3 feet. The wider the leftturn lane line used to offset vehicles, the greater the effect on improving sight distance. (McCoy et al., 1999).

EXHIBIT V-10
Positive Offset of Left-Turn Lanes
Source: Federal Highway Administration, in press.

Delineate Turn Path. Even at signalized intersections, where the traffic signals help to eliminate confusion about ROW, driver confusion can exist in regard to choosing the proper turn path. This is especially relevant at intersections where multiple left-turn lanes are provided, the overall pavement area of the intersection is large, or other unfamiliar elements are presented to the driver. Delineation of turn paths (see Exhibit V-11) is especially useful to drivers making simultaneous opposing left turns, as well as some cases involving drivers turning right for which a clear path is not readily apparent. This strategy is also appropriate for application where the roadway alignment may be confusing or unusual, such as a deviation in the path for through vehicles. Providing positive guidance to the driver in the form of pavement markings can help eliminate driver confusion and eliminate vehicle conflict by “channeling“ vehicles in their proper path.

EXHIBIT V-11
Delineation of Turn Paths for Double Left-Turn Lanes
Source: Federal Highway Administration, in press.

EXHIBIT V-12
Strategy Attributes for Providing or Improving Left-Turn Channelization (P, T)

Attribute	Description
Technical Attributes
Target	This strategy targets intersections where crashes related to left-turn movements are an issue. Crash types that could be reduced by providing or improving left-turn channelization include angle, sideswipe (both same and opposite direction), rear-end, and head-on crashes.
Expected Effectiveness	Recent research has demonstrated the substantive safety effect of providing left-turn lanes. The safety effectiveness varies with the location (rural vs. urban), number of legs, type of traffic control, and number of approaches for which the lane is installed. Exhibit V-12A (rural) and Exhibit V-12B (urban) below provide the best estimates of the relative effectiveness of left-turn lanes (Harwood et al., 2002). The full report should be consulted, since the accident modification factors (AMFs) are to be applied to a base model that is provided therein. Also, there are a variety of effectiveness estimates made for varying types of crashes and left-turn lane treatments. EXHIBIT V-12A Recommended Accident Modification Factors for Installation of Left-Turn Lanes on the Major-Road Approaches to Rural Intersections Intersection Type Intersection Traffic Control Number of Major-Road Approaches on Which Left-Turn Lanes Are Installed One Approach Both Approaches Three-leg Stop signa Traffic signal 0.56 0.85 —— —— Four-leg Stop signa Traffic signal 0.72 0.82 0.52 0.67 a On minor-road approach(es). EXHIBIT V-12B Recommended Accident Modification Factors for Installation of Left-Turn Lanes on the Major-Road Approaches to Urban Intersections Intersection Type Intersection Traffic Control Number of Major-Road Approaches on Which Left-Turn Lanes Are Installed One Approach Both Approaches Three-leg Stop signa Traffic signal 0.67 0.93 —— —— Four-leg Stop signa Traffic signal 0.73 0.90 0.53 0.81 a On minor-road approach(es). The safety effectiveness of providing a positive offset between opposing left-turn lanes has not been quantitatively demonstrated. The positive offset increases visibility of the sight lines, enabling drivers to perceive safe gaps when operating as permissive movements at signalized intersections. Harwood et al. (2002) report some findings related to the use of painted vs. curbed channelization of left-turn lanes. Delineation of turn paths is expected to improve intersection safety, though the quantitative effectiveness of this treatment has not been proven. The additional guidance in the intersection will help separate vehicles making opposing left turns, as well as vehicles turning in adjacent turn lanes.
Keys to Success	Keys to success in implementing left-turn lanes include the appropriate design of all elements (length, width of lane, tapers). Re-striping of available width, including use of all or part of a shoulder or parking lane, can be an effective low-cost way to implement this strategy. The operational effects, on adjacent approaches, of re-striping or redesign should be addressed. Another key to success with left-turn lanes is to incorporate other safety-effective strategies, such as protected-only signal phasing. The key to success in using pavement delineators is to make sure that the pavement markings are visible to, and understandable by, the driver. Delineators should be used as guidance devices, rather than warning devices. It is important to maintain the markings so that they retain their value. Placement of pavement delineators should be based on prevailing speeds, block lengths, distance from intersections, and other factors influencing communication with the driver.
Potential Difficulties	Potential difficulties in providing left-turn lanes where they currently do not exist are the cost and acquisition of space required for the additional lane and the need to relocate the signal heads and hardware. As noted above, use of shoulders and/or parking lanes may be considered, but potential adverse safety concerns, such as lack of a shoulder for emergency stops, should be addressed. In addition, it will be important to address concerns from business owners or other stakeholders concerned about loss of parking. Difficulties may arise due to the need to coordinate design changes with adjacent intersections, as well as accounting for possible operational effects on adjacent intersections. The addition of left-turn lanes may have other adverse effects that should be addressed. Most notably, the greater pavement width and exposure to traffic for pedestrians (including potentially the loss of median protection for pedestrians) represents a risk. Decreased visibility of required pavement delineators can be caused by degraded retroreflectivity during night and/or adverse weather conditions. Rain during the daytime reduces the driver's ability to see the surrounding area. At night, headlight glare from oncoming vehicles, movement of the windshield wiper, slippery pavement surface, and degraded retroreflectivity all add to the decreased visibility of pavement markings. Snow covers pavement markings, and the use of snowplows and chemical and deicing agents affect the durability of the pavement markings.
Appropriate Measures and Data	Key process measures include the number of intersections at which left-turn lanes are implemented and/or improved, the total number and type of left-turn lanes installed, and the number of potential conflicts affected by the improvement. Crash frequency and severity by type of crash are key safety-effectiveness measures. It is important to identify crashes related to the targeted movement and to analyze them separately. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to the targeted maneuvers at the intersection should be analyzed separately. Traffic volume data are needed to represent exposure. It is especially desirable to obtain data on the volume of vehicles using the intersection, turning volumes, and the conflicting volumes. Driver behavior measures may be used as surrogates (e.g., vehicle paths, actual vehicle conflicts, erratic maneuvers).
Associated Needs	Many highway agencies have shifted their focus from constructing new roadways to improvement of existing facilities. The challenge for highway, traffic, and safety engineers is to develop techniques that will result in the decrease of accidents, delays, and other inconveniences on existing facilities. Maintenance of pavement markings is one aspect of this. Development of a roadway delineation management system is a way to track conditions of pavement markings.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway and other agencies should ensure that their design polices for new or reconstructed intersections incorporate consideration of all aspects of left-turn lane design and operation delineation. Highway agencies should review their policies for left-turn warrants and design to consider explicit safety considerations as well as capacity and traffic operational considerations. Highway agencies may also wish to revise their standard intersection design details to accommodate offset left-turn lane treatments as their standard approach. Highway agencies should also review the MUTCD, in addition to their traffic engineering and design policies, regarding use of delineators on routine projects.
Issues Affecting Implementation Time	Improving or implementing left-turn lane treatments can range widely in time. Where no changes to existing pavement or no new construction is needed, implementation can take only weeks or months (including engineering studies). Where redesign or restriping of approaches is performed, time may be longer (6 months or more) depending on the need to reposition or change the location of traffic signal heads or other hardware. Also, gaining acceptance from stakeholders for removal of parking or other actions may require time. Implementation time of applying pavement delineators is relatively quick (1 to 2 days).
Costs Involved	Costs of implementing or improving the design of left-turn lanes can vary. Where reallocation of available width by re-striping is all that is needed, the cost can be relatively low. Where redesign and widening or other construction is necessary, costs will be moderate. Costs may include upgrading and/or relocation of traffic signals and other hardware. Left-turn lane improvements that require ROW acquisition or major reconstruction can be high-cost projects. The cost of delineators is variable and determined largely by the material used for pavement markings (paint, thermoplastic, epoxy, etc.). When using delineators, an issue of concern is the cost-to-service-life of the material. Predicting how long pavement delineators will last is difficult due to the variable factors influencing service life (weather and amount and type of traffic). Materials with a short service life are not desirable, due to the frequent disruption of traffic and threat to worker safety needed to re-mark the pavement.
Training and Other Personnel Needs	Revisions to standard intersection designs such as offset left-turn lanes may require some training of agency engineering personnel. Instituting and following a roadway delineation management system will require training. Personnel will require training to apply, inspect, and maintain pavement delineators.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at signalized and unsignalized intersections. In particular, optimum results from improving or adding left-turn lanes may result by combining it with signalization strategies such as protected phasing and improved clearance intervals.
Other Key Attributes to a Particular Strategy	None identified.

Strategy 17.2 B2: Provide or Improve Right-Turn Channelization (P)

This strategy includes providing right-turn lanes and lengthening right-turn lanes.

General Description

Many collisions at signalized intersections are related to right-turn maneuvers. A key strategy for minimizing such collisions is to provide exclusive right-turn lanes. It is also important to ensure that the right-turn lanes are of sufficient length to allow vehicles to decelerate before turning, ideally without affecting the flow of through traffic. Right-turn lanes remove slow vehicles that are decelerating to turn right from the through-traffic stream, thus reducing the potential for rear-end collisions.

Provide Right-Turn Lanes. The provision of right-turn lanes can minimize collisions between vehicles turning right and following vehicles, particularly on high-volume and high-speed major roads. A right-turn lane may be appropriate in situations where there are an unusually high number of rear-end collisions on a particular approach. Installation of a right-turn lane on one major road approach at a signalized intersection is expected to reduce total crashes according to the AMFs in Exhibit V-13 (Harwood et al., 2002).

The benefits of a right-turn lane are not provided just by the presence of the lane but also by the specific design. Key design issues addressed in design guides include entering taper, deceleration length, and storage length. Design criteria for selecting an appropriate rightturn lane length are presented in both the AASHTO Policy on Geometric Design for Highways and Streets and the FHWA'S 2003 MUTCD, as well as in the policies of individual highway agencies. Through drivers may enter an excessively long right-turn lane by mistake, without realizing it is a turn lane. Upstream signing and marking of the turn lane may address this.

EXHIBIT V-13
Recommended Accident Modification Factors for Installation of Right-Turn Lanes on the Major-Road Approaches to Rural and Urban Intersections

It is possible that installation of a right-turn lane could create other safety or operational problems at the intersection. For example, vehicles in the right-turn lane may block the cross street right-turning drivers' view of through traffic; this would be a significant issue where RTOR are permitted on the cross street. If a right shoulder is re-striped to provide a turn lane, there may be an adverse effect on safety due to the decrease in distance to roadside objects. Delineation of the turn lane also should be carefully considered, so that adequate guidance is provided through the intersection. A channelized right-turn roadway may be desirable in some locations. Channelization of the right turn with a raised or painted island can provide larger turning radii and also an area for pedestrian refuge. Details on the design of channelizing islands for turning roadways can be found in the AASHTO Policy on Geometric Design for Highways and Streets.

Channelizing islands can be raised or flush with the pavement. A Georgia study evaluated the effects of right-turn channelization in the form of painted islands, small raised islands, and large raised islands. Results from the study show that traffic islands appear to reduce the number of right-turn angle crashes and that the addition of an exclusive turn lane appears to correspond to an increased number of sideswipe crashes given the introduction of a lane change (Dixon et al., 1999). Harwood et al. (2002) report some potentially significant differences in the safety performance of painted versus curbed channelization.

Visibility of channelizing islands is very important. Islands can be difficult for drivers to see, especially at night and in inclement weather. This is particularly true for older drivers. Raised islands have been found to be more effective than flush painted islands at reducing nighttime collisions, since they are more easily seen.

Older drivers in particular benefit from channelization as it provides a better indication of the proper use of travel lanes at intersections. However, older drivers find that making a right turn without the benefit of an acceleration lane on the crossing street is particularly difficult (Staplin et al., 1998).

Right-turn roadways can reduce the safety of pedestrian crossings. Crossing distances are increased, as is pedestrian exposure to traffic. Elderly and mobility-impaired pedestrians may have difficulty crossing intersections with large corner radii. Right-turn channelization also makes it more difficult for pedestrians to cross the intersection safely, adequately see oncoming traffic that is turning right, and know where to cross. Proper delineation of the turning roadway may help, particularly at night.

Where curbed islands are provided, they offer a refuge for pedestrians. Where it is known that channelization islands are being used by pedestrians, crossing paths should be clearly delineated, and the island itself should be made as visible as possible to passing motorists.

Removal of small right-turn triangular channelizing islands may be an appropriate method for improving right-turn channelization. Often, these islands were installed in urban areas as a location to place a signal pole. Right-turning drivers may not see this island when approaching the intersection and may stop suddenly, increasing the potential for rear-end collisions. Removal of this island may be appropriate, especially if the road becomes less urban in nature. The city of Winston-Salem, North Carolina, has removed such triangular islands in an attempt to reduce rear-end collisions related to right turns. To provide positive guidance through the turn, the city installed flexible delineators along the gore stripe. The “after” study has not been completed, but it does appear right-turn rear-end crashes are becoming less common.

Other issues to consider when designing a right-turn lane include provision of clear sight triangles, increased crossing distance for pedestrians, potential conflicts between turning vehicles and cyclists proceeding through the intersection, and the potential need to move the stop bar on the cross street. Transit stops may also need to be moved from the near side to the far side of an intersection due to possible conflicts between through buses and rightturning vehicles.

Lengthen Right-Turn Lanes. Lengthening a right-turn lane can help improve operations and safety by providing additional sheltered space for vehicles to decelerate or wait to turn. If the length of a right-turn lane is inadequate, vehicles waiting to turn may be doing so from the through-traffic stream, thus increasing the potential for rear-end collisions. Providing longer entering tapers and deceleration lengths can reduce the potential for rear-enders. Also, if access to a right-turn lane is blocked by a queue of through vehicles at a signal, the right-turners may block the movement of through traffic, if the two movements operate on separate or split phases. This could lead to unsafe lane changes and added delay.

The length of a right-turn lane consists of three components: entering taper, deceleration length, and storage length. Design criteria for selecting an appropriate right-turn lane length are presented in both the AASHTO Policy on Geometric Design for Highways and Streets and the TRB Highway Capacity Manual, as well as in the policies of individual highway agencies. A detailed analysis of traffic conditions should be performed to ensure that a right-turn lane is of proper length.

Improvements to right-turn lanes are discussed in greater detail in the Unsignalized Intersection guide.

Strategy 17.2 B3: Improve Geometry of Pedestrian and Bicycle Facilities (Combination of Tried and Proven Strategies)

The mix of travel modes at intersections, along with the vehicle-vehicle conflicts possible, can create safety and operational concerns for nonmotorists. A variety of relatively low-cost treatments can be implemented to help pedestrians and bicyclists proceed through the intersection more safely and more efficiently. Multivehicle crashes (specifically rear-ends) can be reduced if pedestrians are more visible and more drivers expect to encounter them.

Geometric or physical improvements that can be made to an intersection to increase pedestrian safety include the provision of the following:

• Continuous sidewalks,

• Signed and marked crosswalks,

• Sidewalk set-backs,

• Median refuge areas,

• Pedestrian overpasses,

• Intersection lighting,

• Physical barriers to restrict pedestrian crossing maneuvers at higher-risk locations,

• Relocation of transit stops from the near side to the far side of the intersection, and

• Other traffic calming applications to reduce vehicle speeds or traffic volumes on intersection approaches.

Improvements to pedestrian facilities are discussed in greater detail in the guide for crashes involving pedestrians. Several strategies used by European countries to improve pedestrian safety at signalized intersections are described in Appendix 4.

Some of the problems facing bicyclists at intersections include high traffic volumes and speeds as well as the lack of space for bikes. Possible improvement projects include the following:

• Widening outside through lanes (or adding bike lanes),

• Providing median refuge areas,

• Providing independent crossing structures,

• Upgrading storm drain grates with bicycle-safe designs, and

• Implementing lighting.

Additional improvements for bicyclists are listed in the guide for crashes occurring at unsignalized intersections.

Strategy 17.2 B4: Revise Geometry of Complex Intersections (Combination of Tried and Proven Strategies)

This strategy includes a series of mostly higher-cost solutions:

• Converting a four-leg intersection to two T intersections,

• Converting two T intersections to one four-leg intersection,

• Improving intersection skew angle, and

• Improving deflection in the through-vehicle travel path.

A fifth solution, closing an intersection leg, is one commonly tried when addressing the problem of complex intersections. This can be a low-cost solution because it does not typically require major reconstruction. A detailed description of this strategy follows.

General Description

Some geometric problems with signalized intersections will not be remedied using signing, channelization, or signal phasing. Physical modifications to all or part of an intersection may be needed to reduce severe crash rates. There may be multiple problems associated with one or more movements at the intersection that can be best addressed with significant improvements to intersection design. Because of the extensive reconstruction required to implement these strategies, they will not be appropriate for agency programs designed for quick low-cost action.

Convert a Four-Leg Intersection to Two T Intersections. (T) For some signalized four-leg intersections with very low through volumes on the cross street, the best method of improving safety may be to convert the intersection to two T intersections. This strategy should help reduce crashes related to the intersection layout, such as angle crashes involving left-turning vehicles in which drivers are not expecting to encounter one of the infrequent through-vehicles. This conversion to two T intersections can be accomplished by realigning the two cross-street approaches an appreciable distance along the major road, thus creating separate intersections that operate relatively independently of one another. The intersections should be separated enough to ensure the provision of adequate turn-lane channelization on the major road.

If through volumes are high, the intersection may be safer if left as a conventional four-leg intersection. Converting it to two T intersections would only create excessive turning movements at each of the T intersections.

In a study conducted by Hanna et al., (1976) offset intersections had accident rates that were approximately 43 percent of the accident rate at comparable four-leg intersections. Thus, it is expected that this strategy would reduce the accident experience of targeted four-leg intersections.

Convert Two T Intersections to One Four-Leg Intersection. (T) For some signalized offset T intersections with very high through volumes on the cross street, the best method for improving safety may be to convert the intersection to a single four-leg intersection. This can be accomplished by realigning the two cross-street approaches to meet at a single point along the major road. It is expected that this strategy would reduce accidents involving left-turning traffic from the major road onto the cross street at each of the two T intersections.

Improve Intersection Skew Angle. (P) Roads that intersect with each other at angles less than 90 degrees can present sight distance and operational problems for drivers. A high incidence of right-angle accidents, particularly involving vehicles approaching from the acute angle, may be the result of a problem associated with skew. Vehicles have a longer distance to travel through the intersection (increasing their exposure to conflicts), and drivers may find it difficult to turn their head and neck to view an approach on an acute angle. Furthermore, vehicles turning right at an acute angle may encroach on the lane for vehicles approaching from the opposite direction. When RTOR are permitted, drivers may have more difficulty judging gaps when turning. Also, crossing distances for pedestrians are increased.

Skewed intersections (with the angle of intersection less than 75 degrees) pose particular problems for older drivers, as many older drivers experience a decline in head and neck mobility. A restricted range of motion reduces the older driver's ability to effectively scan to the rear and sides of their vehicle to observe blind spots. They may also have trouble identifying gaps in traffic when making a left turn or safely merging with traffic when making a right turn. More information on converting a four-leg intersection to two T intersections, or vice versa, and eliminating intersection skew is contained in the guide for crashes occurring at unsignalized intersections.

Remove Deflection in Through-Vehicle Travel Path. (T) Intersections with substantial deflections between approach alignments can produce operational and safety problems for through-vehicles as they navigate through an intersection (see Exhibit V-14). Forced path changes for through-vehicles violate driver expectations and may be difficult for unfamiliar drivers to navigate. Violation of driver expectancy can result in reduced speed of the vehicle through the intersection. Crashes influenced by a deflection in travel path are likely to include rear-end, sideswipe, head-on, and angle. Acceptable deflection angles through intersections vary by individual agency, but are typically related to the design and/or posted speed on an intersection approach. Typical maximum deflection angles are 3 to 5 degrees.

Pavement markings can be a low-cost solution to guide through vehicles through the intersection. Dashed lines similar to those used to delineate left-turn paths are appropriate for delineation of the through path.

Redesign of an intersection approach is a relatively high-cost solution. Proper design of an intersection involves providing traffic lanes that are clearly visible to drivers at all times, clearly understandable for any desired direction of travel, free from the potential for conflicts to appear suddenly, and consistent in design with the portions of the highway approaching the intersection. The sight distance should be equal to, or greater than, minimum values for interchange conditions.

EXHIBIT V-14
Deflection in through-Vehicle Travel Path through Intersection
Source: Federal Highway Administration, in press.

Close Intersection Leg. (T) For some signalized intersections with crash histories, the best method for improving safety may be to close access to a leg of the intersection. This is a radical approach to safety improvement that should generally be considered only when less restrictive measures have been tried and have failed. Closure of access to an intersection leg can be accomplished by closing and abandoning a minor approach using channelizing devices or by reconstructing the minor approach so that it dead-ends before reaching the intersection with the major street. An alternative to closing the entire intersection leg is to convert the leg to a one-way street that departs the intersection. Though it is a significant modification to an intersection, it can be a low-cost treatment. A major consideration in deciding to implement this strategy is the impact closure will have on traffic patterns and volumes at other locations. This treatment may be most applicable to those intersections with more than four legs.

EXHIBIT V-15
Strategy Attributes for Revising Geometry of Complex Intersections—Closing Access to an Intersection Leg (T)

Attribute	Description
Technical Attributes
Target	The target of this strategy should be signalized intersections with high levels of crashes on a leg where other strategies have not been successful or are not considered appropriate. Any crash type could be targeted by this strategy, since reasons for closing an intersection leg can vary.
Expected Effectiveness	Closure of an intersection leg would be expected to eliminate crashes related to that leg. Determination of the effectiveness is site specific, due to the varying conditions at intersections where leg closure might be considered. In addition, consideration must be given to the adjacent intersections, to alternative routes onto which traffic would be diverted, and to the potential impact to safety on those routes. Where properly applied, a net safety benefit can be expected.
Keys to Success	The key to success for a project of this type is conducting an adequate system traffic study to ensure that the safety and other operational problems are not merely transferred from the intersection being treated to other locations. Such a study should involve representatives from the affected neighborhood, businesses, and road. Their input should be sought early in the decision-making process and maintained through implementation of the agreed-upon plan.
Potential Difficulties	Diverted traffic may contribute to safety or operational problems at adjacent intersections or on alternative routes, resulting in no net benefit. Owners of properties where access would be reduced, especially commercial operations, may oppose this strategy. Care should be taken during the transition period, both before and after the intersection leg is closed, to alert drivers to the changes as they approach the section involved. Another potential difficulty is in loss of local access for emergency vehicles on the approach being closed. Design solutions may need to be considered, including mountable curbs as well as vegetation or other barriers that can be driven through or over in an emergency.
Appropriate Measures and Data	Key process measures include the number of intersections with legs closed and the change in the number of conflicts due to closure. The latter may also be used as a surrogate safety measure. Crash frequency and severity by type are key safety effectiveness measures. Separate analysis of crashes targeted by such intersection relocations is desirable. Where issues of potential effect on commercial operations exist, economic impact measures may be needed that reflect the change in sales or other measures of economic activity. Crash frequency and severity data are needed for the existing intersection and the intersection with the closed leg. Traffic volume data are needed to represent exposure. In some cases, sales and other economic data may be needed to assess impacts on commercial operations whose access is affected. Net change in conflicts may be used as a surrogate measure of safety until adequate crash data are available.
Associated Needs	PI&E is central to successful use of this strategy.
Organizational and Institutional Attributes
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agency policies concerning geometric design of intersections should address the appropriate application and potential benefits of closure of intersection legs. Nearly any highway agency can participate in the implementation of this strategy. While the strategy is applicable to both rural and urban locations, the most likely use of this strategy will be by agencies that operate extensive systems of urban and suburban arterials. In some cases, public transit service may be affected by the closure. Therefore, care must be taken to establish communication and participation among all public agencies potentially affected.
Issues Affecting Implementation Time	This strategy will likely require an implementation time of at least 1 year to provide time to work out the details of street closure and to communicate the plan to affected businesses and residents.
Costs Involved	Costs to implement this strategy are highly variable. Where mere closure of an intersection leg is all that is needed, costs are low, especially if the closure will be implemented with barricades or other low-cost devices. In other cases, modifications to the closed street and improvements required due to diversion of traffic to a different intersection (such as signing, improved signal timing at nearby intersections, etc.) may require substantially higher expenditures.
Training and Other Personnel Needs	Use of this technique should be included in training concerning geometric design and traffic control issues.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Closure of an intersection leg is compatible with most strategies for improving signalized intersection safety.
Other Key Attributes to a Particular Strategy	This strategy is primarily appropriate for urban and suburban intersections where reasonable alternative access or routes are readily available.

Strategy 17.2 B5: Construct Special Solutions (T)

This strategy includes the following:

• Providing indirect left turn,

• Reconstructing intersections, converting intersections to roundabouts,

• Convert two-way streets to a one-way pair, and

• Constructing interchanges.

General Description

Signalized intersections may have such a significant crash problem that the only alternative is to change the nature of the intersection itself. These types of projects will be high cost and require substantial time for implementation. As such, they will generally not be applicable for agency programs focusing on low-cost, short-term solutions. Note that implementing these strategies will also necessitate significant public involvement and stakeholder activity. Nonetheless, these strategies are outlined here to provide a complete picture of the range of solutions to signalized intersection safety.

Provide Indirect Left Turn. As traffic growth on arterial roadways continues to result in congestion and safety problems at major (high-volume) at-grade intersections, indirect leftturn designs are increasingly being considered and constructed. A few indirect left-turn designs are relatively common to some areas, while many involve rather innovative solutions. These projects may result in major reconstruction of an intersection or conversion to interchanges. ROW restrictions are commonly a determining factor when choosing an alternative. A longer-term, higher-cost design may be the best solution to severe operational and/or safety problems at an intersection.

EXHIBIT V-16
Jughandle Intersection
Source: Federal Highway Administration, in press.

Safety problems associated with left-turns at signalized intersections are magnified at high-volume intersections—or, at least, intersections with high volumes of left turns. Indirect left-turn treatments, such as jughandles before the crossroad, directional median crossovers, and loop roadways beyond the crossroad, can address both safety and operational problems related to left turns. These treatments remove the leftturning vehicles from the traffic stream without causing them to slow down or stop in a through-traffic lane, thereby reducing the potential for rear-end crashes with through vehicles. Right-angle crashes are also likely to decrease after indirect leftturn treatments are implemented, since the turning movement is relocated or changed to a different maneuver. Such treatments are effective on divided highways with medians too narrow to accommodate left-turn lanes and on approaches without enough room for a turn lane long enough to provide sufficient storage capacity. In some cases, it is possible to implement indirect left turns using appropriate signing. However for other designs, implementation costs could be quite high, and the time required to implement could be quite lengthy. If ROW must be acquired, care should be taken to ensure that safety problems are not transferred to nearby intersections if drivers choose alternative routes, such as in cases where less convenient turn arrangements result. Clear signing is a necessity for indirect leftturn designs, especially if there are not similar treatments at other intersections in an area.

This strategy should reduce rear-end collisions resulting from the conflict between vehicles waiting to turn left and following vehicles as well as right-angle collisions resulting from the conflict between vehicles turning left and oncoming through-vehicles.

Alternative left-turn designs are discussed in various publications and will be included in the forthcoming FHWA Signalized Intersection Guide. One option is to convert the intersection to a roundabout (see next substrategy), and various other options for alternative left-turn designs that may be considered are shown in Exhibit V-17 through Exhibit V-21.

EXHIBIT V-17
Median U-Turn Crossover
Source: Federal Highway Administration, in press.

EXHIBIT V-18
Super Street Median Crossover
Source: Federal Highway Administration, in press.

EXHIBIT V-19
Quadrant Roadway Intersection
Source: Federal Highway Administration, in press.

EXHIBIT V-20
Split Intersection
Source: Federal Highway Administration, in press.

EXHIBIT V-21
Continuous Flow Intersection
Source: Federal Highway Administration, in press.

Convert to Roundabout. A roundabout can potentially have a better crash experience than a conventional signalized intersection. The FHWA publication Roundabouts: An Informational Guide (Robinson et al., 2000, available online at http://www.tfhrc.gov/safety/00-0671.pdf), summarizes the current state of practice and should be consulted as a source of information on the design, operation, and safety of roundabouts. The types of conflicts that occur at roundabouts are different from those occurring at conventional intersections; namely, conflicts from crossing and left-turn movements are not present in a roundabout. The geometry of a roundabout forces drivers to reduce speeds as they proceed through the intersection. This helps keep the range of vehicle speed narrow, which helps reduce the severity of crashes when they do occur. Pedestrians only have to cross one direction of traffic at a time at roundabouts, thus reducing their potential for conflicts. However, vehicles in the circulating roadway are not required to stop by virtue of the presence of a positive traffic control device.

The safety performance of roundabouts in the United States is not well documented due to the small number that have been built here. A current NCHRP study (Project 3-65) involves a comprehensive analysis of the operation and safety of roundabouts in the United States. However, recent conversion of several intersections in the United States to roundabouts showed a reduction in crash rates after construction. Data collected after intersections were converted to roundabouts in Europe and Australia show that reductions in crash rates resulted. Single-vehicle crash rates for roundabouts, however, tend to be higher than for conventional intersections (Robinson et al., 2000)

A comparison was made of crash rates for conventional intersections (using a U.S. crashprediction model) and rates for roundabouts (using a model from the United Kingdom). For volumes entering the intersection of 20,000 vehicles/day, the crash rate was 33 percent lower for roundabouts than for signalized intersections in urban/suburban areas and 56 percent lower in rural areas. For an entering Average Daily Traffic (ADT) of 40,000 vehicles/day, approximately 15 percent fewer crashes are predicted for roundabouts. At higher volumes (50,000 or more entering vehicles per day), the safety performance of roundabouts and signalized intersections is probably comparable, and it is less likely there will be a safety benefit of conversion to a roundabout. (ITE Traffic Safety Toolbox [Institute of Transportation Engineers, 1999b]).

Studies do not show an improvement in rates for crashes involving bicyclists. In addition, since single-vehicle crash rates tend to increase after conversion to a roundabout, central islands, splitter islands, and the clear zone on the perimeter of the roundabout should be kept clear of obstacles.

Roundabouts are not appropriate for every intersection and will not solve all safety problems at specific intersections. Volumes of traffic entering the intersection are an important factor in the effectiveness of a roundabout, along with turning movements and other operational characteristics. All of these should be studied to ensure that a roundabout is an appropriate design for a given intersection.

Indirect left-turn treatments and roundabouts are both discussed in further detail in the guide for crashes occurring at unsignalized intersections. European uses of roundabouts are described in Appendix 4.

Convert Two-Way Streets to a One-Way Pair. When two-way streets are converted to oneway streets, it is generally for the purpose of increasing capacity, but the removal of opposing traffic flows can improve safety as well. Removal of one direction of traffic from a two-way street allows for better signal synchronization and progression of platoons. Smooth progression and reduced congestion can reduce rear-end crashes. In addition, the removal of one direction of traffic can reduce congestion and improve safety by

• Reducing the number of vehicle/vehicle conflict points at intersections,

• Allowing for unopposed turn maneuvers,

• Simplifying operations and signal phasing at multileg intersections,

• Allowing pedestrians to only have to deal with traffic from one direction, reducing conflicts with vehicles, and

• Providing more gaps for vehicles and pedestrians at unsignalized crossings.

The ITE Traffic Safety Toolbox (Institute of Transportation Engineers, 1999b) reports that studies have shown a 10- to 50-percent reduction in total crashes after conversion of a twoway street to one-way operation. At the same time, this strategy increases capacity significantly; a one-way street pair can handle up to 50 percent more volume than two parallel two-way streets.

Safety-related drawbacks to conversion to one-way streets may include the following:

• Pedestrians not looking in the correct direction for oncoming vehicles, and

• Minor sideswipe crashes related to weaving maneuvers as drivers attempt to park or reach a turn lane.

Supplemental and redundant signing is recommended on one-way streets converted from two-way operation.

Transit operations may be adversely affected with the introduction of one-way operation. Special care should be taken to avoid contraflow bus lanes on such streets, as these present a special hazard to crossing pedestrians. One key element of creating a one-way pair is the design of the transitions at each end. Care must be taken not to create conditions that cause driver confusion and erratic maneuvers.

Construct Interchange or Grade Separation. At some signalized intersection locations with extremely high volumes, extremely poor crash histories, or other mitigating factor(s), provision of a grade separation or interchange can be considered. This is an expensive approach to safety improvement and should generally be considered when other, less restrictive measures have been tried and have failed. Often this solution is applied for capacity and operational reasons, where the capacity of a signalized intersection is insufficient to accommodate the volume passing through it. If a grade separation alternative is considered, maintaining access to existing development is a key element to successfully implementing the improvement.

By separating the grades of intersecting roadways, volumes of crossing and turning traffic, as well as bottlenecks and spot congestion, may be reduced. This can lower the number and severity of crashes caused by these movements and intersection conditions, specifically rearend and angle crashes.

Guidance on constructing and designing interchanges is discussed in the AASHTO Policy on Geometric Design of Highways and Streets (American Association of State Highway and Transportation Officials, 2001). Time required for the design and construction of an interchange could range anywhere from 4 to 10 years, depending upon location constraints and environmental factors. Costs would likely be high and variable. Simple service interchanges are less expensive than system interchanges, but construction costs may still be several million dollars.

Objective 17.2 C—Improve Sight Distance at Signalized Intersections

Adequate intersection sight distance contributes to the safety of the intersection. In general, sight distance is needed at signalized intersections for the first vehicle stopped at an approach to be able to see the first vehicles stopped at the other approaches, for drivers making permitted left turns, and for right-turning vehicles. Where RTOR are allowed, adequate sight distance should be available. Improvements in sight distance can lead to a reduction in crashes caused by drivers stopping suddenly (rear-end), drivers proceeding through the intersection when the signal has not assigned them the right-of-way (angle), and drivers turning through an inadequate gap in opposing traffic (angle).

Strategy 17.2 C1: Clear Sight Triangles (T)

General Description

Sight distance improvements can often be achieved at relatively low cost by clearing sight triangles to restore sight distance obstructed by vegetation, roadside appurtenances, buildings, bus stations, or other natural or man-made objects.

The most difficult aspect of this strategy is the removal of sight restrictions located on private property. The legal authority of highway agencies to deal with such sight obstructions varies widely, and the time (and possibly the cost) to implement sight distance improvements by clearing obstructions may be longer if those obstructions are located on private property. If the object is a mature tree or planting, then local concerns over adverse environmental consequences may arise. For a more detailed discussion of trees, see the guide for crashes involving trees in hazardous locations.

Research has established a relationship between intersection safety and sight distance at unsignalized intersections. No such research quantifies the effectiveness of improving sight distance at signalized intersections. One may expect that crashes related to inadequate sight distance (specifically, angle and turning related) would be reduced if the sight distance problems are improved. However, as the signal assigns ROW for most vehicles crossing paths at right angles and because traffic volumes affected by the other situations cited above are low, the overall impact on crashes could be relatively small.

Since sight distance is a greater issue at intersections with stop control than at signalized intersections, more research has been performed on the effectiveness of sight distance improvements at stop-controlled intersections (several of the studies are summarized in NCHRP Report 440 [Fitzpatrick et al., 2000]). There are several movements at signalized intersections that operate similarly to stop-controlled intersections (such as RTOR and permitted left turns) for which expected effectiveness of sight distance improvements at signalized intersections may be inferred from similar studies at stop-controlled intersections. Such estimates should be performed with caution, taking into consideration the other characteristics of signalized intersection operation that would alter the effectiveness estimates.

More information on clearing roadside and median intersection sight triangles is presented in the guide for crashes at unsignalized intersections.

Strategy 17.2 C2: Redesign Intersection Approaches (P)

General Description

Signalized intersections with sight-distance-related safety problems that cannot be addressed with less expensive methods (such as clearing sight triangles, adjusting signal phasing, or prohibiting turning movements) may require horizontal or vertical (or both) realignment of approaches. Realigning both of the minor-road approaches so that they intersect the major road at a different location, or a different angle, can help address horizontal sight distance issues.

This is a high-cost, longer-term treatment for the intersection, but if completed according to applicable design policy, it should help alleviate crashes related to sight distance. The 2001 AASHTO Policy on Geometric Design of Highways and Streets contains updated sight distance guidelines, and these guidelines should be considered when revising intersection approach geometry. There are significant ROW and property access issues involved in this strategy, and public information campaigns are vital to the success of the intersection improvements.

An intersection leg can be closed in order to address sight distance issues related to that particular leg. This strategy is covered in the previous section of this guide.

Intersection relocation and closure, elimination of intersection skew, and offsetting of leftturn lanes are all strategies that involve improvements to approach alignment to improve sight distance. These strategies are each covered in greater detail in the unsignalized intersection guide.

Objective 17.2 D—Improve Driver Awareness of Intersections and Signal Control

Driver awareness of both downstream intersections and traffic control devices is critical to intersection safety. The inability to perceive an intersection or its control or the back of a stopped queue in time to react as necessary can result in safety problems. Drivers caught unaware could be involved in serious crashes, especially at intersections with high speeds on the approaches. This objective details strategies aimed at improving driver awareness of signalized intersections and the traffic control in place.

Strategy 17.2 D1: Improve Visibility of Intersections on Approach(es) (T)

This strategy includes the following:

• Improving signing and delineation,

• Installing larger signs,

• Providing intersection lighting,

• Installing rumble strips on approaches, and

• Installing queue detection system.

General Description

Some crashes at signalized intersections may occur because drivers are unaware of the presence of an intersection or are unable to see the traffic control device in time to comply. These crashes are generally rear-end or angle collisions. The ability of approaching drivers to perceive signalized intersections immediately downstream can be enhanced by signing, delineation, and warning devices. Other strategies to improve the visibility of an intersection include providing lighting, improving the visibility of the signals, and using devices to call attention to the signals. All of these strategies are discussed in detail in the unsignalized intersection guide, and that guide should be consulted for additional information.

The FHWA report Synthesis of Human Factors Research on Older Drivers and Highway Safety: Volume 2 (Staplin et al., 1997) reviews research on older drivers' visual abilities related to driving. Research shows that recognition and legibility distances as well as response speeds are lower for older drivers than for younger ones. The Synthesis summarizes recent research by stating that if recognition of an intersection is based on signs being legible to drivers, older drivers will take longer to recognize intersections. Therefore, consideration should be given to providing traffic control devices that contribute to improved legibility and response times for older drivers. This may include redundant signing, overhead signing, and advanced route signing. The older drivers guide should be consulted for more information.

Improve Signing and Delineation. Installing or upgrading signs and pavement markings on intersection approaches can help better prepare drivers for the intersection ahead. This may include advance guide signs, advance street name signs (Exhibit V-22), warning signs, pavement markings, overhead street signing, and post-mounted delineators. Advance warning signs, such as the standard intersection warning sign or the standard sign with flashers, can also alert drivers to the presence of an intersection. Installing advance warning signs on both sides of the roadway to provide redundancy in signing may be appropriate in some situations, such as when the intersection approach in on a curve. Street name and lane assignment signs in advance of the intersection prepare drivers for choosing and moving into the lane they will need to use for their desired maneuver. Signs and flashers warning drivers they are approaching a red signal might improve both awareness of the intersection and a red signal (see Exhibit V-23).

EXHIBIT V-22
Advance Street Name Sign
Source: Federal Highway Administration, in press.

EXHIBIT V-23
Advanced Warning Sign for Red Signal
Source: Federal Highway Administration, in press.

Providing a break in pavement markings at intersections also helps to alert drivers to their presence. This includes centerlines, lane lines, and edge lines. Close spacing of progressive signals can lead to drivers focusing on a downstream signal and not noticing an intermediate signal. Signing, pavement markings, lighting, or alteration of the appearance of a signal to make it more noticeable should be considered in this situation.

Maintenance of signs and pavement markings is also important to the success of this strategy. Retroreflectivity of older pavement markings and signs should be checked periodically to determine whether replacement is needed.

Install Larger Signs. The visibility of intersections with existing regulatory and warning signs and the ability of drivers to perceive the signs can be enhanced by installing signs with larger letters. Such improvements may include advance guide signs, warning signs, pavement markings, and post-mounted delineators. The FHWA Older Driver Highway Design Handbook (Staplin et al., 1998; available at http://www.fhwa.dot.gov/tfhrc/safety/pubs/older/intro/) encourages installation of larger signs to contribute to a better driving environment for older drivers. The older driver guide contains additional information about engineering improvements to aid older drivers.

Provide Intersection Lighting. Providing lighting at the intersection itself or at both the intersection and on its approaches can make drivers aware of the presence of the intersection and reduce nighttime crashes. Crash data should be studied to ensure that safety at the intersection could be improved by providing lighting (this strategy would be supported by a significant number of crashes that occur at night). The costs involved with intersection lighting may be moderately expensive, especially since maintenance is needed to keep the equipment in working order.

Intersection lighting is of particular benefit to police officers. Lighting not only helps them perform their duties, such as traffic stops, but also helps drivers see them better, especially when out of the vehicle.

Install Rumble Strips on Approaches. Rumble strips can be installed in the roadway on intersection approaches transverse to the direction of travel to call attention to the presence of the intersection and the traffic control used. Rumble strips are particularly appropriate on intersections where a pattern of crashes related to lack of driver recognition of the presence of the signal is evident, often on high-speed approaches. This strategy should be used sparingly, as the effectiveness of rumble strips is dependent on their being unusual. Rumble strips are normally applied when less intrusive measures, such as “signal ahead” signs or flashers, have been tried and have failed to correct the crash pattern, and they are typically used in combination with the advance warning signs. For example, a rumble strip can be located in the roadway so that when the driver crosses it, a key traffic control device such as a “signal ahead” sign is directly in view. Rumble strips in the traveled way can also be used on a temporary basis to call attention to changes in traffic control devices, such as installation of a signal where none was present before. Care must be taken to avoid use of rumble strips where the noise generated will be disturbing to adjacent properties. NCHRP Synthesis of Highway Practice 191 (Harwood, 1993) reviews the state of the art of rumble strip usage.

Install Queue Detection System. Queue detection systems are standard tools for operation of traffic signals. In normal practice, queue detection is used for actuated signal systems to “call-up” a phase given the presence of a vehicle in a specific lane or movement.

The application of queue detection systems as safety devices is a new and potentially effective device. One such system has been implemented in Oregon on an approach to a signalized intersection in a rural setting that regularly experiences significant queues, especially during the summer when seasonal traffic increases. Two loop detectors in each lane on the intersection approach detect when a vehicle is stopped at that location. The detectors are connected to an overhead sign with beacons located a half mile upstream. The sign contains the message “Prepare to stop when lights flash.” When a vehicle is continuously present at a detector, beacons on the overhead sign flash to warn drivers of the stopped vehicle ahead. A preliminary evaluation indicates a reduction in crashes after installation of this system, but additional data are needed to determine if other factors contributed to this decrease. For additional information on this system, refer to the FHWA report Safety Applications of ITS in Rural Areas, (Federal Highway Administration, 2002), available online at http://www.itsdocs.fhwa.dot.gov/JPODOCS/REPTS_TE/5_1_1.htm.

Strategy 17.2 D2: Improve Visibility of Signals and Signs at Intersections (T)

General Description

Lack of visibility of traffic control devices may contribute to crash experience at signalized intersections. Visibility of traffic signals and signs at intersections may be obstructed by physical objects (such as signs or other vehicles) or may be obscured by weather conditions, such as fog or bright sunlight. Also, drivers' attention may be focused on other objects at the intersection, such as extraneous signs. Poor visibility of signs and signals may result in vehicles not being able to stop in time for a signal change or otherwise violating the intended message of a regulatory or directional sign. Providing adequate visibility of signs and signals also aids in drivers' advance perception of the upcoming intersection. The FHWA Older Driver Highway Design Handbook should be consulted to ensure that improvements to visibility of traffic control devices will be adequate for older drivers (Staplin et al., 1998; available at http://www.fhwa.dot.gov/tfhrc/safety/pubs/older/home/index.htm).

In addition to potentially restricting driver sight lines, large numbers of appurtenances and signage not associated with the driving task in the vicinity of an intersection can impose a high workload. This may ultimately distract the driver from the task at hand (i.e. safely navigating the intersection). This “visual clutter” can make it difficult for the driver to extract the information from the signs required to execute the driving task (directional information, speed information, etc.). Enforcement of existing sign restrictions, and/or the creation of new restrictions limiting the placement of signs near intersections, can reduce the amount of information provided to the driver near intersections.

Maintenance of signals and signs is important to the visibility of the devices. If visibility of traffic control devices is considered to be a potential factor in crashes, a field review should be performed to determine if part of a sign's message is covered, obliterated, or blocked, as well as to check the reflectivity of the sign.

Methods for improving visibility of traffic signals and signs include the following:

• Install additional signal head (see Exhibit V-24),

• Provide visors to shade signal lenses from sunlight,

• Provide louvers, visors, or special lenses so drivers are able to view signals only for their approach,

• Install backplates,

• Install larger (12-in.) signal lenses,

• Remove or relocate unnecessary signs, and

• Provide far-side left-turn signal.

Methods for improving visibility of traffic signals and signs are discussed further in Appendix 7. Additional information on improving signal visibility to reduce red-light running can be found in Making Intersections Safer: A Toolbox of Engineering Countermeasures to Reduce Red-Light Running (McGee, 2003; available online at http://www.ite.org/library/redlight/MakingInt_Safer.pdf).

EXHIBIT V-24
Signal Head with Double Red Section
Source: Federal Highway Administration, in press.

EXHIBIT V-25
Strategy Attributes for Improving Visibility of Signals and Signs at Intersections

Attribute	Description
Technical Attributes
Target	This strategy is targeted at crashes that occur because drivers are unable to see traffic signals and signs sufficiently in advance to safely negotiate the intersection being approached. Crash types would include angle and rear-end crashes.
Expected Effectiveness	Improved visibility and awareness of traffic control information are expected to reduce conflicts related to drivers not being able to see the device well or in enough time to comply with the signal indication or sign message (such as those resulting in rear-end and right-angle crashes).
Keys to Success	Visibility and clarity of the signal should be improved without creating additional confusion for drivers. Additional signing to warn drivers should not clutter the intersection and should not present confusing or conflicting messages to drivers.
Potential Difficulties	Care should be taken to ensure that new or relocated signs do not present additional sight distance, roadside, or driver distraction hazards. If rumble strips are used in an area with adjacent residences, the noise may be objectionable, creating public resistance. Bicyclists may also object to rumble strips as the treatment may force them to ride in the roadway travel lanes. If some of the devices recommended are not maintained properly, the expected benefits may be lost.
Appropriate Measures and Data	Key process measures include the number of intersection approaches for which improvements are implemented, which increase driver perception of traffic signals located immediately downstream, and the number of conflicts potentially eliminated by the improvement. Crash frequency and severity by type of crash are key safety effectiveness measures. It is especially useful to identify crashes related to unseen signals (inadvertent red light running, etc). Driver behavior measures (e.g., conflicts, erratic maneuvers, speeds, and braking) may be used as surrogate safety measures. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to the targeted improvements at the intersection should be analyzed separately. Traffic volume data are needed to represent exposure, especially the volumes of movements of interest and the opposing through volumes.
Associated Needs	Removing signs and other elements contributing to roadside clutter may require public involvement activities.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding use of traffic control devices to ensure appropriate action is being taken on routine projects. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas.
Issues Affecting Implementation Time	Implementation time will be relatively short for procedures to install new signs, improve signals, and remove or relocate signs.
Costs Involved	Costs will be low for most procedures to install or upgrade signs and signals to improve visibility and awareness of the traffic control devices. Ongoing maintenance costs should be included when considering use of these devices.
Training and Other Personnel Needs	Visibility of traffic control devices should be addressed in highway agency training concerning traffic signal installation and human factors.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Actions taken to improve visibility of signals are compatible with most other strategies to improve signalized intersection safety but are not appropriate in conjunction with removal of a signal.
Other Key Attributes to a Particular Strategy	None identified.

Objective 17.2 E—Improve Driver Compliance with Traffic Control Devices

Safety problems at signalized intersections cannot always be solved only with engineering countermeasures. Enforcement of traffic regulations or public information campaigns may be the best way to improve intersection safety. This section details information on strategies to improve compliance with traffic control devices, focusing mainly on red-light running and speeding on approaches to signalized intersections.

Strategy 17.2 E1: Provide Public Information and Education (T)

General Description

Providing targeted public information and education (PI&E) on safety problems at intersections is a preventive measure that can help improve driver compliance with traffic control devices and traffic laws. PI&E programs generally add effectiveness to targeted enforcement programs, as well.

Another option is to develop public information campaigns aimed at specific drivers who violate regulations at intersections, even though it is often difficult to identify and focus upon a subset of the driving population using a specific intersection. Therefore, an areawide program is often the preferred approach. A key to success for this strategy is reaching as much of the targeted audience as possible, whether it is through television, radio, distribution of flyers, driver education classes, or other methods. Targeted drivers need to be defined both in terms of the location of the hazardous intersection(s) and the attributes of the drivers who may have been identified as over represented in the population involved in crashes. More information on public information that is targeted at specific drivers is provided in the guide for crashes occurring at unsignalized intersections.

EXHIBIT V-26
Strategy Attributes for Providing Public Information and Education (T)

Attribute	Description
Technical Attributes
Target	The target for this strategy is crashes related to drivers either being unaware of, or refusing to obey, traffic laws and regulations that impact traffic safety. Crashes related to red-light running, speeding, and not yielding to pedestrians could be reduced with PI&E campaigns.
Expected Effectiveness	Data on the effectiveness of this strategy for this specific application are not known, but it is expected that providing information to drivers will help improve safety at intersections. It may not be possible to identify or reach the entire audience that would benefit from a PI&E campaign.
Keys to Success	Keys to success include identifying and reaching as much of the intended audience as possible, providing information in nontechnical terms, and providing agency personnel to answer questions and calls from the public. It is important to motivate people to drive (and bike or walk) safely. Since unsafe actions do not always result in crashes, road users may have a false sense of security and may not see the need to drive more safely or follow traffic regulations in all circumstances. Use of trained public information specialists is important for program success. Establishing good relationships with media representatives will be extremely helpful for maximizing coverage and impact.
Potential Difficulties	The primary potential difficulty associated with this strategy is relating the importance of informational/educational programs to the public. Many people may see a notice for a public meeting and think that it is a waste of time and not attend. Brochures, posters, and advertisements can be effective if they are conspicuous and readily available. Use of electronic media is expensive, unless strategies are employed for receiving donated time. Consideration should be given to people who may need materials in languages other than English or in alternative formats to accommodate disabilities. Another difficulty is maximizing the reach of a public involvement program.
Appropriate Measures and Data	Key process measures include identifying the number and frequency of different media used (radio ads, brochures, etc.) and measuring the population exposed to the message. Level of expenditure is another possible process measure and can be used to produce a productivity measure. Crash frequency and severity by type of crash should be tracked before and after implementation of the public information campaign. Traffic volume data are needed to represent exposure. Surrogate safety measures can include driver behavior (e.g., change in unsafe targeted and untargeted driving acts).
Associated Needs	None identified.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should ensure that education and information programs are scheduled when most likely to maximize the exposure of the message to the target population. Coordination and cooperation with other parts of an organization that have established marketing skills can be useful. It will be helpful to enlist media representatives as part of the group of stakeholders involved in planning and implementing the program. Nearly all highway agencies can make use of this strategy, which is applicable in rural, urban, and suburban areas.
Issues Affecting Implementation Time	Implementation time for this strategy should be short to moderate. Extensive planning of the program and design of the educational materials can lengthen the implementation time.
Costs Involved	Cost will generally be low to moderate and depend upon the kinds of materials developed (brochures, posters, radio, or television advertisements), the extent of effort spent on designing the materials, and the amount of free media coverage that can be achieved. Strategies may be used to attract media attention without purchasing air time.
Training and Other Personnel Needs	Training for highway personnel in providing PI&E should be included as a part of developing and implementing the program.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at signalized intersections. It may be used in conjunction with overall traffic safety public service campaigns.
Other Key Attributes to a Particular Strategy	None identified.

Strategy 17.2 E2: Provide Targeted Conventional Enforcement of Traffic Laws (T)

General Description

Enforcement is a potential countermeasure to unsafe and illegal motorist behavior at intersections. Studies report the reduction of traffic law violations when enforcement is used (Traffic Engineering Handbook [Pline, 1999]). Traffic law enforcement agencies will often select locations for targeted enforcement when crash, citation, or other sources of information suggest that the site is unusually hazardous due to illegal driving practices, such as speeding or red-light running. These actions can lead to rear-end, head-on, sideswipe, angle, and pedestrian or bicycle-related crashes.

Traffic law enforcement methods vary depending upon the type of program being implemented. For background on methods and approaches please refer to the publications available at the following Web sites: http://www.nhtsa.dot.gov/people/injury/enforce/DESKBK.html and http://www.nhtsa.dot.gov/people/injury/enforce/millennium/index.htm).

Targeted enforcement of traffic laws is a short-term, moderate-cost measure to address sitespecific signalized intersection safety. Though this is an effective strategy, the effectiveness has often been found to be short-lived. It is difficult—if not impossible—to provide constant enforcement of traffic regulations due to funding and staffing reasons, so periodic enforcement may be necessary to sustain the effectiveness of this strategy. For European experience on the effectiveness of traffic law enforcement, including speeding laws, see Appendix 11 and Appendix 12.

It is important to correctly identify intersections that would benefit from enforcement. Care should be taken to first ensure that the existing signals are operating properly, are visible, and meet MUTCD requirements, as well as that timing plans—including clearance intervals—are appropriate. Analysis of crash statistics can help with this process, as can spot speed or conflict studies. In some cases, public input or observations by law enforcement personnel may suggest that a location should be targeted for enforcement.

Police officers providing targeted enforcement of red-light running can be aided by “telltale” or “tattle-tale” lights. These lights are placed at traffic signals, but facing away from oncoming traffic. Police officers are able to wait in their vehicles on the downstream side of the traffic signal and view the tattle-tale light. This way, they are able to pursue red-light runners without also running through the red light themselves (and possibly into vehicles entering the intersection from the cross street).

Targeted enforcement at intersections is covered in more detail in the unsignalized intersection guide.

Strategy 17.2 E3: Implement Automated Enforcement of Red-Light Running (Cameras) (P)

General Description

Red-light running is a well-documented and growing traffic safety problem. Various engineering countermeasures can be used to address red-light running; refer to Strategy 17.2 A2 for a discussion of optimizing clearance times, as well as the ITE and FHWA report Making Intersections Safer: A Toolbox of Engineering Countermeasures to Reduce Red-Light Running (McGee, 2003; available online at http://www.ite.org/library/redlight/MakingInt_Safer.pdf). While some occurrence of this can be addressed by engineering, in many instances inappropriate driver behavior is the primary problem. Because it is not feasible to provide police officers to enforce traffic signals as often or in as many locations as an agency might need, automated enforcement is an attractive alternative (see Exhibit V-27).

Automated enforcement refers to the use of photo radar and video camera systems connected to the signal control. Such systems record vehicles proceeding through the intersection after the signal displays red. Red-light-running cameras turn on after the signal turns red. A detector senses approaching vehicles and sends a signal to the camera, which photographs the vehicles as they enter the intersection. It is possible to set a grace period (generally one second) so that the cameras do not photograph people who were caught in the dilemma zone and enter the intersection just after the signal turns red. Data on the violation, such as date, time, speed of vehicle, and the time that had elapsed since the signal turned red, are printed on the photograph. Police officers review the photos to determine if a violation occurred, and if so, a citation is mailed to either the driver or the vehicle owner, depending on the legislation for the jurisdiction.

EXHIBIT V-27
Red-Light-Running Camera
Source: Federal Highway Administration, in press.

Automated enforcement of red-light running has been shown to significantly decrease violations, not only at intersections where cameras are installed, but also at other intersections in the area.

There are many applications of red-light-running technology throughout the world. For additional information on red-light-running cameras available on the Internet, see http://www.iihs.org/safety_facts/rlc.htm, http://safety.fhwa.dot.gov/programs/srlr.htm, and http://www.ite.org/library/redlight/index.asp.

In Canada, cameras are being used for the purpose of speed enforcement as well as red-light running. Strathcona County in Alberta was the first jurisdiction in North America to use redlight cameras to record speeding violations. A red-light-running camera installed at the intersection of Wye Road and Ordze Avenue in 1998 was used for speed enforcement beginning in 2000. This strategy has been effective in reducing speeding: a 75-percent drop in violations was experienced from 2001 to 2002. For additional information, visit the Royal Canadian Mounted Police Web site for Strathcona County ( http://www.strathconacountyrcmp.ca/redlightcam.htm).

Red-light-running camera equipment can be used not only to record violators but also to protect cross-street vehicles. There are systems capable of using a vehicle's speed to predict whether the vehicle will run the red light, and if so, the system can extend the cross-street red indication to prevent cross-street vehicles from entering the intersection while the violating vehicle is still in the intersection.

Red-light-running cameras are used in Europe as well; refer to Appendix 4 for additional information on several countries' experiences. Automated enforcement offers the opportunity to address a systemwide problem in an efficient manner, using technology as a substitute for law enforcement personnel time. While the advantages in terms of safety and cost-effectiveness are clear, there are problems and in some cases controversies associated with automated enforcement. Concerns over invasion of privacy and the ability to identify and cite the driver (most systems identify the vehicle through identification of the plates), as well as the belief by some that the systems are unfair or intended to generate fine revenue versus address safety problems are all issues that have arisen. In some jurisdictions, enabling legislation may be needed for successful prosecution of red-light-running camera violations. The National Conference of State Legislatures tracks status of legislation and provides examples of model legislation (http://www.nhtsa.dot.gov/ncsl/Index.cfm).

EXHIBIT V-28
Strategy Attributes for Providing Automated Enforcement of Red-Light-Running (Cameras) (P)

Attribute	Description
Technical Attributes
Target	The target for this strategy is drivers who intentionally disobey red signal indications. Intersections where observations or crash histories indicate a red-light-running problem may be suitable for installation of a red-light-running camera. Crashes of this type are likely to be angle and rear-end collisions.
Expected Effectiveness	Several studies have shown the effectiveness of automated red-light enforcement in reducing red-light violations and crashes related to those violations. Fairfax, Virginia, experienced a 44-percent reduction in violations during the first year of operation. Two other sites in the city that did not have cameras experienced decreases in violations of 34 percent. Control sites in nearby counties experienced little change. (Retting et al., 1999a). Oxnard, California, experienced approximately 41 percent fewer red-light violations within a few months after beginning to use the cameras to enforce the signals (Retting et al., 1999b). FHWA has made a general estimate of a 15-percent reduction in red-light-running incidents resulting from these programs. The ITE Informational Report, Automated Enforcement in Transportation (Institute of Transportation Engineers,1999a), contains information on experiences with red-lightrunning cameras in other jurisdictions. The automated enforcement programs highlighted in this document experienced a range of reduction in violations of 23 to 83 percent. An evaluation of a program in Victoria, Australia, showed a decrease in violations of 35 to 60 percent, accompanied by a 35-percent reduction in right-angle crashes, 25-percent reduction in right-angle turning crashes, 31-percent reduction in rear-end crashes, and 28-percent reduction in rear-end turning crashes.
Keys to Success	Keys to success of red-light-running programs primarily relate to acceptance of local stakeholders, including both officials and the public. Acceptance by local law enforcement is another critical element necessary for the success of an automated enforcement program. Indeed, local law enforcement needs to be seen as central to such a program. So, incorporation of a public information campaign explaining the program, the need for it, how the cameras work, and what the benefits may be is a key to successful implementation. Successful red-light camera programs have generally begun as safety improvement programs. Programs that are perceived as revenue generators (i.e., through collection of fines) are generally not well-accepted. Therefore, there should be clear justification of the installations based upon documented violation levels. It is important that both the highway agency and the law enforcement agency(ies) in the jurisdiction be involved jointly in planning and operating the program. Moreover, where private contractors are used to implement parts of the program, it is important that their contract and compensation not be directly linked to revenue or tickets issued. Some programs have lost public support because it was perceived that a private company was profiting from traffic ticket revenue. Avoiding controversial contract provisions and maintaining clear control over administration of the program by the appropriate police agency are keys to success. FHWA and NHTSA developed guidelines for implementation of red-light-running cameras, which can be found online at http://safety.fhwa.dot.gov/rlcguide/index.htm. See Appendix 8 for a description of a successful red-light camera program in Howard County, Maryland. Adequate legislative authority is needed to conduct such a program (see the section below on “Legislative Needs”).
Potential Difficulties	here are many opponents to red-light-running enforcement cameras. Arguments against this strategy include violation of personal privacy, violation of constitutional rights, lower effectiveness than other types of enforcement, high cost outweighing the benefits, and implementation solely to generate revenue. Recent challenges also include questioning the precision of the cameras and the proper setting of the camera. Counter-arguments to all of these issues are presented in ITE's Automated Enforcement in Transportation. Other potential difficulties to overcome include administration of the program through the use of contractors. The program needs to be clearly identified as a public safety program, with administrative responsibility remaining clearly in the hands of the police. Timeliness of the citation is important. Administrative systems may slow the processing and create a lengthy gap between the moment of the violation and the moment the sanction is received by the driver. Principles of effectiveness of enforcement suggest that the time between violation and punishment should be short to be effective. In addition, the technology has spawned an industry directed at defeating it. For example, see http://www.phantomplate.com.
Appropriate Measures and Data	A key process measure is the number of intersection approaches on which red-lightrunning cameras are installed. A more detailed measure is the number of citations issued from the program, as well as the number of traffic convictions resulting. Crash frequency and severity data by type are key safety effectiveness measures. Data describing these crashes and data on the frequency of violations are needed for periods both before and after installation of the cameras. Traffic volume data are needed to represent exposure. Where feasible, the effect of automated enforcement on total crashes and crash types potentially related to signal violations should be evaluated separately. Surrogate safety measures include violation frequency and the number of intersections where potential benefits may be observed even though cameras are not installed.
Associated Needs	PI&E is needed to make automated enforcement successful. Public opinion and acceptance can “make or break” an automated enforcement program. Information and awareness efforts and materials typically include the following: (1) documentation of the problem (in nontechnical terms), (2) objectives of the automated enforcement program, (3) advantages of automated enforcement or conventional enforcement, (4) general locations or areas of automated enforcement systems, (5) uses of revenue generated by automated enforcement, and (6) information on what to do when a citation is received in the mail. Signs before each approach to the intersection, informing the public that automated technology is being used, have been used to make the public aware of an automated enforcement system. Some jurisdictions have even painted the cameras in highly visible colors. Having members of local law enforcement speak on television and radio shows, or on panels at local meetings, has been helpful to some agencies installing cameras.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agency crash analysis procedures should include methods to identify the need for automated red-light enforcement. It is important that the program be handled in a coordinated manner by the highway, law enforcement, and judicial agencies.
Issues Affecting Implementation Time	The time to implement red-light-running cameras can vary depending upon the extent of public involvement and whether new legislation is needed.
Costs Involved	Costs may vary, depending upon the effort put into public information and need for additional legislation. Equipment costs can vary somewhat, due to the type of camera selected (i.e., 35-mm, video, digital). Costs also include monitoring of the videotapes, issuance of citations, collections and records maintenance, maintenance of equipment, maintaining quality control, and rotating or moving the equipment from location to location. Some agencies have established staffs or hired consultants to perform work associated with program implementation. Some cost information may be found at http://www.iihs.org/safety_facts/rlc.htm Proceeds from citations can be used to cover all or some of the costs of implementing, operating, and maintaining the system. If fines are set sufficiently high, additional monies can be put into a general fund or a special fund targeted for safety improvements. It is important, however, for the program revenue to not be the reason for the program itself.
Training and Other Personnel Needs	Training for highway engineers, safety analysts, and police officers should address automated red-light enforcement. Implementation of the program either through the jurisdiction or private contract requires acquisition of the necessary hardware and software, training in their use, implementation of quality control processes and procedures, and development of new processes to measure program costs and success.
Legislative Needs	Legislation may be necessary before implementing an automated enforcement program. The legislation is necessary to meet constitutional standards, state legal standards, state vehicle code standards, and local jurisdiction standards. A state's enabling legislation should address the broad constitutional issues (federal and state) within a framework that includes elements such as definitions of acceptable automated enforcement devices, any restricted uses, description of acceptable photographic evidence, and penalty provisions. Local legislation should cover requirements in much more detail. This should include issues such as operating criteria, the agency that is responsible for camera operation, restricted uses in that jurisdiction, and requirements for advance notification. An example of both state and local legislation authorizing red-light-running programs may be found at the site explaining the Safelight program in Charlotte, NC. (available at http://www.charmeck.org/departments/transportation/special+programs/safelight.htm). A summary of state legislation on automated enforcement may be found at: http://www.iihs.org/safety_facts/state_laws/auto_enforce.htm. The National Conference of State Legislatures tracks status of legislation and provides examples of model legislation (http://www.nhtsa.dot.gov/ncsl/Index.cfm). Additional information on related legislation can be found at www.stopredlightrunning.com.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at intersections. Indeed, as noted above, this strategy should be accompanied by a public information or outreach campaign to explain the program.
Other Key Attributes to a Particular Strategy	None identified.

Information on Current Knowledge Regarding Agencies or Organizations That Are Implementing This Strategy

See Appendix 8 for a description of a red-light camera program in Howard County, Maryland.

Charlotte, North Carolina, has had a red-light running program in operation since 1998. It is called the “Safelight” program. Detailed information on it may be found at http://www.charmeck.org/departments/transportation/special+programs/safelight.htm.

The City of Portland, Oregon, has a summary of its red-light running program at: http://www.portlandonline.com/police/index.cfm?c=30592.

Also see http://www.ite.org/library/redlight/index.asp for descriptions of other programs.

Strategy 17.2 E4: Implement Automated Enforcement of Approach Speeds (Cameras) (T)

Enforcement of traffic regulations is an important part of an overall intersection safety improvement strategy, but limited resources constrain the efforts police can devote to providing speed enforcement. Traffic law enforcement agencies will often select locations for targeted enforcement when crash, situation, or other sources of information suggest that the site is unusually hazardous due to illegal driving practices. Crash types that might indicate speeding as a concern include right-angle and rear-end collisions. Speed-enforcement cameras (also known as photo radar) are a potential method to use in these locations.

EXHIBIT V-29
Strategy Attributes for Implementing Automated Enforcement of Approach Speeds (Cameras) (T)

Attribute	Description
Technical Attributes
Target	The target for this strategy is drivers who speed on approaches to signalized intersections. Crash types related to these actions include angle and rear-end crashes.
Expected Effectiveness	Automated enforcement of speeds may provide a longer-term effect than on-site enforcement by police officers. It is not feasible to provide officers to constantly enforce speed limits, but a camera is more flexible regarding the duration it can operate. There is very little evidence available concerning the crash-reduction benefits of this strategy. Most studies use surrogates, such as speed reduction, to measure the effectiveness of automated speed enforcement. Several agencies have shown reductions in crashes after speed enforcement cameras were installed. Paradise Valley, Arizona, experienced a 40-percent decrease in crashes after it began using a camera mounted in a mobile vehicle. In National City, California, a 51-percent decrease in crashes was experienced in the 6-year period following installation of a camera unit in a mobile vehicle in 1991 (Institute of Transportation Engineers, 1999a). However, these individual evaluations may have methodological problems. “Other research by Bloch (1998) questions the effectiveness of automated speed enforcement versus other enforcement strategies (e.g., speed display boards or periodic police patrols). Bloch claims that more than half of the 18 studies evaluating automated enforcement programs have serious methodological problems, thereby negating the validity of their positive results” (Popolizio, 1995). Furthermore, many applications of automated speed enforcement are not directed at approaches to intersections. While results may be similar, there is no sound evidence that this would in fact be the case. In an effort to provide a more definitive answer to the question of effectiveness, a 1993 study in Riverside, California examined “the effect of photo-radar and speed display boards on traffic speed . . . on comparable streets. . . .” The study sought to determine which device is more effective (including more cost-effective) and “whether supplementing speed display boards with police enforcement makes them more effective.” Bloch reported the study's results (1998), the primary conclusion of which was “[W]hile both photo-radar and speed display boards can be effective in reducing vehicle speeds, display boards offer better overall results.” Findings showed that photo radar and speed display boards had about the same effectiveness, reducing mean speeds by 5.1 and 5.8 miles per hour (mph), respectively, where baseline speeds averaged 34 to 35 mph in 25-mph zones. All speed control devices produced more noteworthy results on speeds 10 mph or more over the 25-mph speed limit. At the experimental site, the photo radar reduced these excessive speeds by 30.2 percent; the speed display board reduced them by 34.9 percent, and the enforced display board by 31.8 percent. However, these significant speed reduction capabilities were not sustained after the devices were removed. Researchers noted one long-term, statistically significant effect with the unenforced display board: a 1.7-mph decrease in speed continued at the experimental site after the display board was gone. The study also analyzed the cost-effectiveness in three areas of the three speed controls. Cost per deployment represented an overall estimate for a speed control program, while cost per mph of speed reduction determined whether a device had been cost-effective in achieving speed reductions. Cost per driver exposed assessed “the cost of exposing an individual driver to a speed management device.” Exhibit V-29A illustrates the overall cost estimates for the three areas. As indicated, the unenforced speed display board was the most cost-effective device on both an hourly and daily basis, and photo radar was the least cost-effective of the three speed control devices (TransSafety, 1998). EXHIBIT V-29A Recommended Accident Modification Factors for Installation of Left-Turn Lanes on the Major-Road Approaches to Urban Intersections Type of Speed Control Cost-Effectiveness Measure Cost per MPH of Speed Reduction Per Deployment Per Hour of Deployment Per 12-Hour Day Per Driver Exposed Photo radar (police costs only) $155.00 $8.42 $119.23 $0.39 Photo radar (police and equipment) $220.36 $11.98 $169.51 $0.55 Unenforced speed display board $10.29 $0.20 $2.39 $0.01 Enforced speed display board $91.79 $1.27 $16.39 $0.08
Keys to Success	A key to the success of this strategy is planning the enforcement and prioritizing the intersections that need it (Transportation Research Board, 1998). Such intersections should have a combination of high-speed violation rates and related crash patterns. In some cases, public input or observations by law enforcement personnel may suggest that a location should be targeted with enforcement. It is important that both the highway agency and the law enforcement agency(ies) in the jurisdiction be involved jointly in planning and operating the program. Another critical key to the success of an automated enforcement program is public awareness and acceptance. Acceptance by local law enforcement is another critical element necessary for the success of a program.
Potential Difficulties	There are many opponents to speed enforcement cameras. Arguments against this strategy include violation of personal privacy, violation of constitutional rights, lower effectiveness than other types of enforcement, high cost outweighing the benefits, accuracy of the devices and the settings, and implementation solely to generate revenue. For an example of organized opposition, see http://www.sense.bc.ca/. In addition, this technology has spawned an industry focused upon defeating it. For example, see http://photo-radar.net/ and http://www.phantomplate.com.
Appropriate Measures and Data	A key process measure is the number of intersection approaches on which automated speed enforcement is applied. A more detailed measure is the number of citations issued from the program, as well as the number of traffic convictions resulting. Crash frequency and severity data by type are key safety effectiveness measures. Data describing these crashes and data on the frequency of violations are needed for periods both before and after installation of the cameras. Traffic volume data are needed to represent exposure. Where feasible, the effect of automated speed enforcement on total crashes and crash types potentially related to speed violations should be evaluated separately. Surrogate safety impact measures include mean speed, 85th-percentile speed, and percentage of drivers exceeding the speed limit by specific amounts.
Associated Needs	PI&E is needed to make automated enforcement successful. Public opinion and acceptance can “make or break” an automated enforcement program. Information and awareness efforts and materials typically include the following information: (1) documentation of the problem (in nontechnical terms), (2) objectives of the automated enforcement program, (3) advantages of automated enforcement or conventional enforcement, (4) general locations or areas of automated enforcement systems, (5) uses of revenue generated by automated enforcement, and (6) information on what to do when a citation is received in the mail. As one approach, the public is being informed about the presence of automated technology by placing signs on each approach to an intersection. Having members of local law enforcement speak on television shows, radio shows, or panels at local meetings has been helpful to some agencies installing cameras. The City of Tempe, Arizona, publishes a schedule of locations at which photo enforcement will be occurring (see http://www.tempe.gov/police/Public%20Information%20Office/photo_radar_shedule.htm). The City of Calgary, Canada, also lists locations and information at: http://www.gov.calgary.ab.ca/police/news/photoradarf.html.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agency crash analysis and field reconnaissance procedures should include methods to identify the need for automated speed enforcement. It is important that the program be handled in a coordinated manner by the highway, law enforcement, and judicial agencies. Nearly every highway and police agency has intersections under its jurisdiction where this strategy may be applied. Any speed control program should be based upon wellestablished policies and procedures regarding the setting of speed limits. Speed limits should reflect sound principles and application of current scientific knowledge on what speeds are considered safe, as well as protect against demands based solely on political considerations.
Issues Affecting Implementation Time	The time to implement speed enforcement cameras can vary somewhat, depending upon the extent of public involvement, the need to purchase new equipment, and whether new legislation is needed.
Costs Involved	Costs may vary depending upon the effort put into public information and need for additional legislation. Equipment costs can vary somewhat due to the type of camera selected (i.e., 35-mm, video, or digital), collections and records maintenance, and maintenance of equipment. Funding may be available at the national level through NHTSA. Information on grants obtained from NHTSA under TEA 21 may be found at http://www.nhtsa.dot.gov/nhtsa/whatsup/tea21/tea21programs/402Guide.html. While any cost data will soon become outdated, Exhibit V-29A quotes costs in a costproductivity context.
Training and Other Personnel Needs	Training for highway engineers, safety analysts, and police officers should address automated speed enforcement.
Legislative Needs	Legislation may be necessary before implementing an automated enforcement program if such legislation has not already been enacted. The legislation is necessary to meet constitutional standards, state legal standards, state vehicle code standards, and local jurisdiction standards. A state's enabling legislation should address the broad constitutional issues (federal and state) within a framework that includes elements such as definitions of acceptable automated enforcement devices, any restrictive uses, description of acceptable photographic evidence, and penalty provisions. Local legislation should cover requirements that address local needs for automated enforcement programs in much more detail and should include issues such as operating criteria, the agency that is responsible for camera operation, restrictive uses to that jurisdiction, and requirements for advance notification.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at intersections.
Other Key Attributes to a Particular Strategy	None identified.

Information on Current Knowledge Regarding Agencies or Organizations That Are Implementing This Strategy

Portland and Beaverton, Oregon have jointly published an excellent report on the Internet regarding their experience with photo radar. Although this was not focused upon intersection approaches, it should provide insights that are equally applicable thereto. http://www.portlandonline.com/police/index.cfm?&a=32388&c=29870.

A description of the program in Boulder, Colorado, may be found at: http://www.ci.boulder.co.us/publicworks/depts/transportation/safety/photoradar.html.

Refer to Appendix 4 for information on several European countries' experiences with automatic speed enforcement.

Strategy 17.2 E5: Control Speed on Approaches (E)

General Description

Since speed contributes to crash severity, lowering speeds on approaches to intersections can help reduce the severity of crashes. Slowing vehicle speeds on intersection approaches can improve safety for motorists, pedestrians, and bicyclists. Various techniques for attempting to control speeds on approaches involve geometric design, signal control technology, and other traffic calming treatments.

While warning signs or reduction of speed limits on an intersection approach cannot be expected to be extremely effective in lowering speeds, redesign of the approach can be more effective. Construction of a horizontal curve with an appropriate design speed could accomplish speed reduction. However, the curve should be designed so as not to create problems related to violations of driver expectancy or limited sight distance to the intersection.

Some jurisdictions are using signal control technology to change the signal indication to red when a vehicle is detected traveling at a speed significantly over the speed limit on the approach to the intersection. These systems can be accompanied by a sign warning drivers that the technology is in use. Speeding vehicle activated traffic signals have been deployed in the northern Virginia suburbs of Washington, D.C. Additional information can be found on the US DOT's Intelligent Transportation System Joint Program Office (ITS JPO) Web site. (http://www.its.dot.gov/inform/p79.htm). This technology is also used by several European countries. Refer to Appendix 4 on this and other strategies used in Europe to improve intersection safety.

A raised intersection is another example of a design that could be implemented to slow vehicles. Traffic calming is not intended to be used in place of a signal that meets warrants but can be used as a method of addressing crash severity if designed to slow vehicle speeds.

Roadway treatments such as chicanes, speed tables, and reduced lane widths through widening sidewalks or landscaped areas can be used to slow speeds on roadway approaches to intersections. These are discussed in more detail in the pedestrian guide.

Traffic calming strategies are typically intended to reduce vehicle speeds or traffic volumes on collector and local streets. A main benefit of traffic calming is the potential improvement in pedestrian safety. The history of traffic calming is one centered upon neighborhood traffic management rather than collector and arterial streets. Care must be taken not to extend these methods beyond their range of appropriate application.

The Institute of Transportation Engineers has assembled information on traffic calming on its Web site, which is also sponsored by FHWA: http://www.ite.org/traffic/index.html. The ITE site includes links to Web sites for organizations that are implementing traffic calming strategies. Traffic calming is discussed in the guide for crashes at unsignalized intersections and in even more detail in the guide for crashes involving pedestrians.

Objective 17.2 F—Improve Access Management near Signalized Intersections

Effective access management is a key to improving safety at, and adjacent to, intersections. The number of access points, coupled with the speed differential between vehicles traveling along the roadway and vehicles using driveways, contributes to rear-end crashes. The AASHTO Policy on Geometric Design states that driveways should not be located within the functional area of an intersection. The ITE Traffic Engineering Handbook suggests that the functional area include storage lengths for turning movements and space to maneuver into turn lanes, and consideration should be given to locating driveways, so as to provide enough space to store queues ahead of or behind driveways.

Closing or relocating driveways will reduce turning movements near intersections. Prohibiting turn movements is another strategy to address access management at intersections.

Strategy 17.2 F1: Restrict Access to Properties Using Driveway Closures or Turn Restrictions (T)

General Description

Restricting access to commercial properties near intersections by closing driveways on major streets, moving them to cross streets, or restricting turns into and out of driveways will help reduce conflicts between through and turning traffic. Such conflicts can lead to rear-end and angle crashes related to vehicles turning into and out of driveways and speed changes near the intersection and the driveway(s).

Locations of driveways on both the cross street and major street should be determined based on the probability that a queue at the signal will block the driveway. Directing vehicles to exits on signalized cross streets will help eliminate or restrict the access to the main roadway. Restricting turns to rights-in and rights-out only will address conflicts involving vehicles turning left from the road and left from the driveway.

Restricting access to properties is discussed in greater detail in the guide for crashes occurring at unsignalized intersections.

Strategy 17.2 F2: Restrict Cross-Median Access near Intersections (T)

General Description

When a median opening on a high-volume street is near a signalized intersection, it may be appropriate to restrict cross-median access for adjacent driveways. For example, left and U-turns can be prohibited from the through traffic stream, and left turns from adjacent driveways can be eliminated. Restrictions can be implemented by signing, by redesign of driveway channelization, or by closing the median access point via raised channelization. When access patterns are changed or restricted, the movements restricted in that location should be accommodated at a safe location nearby.

EXHIBIT V-30
Strategy Attributes for Restricting Cross-Median Access near Intersections (T)

Attribute	Description
Technical Attributes
Target	The target of this strategy is crashes involving drivers making turns across medians on approaches to signalized intersections. Angle crashes between vehicles turning through the median and opposing vehicles, as well as rear-end crashes involving vehicles waiting to turn and following vehicles, are crashes related to the crossmedian movement. Sideswipe crashes may occur when a following vehicle on the major road attempts to pass a vehicle waiting to turn left through the median.
Expected Effectiveness	Restricting cross-median access is expected to eliminate conflicts related to vehicles using the median opening, as well as related rear-end and angle crashes.
Keys to Success	Provision of alternative locations for turning maneuvers is a key to the successful restriction of access at a median opening. Care should be taken to prevent the safety problems related to the median opening from being transferred to another location. It is also important for land owners and affected persons to be involved early in the planning process. The quadrants of many signalized intersections are developed with commercial land uses that rely on pass-by traffic. Demonstrating a linkage to the safety of their customers as well as the operational efficiency of the street serving their business can be a key to overcoming resistance to this strategy. The most successful access management techniques rely on physical barriers to restrict movements. Reliance on only signing and pavement markings requires strong enforcement to be effective, which in many cases will not be feasible.
Potential Difficulties	Restricting access at one location will cause turning movements to shift to another location. Care should be taken to ensure adequate capacity and access are provided to accommodate this and that the diversion to alternative access points will not create a safety problem. Adjacent land owners, particularly commercial businesses, are generally opposed to closing and restricting access that they believe will adversely affect their businesses.
Appropriate Measures and Data	Key process measures include the number of intersection approaches for which median access restriction is implemented and the number of potential or actual conflicts eliminated by improvements. Crash frequency and severity by type of crash are key safety effectiveness measures. It is especially useful to identify crashes related to the median access and analyze them separately. A surrogate safety measure is the actual frequency of conflicts occurring at the target locations. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to the targeted median access points at the intersection should be analyzed separately. Traffic volume data are needed to represent exposure. It is especially desirable to obtain data on the volume of vehicles using the median opening and the conflicting volumes.
Associated Needs	There is a definite need to inform the public, especially adjacent property owners, about the safety benefits of access management techniques, as well as methods available to overcome potentially adverse effects of restricting access. In particular, relating the benefits to the specific location is generally required. Thus, accessible, quality data describing the actual safety performance of the location in question is a strong need.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	The optimal situation is to avoid driveway conflicts before they develop. This requires coordination with local land use planners and zoning boards in establishing safe development policies and procedures. Avoidance of high-volume driveways near congested, or otherwise critical, intersections is desirable. Driveway permit staff within highway agencies also needs to have an understanding of the safety consequences of driveway requests. Any highway agency can participate in implementing this strategy. While this strategy is applicable to both rural and urban locations, the greatest need is for agencies that operate extensive systems of urban and suburban arterials. Highway agencies should establish formal policies concerning driveways located near intersections to guide the planning and permitting process and to provide a basis for remedial treatments at existing locations where driveway-related safety problems occur. Local units of government, working as partners with local highway and transportation agencies, should commit to development and implementation of access management guidelines governing land use and site access near signalized intersections for newly constructed facilities. Avoiding safety problems and conflicts with landowners is the preferred approach.
Issues Affecting Implementation Time	plementation of driveway closures and relocations can require 3 months to 3 years. While an extensive project development process usually is not required, discussions with affected property owners must be carried out to reach agreement on access provisions. Essential aspects of such an agreement may include driveway permits, easements, and driveway sharing agreements. Where agreement cannot be reached, the highway agency may choose to initiate legal proceedings to modify access rights. Contested solutions are undesirable and require considerable time to resolve.
Costs Involved	Costs of closing median access points are low, but the cost of providing access in other locations can vary. The materials and labor needed to install signing or additional median curbs or barriers may be low, but relocation of driveways could increase costs.
Training and Other Personnel Needs	Training for highway agency personnel in access management techniques is important to help ensure that the strategies are properly implemented.
Legislative Needs	The power of a highway agency to modify access provisions is derived from legislation that varies in its provision from state to state. Highway agencies generally do not have the power to deny access to any particular parcel of land, but many do have the power to require, with adequate justification, relocation of access points. Where highway agency powers are not adequate to deal with driveways close to intersections, further legislation may be needed.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at signalized intersections.
Other Key Attributes to a Particular Strategy	Since the safety effectiveness of this strategy has not been adequately quantified, it would be desirable to conduct formal evaluations of any projects that are implemented.

A number of major efforts have produced useful guidance documents on access management. The Transportation Research Board Committee on Access Management (ADA70) recently completed and published an Access Management Manual. (See http://www.accessmanagement.gov/manual.html.) The Florida DOT has developed an Access Management CD-Library which can be obtained through http://www.dot.state.fl.us/planning/systems/sm/accman.

Objective 17.2 G—Improve Safety through Other Infrastructure Treatments

Safety problems at signalized intersections may not be specifically related to traffic control, geometry, enforcement, or driver awareness of the intersection. This section provides information on strategies for special intersection conditions that were not covered in the objectives above.

Strategy 17.2 G1: Improve Drainage in Intersection and on Approaches (T)

General Description

One of the most important principles of good highway design is drainage. Drainage problems on approaches to and within intersections can contribute to crashes just as they can on roadway sections between intersections. However, within an intersection, the potential for vehicles on cross streets being involved in crashes contributes to the likelihood for severe crashes, specifically angle crashes. It is necessary to intercept concentrated storm water at all intersection locations before it reaches the highway and to remove over-the-curb flow and surface water without interrupting traffic flow or causing a problem for vehicle occupants, pedestrians, or bicyclists.

Where greater volumes of truck traffic cause rutting in asphalt pavement, especially in the summer when the pavement is hot, consideration should be given to replacing the asphalt with a concrete pavement. Though this is more expensive than a flexible pavement, less rutting will occur, and repair of pavement damage due to trucks will be needed less frequently.

EXHIBIT V-31
Strategy Attributes for Improving Drainage in Intersection and on Approaches (T)

Attribute	Description
Technical Attributes
Target	The target for this strategy is crashes at signalized intersections that are related to poor drainage. Such crashes involve vehicles that hydroplane and hence are not able to stop when required; these crash types include angle, rear end, and head on. Pedestrians and bicyclists would also be at risk.
Expected Effectiveness	Improved drainage can help improve safety, increase traffic capacity, and increase the load capacity of the pavement. However, no adequate documentation of the effect on crash experience seems to be published. It can be expected that improved drainage would reduce crashes related to hydroplaning.
Keys to Success	A key to success for this strategy is involving hydrologic and hydraulic specialists during the initial phases to ensure that proper considerations are given to drainage aspects. Notification of proposed projects should be communicated to other agencies and the public. Any permits and regulations needed by the project should be identified as soon as possible so there are no delays due to legal processes. The success of this strategy will be significantly aided when provision is made for regular condition surveys of existing structures and hydraulic performance to evaluate the functionality of the improvements.
Potential Difficulties	Problems related to drainage design include (1) lateral encroachments on a channel; (2) disruption of water supplies, irrigation facilities, or storm drainage systems; (3) encroachments into environmentally sensitive areas; and (4) failure to plan for ROW. Pavement cross slopes in intersections should be considered in relation to vehicle speeds. For further information, see Appendix 9. Increased maintenance costs and responsibilities due to change in material costs or drainage systems, regardless of how minor, may present problems in implementing drainage improvements. The responsibilities may include many needs, from mowing grass banks to clearing a channel of debris or ice. A serious potential problem associated with drainage design is the legal implications that may be overlooked or not investigated thoroughly. Overlooking a needed permit or regulation can delay a project for months.
Appropriate Measures and Data	Key process measures include the number of intersections at which drainage improvements have been made and the number of each type of improvement (improving inlet structures, redirecting flow away from pavement). The daily volume of vehicles affected by the change is another process measure to consider. Frequency and severity of crashes related to insufficient drainage should be tracked before and after implementation of the improvements. Traffic volume data are needed to represent exposure.
Associated Needs	There is no need for special PI&E programs. Adjacent property owners could be informed of the safety benefits of proper drainage maintenance procedures that will need to be performed by the agency with jurisdiction in the area.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Nearly all highway agencies can make use of this strategy. This strategy is applicable to rural, urban, and suburban areas. Highway agencies should review their design manuals regarding drainage design to ensure that proper drainage design/techniques are being used on all projects. Policy guidance on drainage design/techniques is discussed in AASHTO's A Policy on Geometric Design of Highways and Streets (2001) and Highway Drainage Guidelines (1993) and other policy manuals. Highway agencies should consider these policies if they are not covered in their own guidelines.
Issues Affecting Implementation Time	Many small projects that could include drainage improvements, such as spot safety improvements, single bridge replacements, and similar work, are often planned and constructed within several months. Longer-term improvements sometimes require as much time to complete as construction of an entirely new section of highway.
Costs Involved	While minor drainage improvements can be low cost, the costs involved in designing and implementing a drainage system is not an incidental or minor task on most roads. Careful attention should be given to adequate drainage and protection of the highway from floods in all phases of location and design; this will be effective in reducing both construction and maintenance costs. Drainage is usually more challenging and costly for urban projects than for rural projects due to more rapid runoff rates and larger volumes of runoff, increases in cost due to potential damage to adjacent property by flooding, greater restrictions because of urban developments, lack of natural areas of water bodies to receive flood water, and higher volumes of traffic or pedestrians.
Training and Other Personnel Needs	Effective drainage techniques should be addressed in highway training concerning design applications/methods of intersections.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	This strategy can be used in conjunction with the other strategies for improving safety at signalized and unsignalized intersections.
Other Key Attributes to a Particular Strategy	None identified.

Strategy 17.2 G2: Provide Skid Resistance in Intersection and on Approaches (T)

Slippery pavement should be addressed to reduce the potential for skidding. The coefficient of friction is most influenced by vehicle speed, vehicle tire condition, and surface condition. Consideration should be given to improving the pavement condition to provide good skid resistance, especially during wet weather. This can be accomplished by

• Providing adequate drainage,

• Grooving existing pavement, and

• Overlaying existing pavement.

Refer to the guide for addressing run-off-road crashes, which contains a discussion of providing skid-resistant pavements.

Strategy 17.2 G3: Coordinate Closely Spaced Signals near at-Grade Railroad Crossings (T)

General Description

At-grade railroad crossings on approaches to intersections have potential safety problems related to vehicle queues forming across the railroad tracks. The railroad and nearby traffic control signals should be coordinated to provide preemption of the traffic signals when trains are approaching the intersection.

EXHIBIT V-32
Strategy Attributes for Coordinating Closely Spaced Signals near at-Grade Railroad Crossings (T)

Attribute	Description
Technical Attributes
Target	This strategy targets crashes related to queues on approaches to signalized intersections in close proximity to at-grade railroad crossings. This situation presents a significant potential for vehicle-train crashes, but vehicle-vehicle crashes could also occur if drivers try to speed through an intersection to avoid waiting in a queue near the railroad. Rear-end and angle crashes between vehicles should be analyzed to determine if they are related to the presence of the railroad crossing.
Expected Effectiveness	Coordination of signals to clear the tracks when a train is approaching should eliminate the potential for vehicles to be trapped on the tracks.
Keys to Success	A key to success is the compatibility of the traffic signal and railroad active warning devices in order to safely control vehicle, train, bicycle, and pedestrian movements. Vehicles must be provided with adequate green time to clear the railroad tracks when a train is approaching. This means that potential queue lengths during congested periods must be considered and train detection systems provided on the railroad tracks far enough upstream of the crossing for the signal preemption to clear all vehicles. A gate is an integral part of the active warning system.
Potential Difficulties	The MUTCD states that warning lights shall flash for at least 20 seconds before a train approaches (for train speeds of 20 mph or more). Train detection may need to occur earlier than when the train is 20 seconds away from the crossing, depending upon the amount of time needed to preempt the nearby signal and clear the tracks (Korve, 1999). The railroad track may be so close to the intersection that a design vehicle cannot fit between the tracks and the intersection if it has to stop for a red signal. A presignal can be used to control traffic approaching the at-grade crossing. Presignals are installed on the near side of an at-grade railroad crossing, upstream of the traffic signal. The presignal turns red as a train approaches; this will occur before the downstream traffic signal turns red in order to allow vehicles to clear the railroad tracks. Care must be taken that a driver with a red presignal does not mistakenly think the green trackclearance signal at the intersection is their signal. A special design of the signal face may be needed to ensure vehicles approaching the track do not misunderstand the signals (see Strategy 17.2 D2). A railroad crossing gate would also contribute to understanding of the presignal, since it would be lowered when the presignal is red. Traffic engineers should communicate with railroad agencies to verify the signal preemption system being designed is compatible with the railroad signal systems. Often there are problems with differences in terminology between various agencies (such as “preemption”), and care should be taken to clarify terminology.
Appropriate Measures and Data	Key process measures include the number of signalized intersections near at-grade railroad crossings for which coordination between the train detection and warning system and the traffic signal is implemented. Crash frequency and severity by type of crash and involvement of trains are key safety effectiveness measures. It is useful to separately analyze crashes that did not involve trains by type and whether they occurred during a preempted signal cycle. Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to signal preemption and vehicles clearing the tracks (mainly rear-end crashes) and to driver unawareness of signals (rearend crashes) should be analyzed. Traffic volume data are needed to represent exposure.
Associated Needs	None identified.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Communication between railroad and highway agencies is an important issue in improving the safety at railroad grade crossings and nearby intersections. The American Association of Railroads has information concerning the type of signal equipment to be used. State and local regulations should also be consulted when determining how the traffic signal will operate.
Issues Affecting Implementation Time	Implementation time can vary, depending on the communication and coordination among railway, highway, and any other agencies that would be involved in improvement of signal control at and near railroad grade crossings.
Costs Involved	Costs involved in improving signal control near at-grade crossings can vary, depending upon the compatibility of existing equipment with the desired treatment. Installation of new equipment that allows coordination of signals will increase costs. Maintenance is another cost element to be considered.
Training and Other Personnel Needs	Coordination of traffic control signals with railroad warning and train detection systems should be addressed in highway agency training concerning intersection operation and railroad grade crossing safety.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Coordination of traffic signals with train detection and warning systems is compatible with most other strategies to improve signalized intersection safety.
Other Key Attributes to a Particular Strategy	A traffic signal preemption system should be designed considering many geometric, traffic flow, and vehicle and train characteristics. The ITE document entitled Preemption of Traffic Signals At or Near Railroad Grade Crossings with Active Warning Devices contains discussion on these items (Institute of Transportation Engineers, 1997). Additional information can be found in NCHRP Synthesis 271: Traffic Signal Operation Near Highway-Grade Crossings, including discussion on traffic signal and train detection systems (Korve, 1999).

Strategy 17.2 G4: Relocate Signal Hardware out of Clear Zone (T)

General Description

Traffic signal hardware represents a potential roadside hazard similar to utility poles, trees, and other large fixed objects. Traffic signal supports and controller cabinets should be located as far from the edge of pavement as is possible, especially on high-speed facilities, as long as this does not adversely affect visibility of the signal indications. Consideration should be given to shielding the signal hardware if it cannot be relocated. Where there is an existing roadside barrier, the cabinet should be located behind the barrier when feasible. If practical, signal supports in medians should be located to provide more than the minimum clearance required by the agency. The signal hardware should not obstruct sight lines.

Post-mounted signals in the median of a road are often deemed appropriate to reinforce the information presented by the overhead signal heads at the intersection, especially at left-turn lanes, but they are a hazard in that location. However, their benefit may outweigh the disadvantage of the location of the post in the median.

EXHIBIT V-33
Strategy Attributes for Relocating Signal Hardware Outside of Clear Zone (T)

Attribute	Description
Technical Attributes
Target	This strategy is targeted at crashes with signal hardware at signalized intersections, especially on high-speed roadways, where signal hardware is located within the clear zone or is a sight obstruction. Single-vehicle run-off-road crashes involving the signal hardware, as well as angle crashes related to insufficient sight distance, could occur when signal hardware is in an improper location.
Expected Effectiveness	Relocating the signal hardware outside the clear zone should reduce the likelihood of vehicles striking the hazard. The effectiveness of this strategy is difficult to estimate given the range of conditions and relative infrequency of such conflicts at any one location.
Keys to Success	The new location of the signal hardware should not present a greater safety hazard than the previous location by creating a sight distance obstruction.
Potential Difficulties	Care should be taken to ensure signal hardware is not relocated to a position where it obstructs sight distance or presents a safety hazard to pedestrians or bicyclists. The Americans with Disabilities Act should be consulted to ensure compliance.
Appropriate Measures and Data	A process measure is the number of intersection approaches for which signal hardware is relocated. Frequency and severity of crashes involving signal hardware are key safety effectiveness measures. Traffic volume data are needed to represent exposure. These data should be collected before and after installation of the system for comparison purposes. Traffic volume data are also needed to establish levels of exposure.
Associated Needs	None identified.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding clear zone and location of signal hardware to ensure appropriate action is being taken on routine projects. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas.
Issues Affecting Implementation Time	Implementation time will be relatively short if additional ROW is not needed in order to move the hardware outside of the clear zone. Acquisition of ROW will increase implementation time.
Costs Involved	Costs will be moderate if acquisition of ROW is not required to move the hardware outside of the clear zone. Acquisition of ROW will increase costs.
Training and Other Personnel Needs	Clear zone issues should be addressed in highway agency training concerning traffic signal installation and roadside design.
Legislative Needs	None identified.
Other Key Attributes
Compatibility of Different Strategies	Relocation of signal hardware is compatible with most other strategies to improve safety at signalized intersections.
Other Key Attributes to a Particular Strategy	None Identified.

Strategy 17.2 G5: Restrict or Eliminate Parking on Intersection Approaches (P)

General Description

Parking adjacent to turning and/or through lanes on intersection approaches may create a hazard. It can cause a frictional effect on the through traffic stream, can often block the sight triangle of stopped vehicles, and may occasionally cause the blocking of traffic lanes as vehicles move into and out of parking spaces. Restricting and/or eliminating parking on intersection approaches can reduce the workload imposed on the driver and limit additional collision opportunities. Parking restrictions can be implemented through signing, pavement markings, or restrictive channelization. Restrictions can be implemented for specific times of day or specific vehicle types. Enforcement of parking restrictions, accompanied by public information, including towing offending vehicles, is a necessary component to this strategy.

EXHIBIT V-34
Strategy Attributes for Restricting or Eliminating Parking on Approaches (P)

Attribute	Description
Technical Attributes
Target	This strategy targets crashes related to parking on intersection approaches. The parking, though currently permitted, may present a safety hazard by blocking sight distance (and contributing to angle crashes) or due to parking maneuvers (contributing to rear-end and sideswipe crashes). On-street parking can decrease pedestrian safety if parked vehicles block drivers' and pedestrians' views of each other. Curb extension can be constructed where pedestrians cross streets, and parking should not be permitted on approaches to crosswalks. Further information on this aspect of the problem is covered in the pedestrian crash guide.
Expected Effectiveness	The ITE Traffic Engineering Handbook (Pline, 1999) states that based upon a review of crash data, 20 percent of nonfreeway crashes in cities are in one way or another related to parking. Midblock crash rates on major streets with parking stalls that are used about 1.6 million hours per year per kilometer could be expected to decrease up to 75 percent after parking is prohibited. An Australian Bureau of Transport Economics study (available at http://www.dotars.gov.au/transprog/downloads/road_bs_matrix.pdf) of a black spot treatment program showed that banning parking adjacent to an intersection resulted in an average decrease in crashes of 10 percent on the approach that formerly had parking (Appendix C of Bureau of Transport Economics, 2001). These decreases were seen in the rear-end-, lane-change-, pedestrian-, and parking-maneuver-related crash types.
Keys to Success	Parking regulation signs need to be posted conspicuously. Consistent and rigorous enforcement of these regulations is necessary as well. Working with adjacent land owners to communicate the reasons for prohibiting parking is also essential to achieving success.
Potential Difficulties	The Uniform Vehicle Code does not require use of No Parking signs in some circumstances. Drivers are often not aware of some of the locations where parking is prohibited, however, and signs should be used to convey this information to drivers. Adjacent land owners, particularly commercial businesses, may be opposed to the removal of on-street parking. Removal of parking requires a commitment to enforcement through ticketing and towing where needed.
Appropriate Measures and Data	Key process measures include the number of signalized intersections for which parking has been prohibited on the approaches, the number of approaches on which parking has been restricted, the number of parking spaces eliminated by restrictions, and the percent of problem parking spaces eliminated by restrictions. Crash frequency and severity by type of crash are also key safety effectiveness measures. It is especially useful to separately analyze crashes that are related, directly or indirectly, to on-street parking on the approach (for example, crashes involving vehicles making parking maneuvers, or limited sight distance due to parked vehicles). Crash frequency and severity data are needed to evaluate such improvements. If feasible, both total crashes and crashes related to on-street parking should be analyzed. Traffic volume data are needed to represent exposure.
Associated Needs	Public involvement activities may be required in order to gain understanding and acceptance of the proposed changes in parking regulations.
Organizational and Institutional Attributes
Organizational, Institutional, and Policy Issues	Highway agencies should review their traffic engineering and design policies regarding on-street parking to ensure appropriate action is being taken on projects. All stakeholders should be involved from the earliest stages of planning, including owners of adjacent properties and representatives of legislative bodies for the jurisdictions involved. Nearly any highway agency can participate in implementing this strategy, which is applicable to rural, urban, and suburban areas.
Issues Affecting Implementation Time	Time to implement parking restrictions is low if no new ordinances are required. Implementation may, however, require passing of ordinances by city councils.
Costs Involved	Costs to implement parking restrictions with signing are low. If enforcement is used to help implement the restrictions, costs will be increased.
Training and Other Personnel Needs	Safety issues related to parking should be addressed in highway agency training concerning intersection design and operation.
Legislative Needs	Approval of appropriate legislative body (mayor, town council, etc.) may be required before no-parking zones can be created.
Other Key Attributes
Compatibility of Different Strategies	Restriction of parking is compatible with most other strategies for improving signalized intersection safety.
Other Key Attributes to a Particular Strategy	On-street parking has a detrimental effect on capacity of the roadway. Improved flow of vehicles to and through the intersection may be enough to warrant parking prohibition as well.

Key References

American Association of State Highway and Transportation Officials. A Policy on Geometric Design of Highways and Streets. Washington, D.C. 2001.

American Association of State Highway and Transportation Officials. Guide for the Development of Bicycle Facilities. Washington, D.C. 1999.

American Association of State Highway and Transportation Officials. Highway Drainage Guidelines. Vols. 1–11. Washington, D.C. 1993.

American Association of State Highway and Transportation Officials. Roadside Design Guide. Washington, D.C. 1996. Bauer, K. M., and D. W. Harwood. Statistical Models of At-Grade Intersection Accidents. Report No. FHWA-RD-96-125. Federal Highway Administration. Washington, DC. November 1996.

Bloch, Steven A. “A Comparative Study of the Speed Reduction Effects of Photo-Radar and Speed Display Boards.” Paper presented at the Transportation Research Board Annual Meeting. January, 1998. Washington, DC.

Bonneson, J., and M. Fontaine. Evaluating Intersection Improvements: An Engineering Study Guide. NCHRP Report 457. TRB, National Research Council, Washington, DC. 2001. Available online at http://gulliver.trb.org/publications/nchrp/esg/esg.pdf (last accessed May 20, 2004).

Box, Paul C., and Paul E. Basha. “A Study of Accidents with Lead versus Lag Left-Turn Phasing.” ITE Journal. May 2003.

Bureau of Transport Economics. The Black Spot Program 1996–2002: An Evaluation of the First Three Years. Canberra, Australia. 2001. Available online at http://www.bte.gov.au/docs/r104/htm/contents.htm (last accessed May 20, 2004).

Collura, J., J. Chang, E. Willhaus, and J. Gifford. “Traffic Signal Preemption and Priority: Technologies, Past Deployments, and System Requirements.” Conference Proceedings. Intelligent Transportation Society of America (ITS America) Annual Meeting. June, 2001.

Compton, R. P., and E. V. Milton. Safety Impact of Permitting Right-Turn-On-Red. Publication No. HS-808 200. National Highway Traffic Safety Administration. Washington, D.C. December 1994.

Dixon, K; J. Hibbard; and H. Nyman. “Right-Turn Treatment for Signalized Intersections,” Urban Street Symposium, Dallas, Texas, June 28–30 1999. Conference sponsored by the Transportation Research Board. Available online at http://www.mackblackwell.org/research/finals/arc9012/righturn.pdf (last accessed May 20, 2004).

Federal Highway Administration. Guidance for Using Red Light Cameras. Washington, D.C. 2003. Available online at http://safety.fhwa.dot.gov/rlcguide/index.htm (last accessed May 20, 2004).

Federal Highway Administration. Manual on Uniform Traffic Control Devices for Streets and Highways. FHWA. Washington, D.C. 2003.

Federal Highway Administration. MUTCD 2000: Manual on Uniform Traffic Control Devices for Streets and Highways: Millennium Edition. Washington, D.C. 2000.

Federal Highway Administration. Safety Applications of ITS in Rural Areas. Washington, D.C. 2002. Available online at http://www.itsdocs.fhwa.dot.gov//JPODOCS/REPTS_TE//13609.html (last accessed May 20, 2004).

Federal Highway Administration. Safety Effectiveness of Intersection Left-and Right-Turn Lanes, Tech Brief, Report No. FHWA-RD-02-089, McLean, Virginia, 2002.

Federal Highway Administration. Signalized Intersections: Informational Guide. In press.

Fitzpatrick, K., K. Balke, D. Harwood, and I. Anderson. Accident Mitigation Guide for Congested Rural Two-Lane Highways. NCHRP Report 440. TRB, National Research Council, Washington, DC. 2000.

Fleck, J., and B. Lee. “Safety Evaluation of Right Turn on Red.” ITE Journal. Volume 72. Issue 6. 2002. Pp. 46–48.

Gallagher, V. P. New Directions in Roadway Lighting. Report No. FHWA-TS-80-223. Illuminating Engineering Society. FHWA, Washington, DC. 1980.

Gluck, J., H. S. Levinson, and V. Stover. Impacts of Access Management Techniques. NCHRP Report 420. TRB, National Research Council, Washington, DC. 1999.

Hanna, J. T., T. E. Flynn, and W. L. Tyler, “Characteristics of Intersection Accidents in Rural Municipalities,” Transportation Research Record 601, TRB, National Research Council, Washington, DC. 1976.

Harwood, D. W. Use of Rumble Strips to Enhance Safety. NCHRP Synthesis of Highway Practice 191. TRB, National Research Council, Washington, D.C. 1993.

Harwood, D. W., F. M. Council, E. Hauer, W. E. Hughes, and A. Vogt. Prediction of the Expected Safety Performance of Rural Two-Lane Highways. Report No. FHWA-RD-99-207. Federal Highway Administration. 1999.

Harwood, D. W., et al. Safety Effectiveness of Intersection Left- and Right-Turn Lanes, FHWA-RD-02-089, July 2002. Available online at http://www.tfhrc.gov/safety/pubs/02089/index.htm (last accessed May 20, 2004).

Harwood, D. W., J. M. Mason, R. E. Brydia, M. T. Pietrucha, and G. L. Gittings. Intersection Sight Distance. NCHRP Report 383. TRB, National Research Council, Washington, DC. 1996.

Harwood, D. W., M. T. Pietrucha, M. D. Wooldridge, R. E. Brydia, and K. Fitzpatrick. Median Intersection Design. NCHRP Report 375. TRB, National Research Council, Washington, DC. 1995.

Highway Capacity Manual. TRB, National Research Council, Washington, DC. 2000. Institute of Transportation Engineers. Automated Enforcement in Transportation. Washington, D.C. December 1999a.

Institute of Transportation Engineers. Determining Vehicle Signal Change and Clearance Intervals. Washington, D.C. 1994.

Institute of Transportation Engineers. Guidelines for Prohibition of Turns on Red. Washington D.C. 1986.

Institute of Transportation Engineers. Preemption of Traffic Signals At or Near Railroad Grade Crossings with Active Warning Devices. Washington, D.C. 1997.

Institute of Transportation Engineers. Traffic Safety Toolbox: A Primer on Traffic Safety. Washington, D.C. 1999b.

Jack E. Leisch and Associates. Planning and Design Guide: At-Grade Intersections. 1990.

Kallberg, V-P, and Ranta, S. “Impacts of Urban Speed-Reducing Measures.” 2nd International Symposium on Highway Geometric Design. 2000. Pp. 93–109.

Kay, J. L., J. C. Allen, and J. M. Bruggerman. Evaluation of First-Generation UTCS/BPS Control Strategy. FHWA-RD-75-26 through 28. FHWA, Washington, D.C. 1975.

Knapp, K. K. “Traffic-Calming Basics.” Civil Engineering. Volume 70. Issue 1. 2000. Pp. 46–49.

Korve, H. Traffic Signal Operation near Highway-Grade Crossings. NCHRP Synthesis 271. TRB, National Research Council, Washington, D.C. 1999.

Koupi, P. A., and A. M. Kothari. Recommended Guidelines for Protected/Permissive Left-Turn Phasing. Conference Proceedings, Institute of Transportation Engineers Annual Meeting. 1999.

Kuciemba, S. R., and J. A. Cirillo. Safety Effectiveness of Highway Design Features, Volume V: Intersections. Report No. FHWA-RD-91-048. Federal Highway Administration. November 1992.

Larsen, J. H. “Develop Your Own In-House Public Relations Program.” ITE Journal. Volume 61. No. 1. January 1991.

Lin, F. B. “Right-Turn-on-Red Characteristics and Use of Auxiliary Right-Turn Lanes.” Transportation Research Record 1010. TRB, National Research Council, Washington, DC. 1985.

Lunenfeld, H. “Human Factors Associated with Interchange Design Features,” Transportation Research Record 1385, TRB, National Research Council, Washington, DC. 1993.

McCoy, P., P. Byrd, G. Pesti, and V. Singh. “Improving Sight Distance from Opposing Left-Turn Lanes.” Institute of Transportation Engineers Annual Meeting. 1999.

McGee, H. Making Intersections Safer: A Toolbox of Engineering Countermeasures to Reduce Red- Light Running. Federal Highway Administration and Institute of Transportation Engineers. 2003. Available online at http://www.ite.org/library/redlight/MakingInt_Safer.pdf (last accessed May 20, 2004).

Migletz, J., Fish, J. K., and J. K. Graham. Roadway Delineation Practices Handbook. Report No. FHWA-SA-93-001. Federal Highway Administration. Washington, D.C. August 1994.

Neuman, T. R. Intersection Channelization Design Guide. NCHRP Report 279. TRB, National Research Council, Washington, DC. 1985.

Pline, J. L. Left-Turn Treatments at Intersections. NCHRP Synthesis 225. TRB, National Research Council, Washington, DC. 1996.

Pline, J. L. Traffic Engineering Handbook. Institute of Transportation Engineers, Washington, DC. 1999.

Popolizio, Rudolph E. “New York City’s Red Light Camera Demonstration.” 1995 Compendium of Technical Papers. Institute of Transportation Engineers, Washington, DC. 1995.

Rakha, H. A. Medina, H. Sin, F. Dion, M. Van Aerde, and J. Jenq. “Traffic Signal Coordination across Jurisdictional Boundaries: Field Evaluation of Efficiency, Energy, Environmental, and Safety Impacts.” Transportation Research Record 1727. TRB, National Research Council, Washington, DC. 2000.

Retting, R., J. Chapline, and A. Williams. Changes in Crash Risk Following Retiming of Signal Change Intervals. Insurance Institute for Highway Safety, Arlington, VA. September 2000.

Retting, R., and M. Greene. “Influence of Traffic Signal Timing on Red-Light Running and Potential Vehicle Conflicts at Urban Intersections.” Transportation Research Record 1595. TRB, National Research Council, Washington, DC. 1997.

Retting, R., M. Nitzburg, C. Farmer, and R. Knoblauch. “Field Evaluation of Two Methods for Restricting Right Turns on Red to Promote Pedestrian Safety.” ITE Journal. Volume 72. Issue 1. 2002. Pp. 32–36.

Retting, R. A., A. F. Williams, C. M. Farmer, and A. F. Feldman. “Evaluation of Red Light Camera Enforcement in Fairfax, Va., USA.” ITE Journal. August 1999a.

Retting, R. A., A. F. Williams, C. M. Farmer, and A. F. Feldman. “Evaluation of Red Light Camera Enforcement in Oxnard, California.” Accident Analysis and Prevention. 1999b.

Robinson, B. W., et al. Roundabouts: An Informational Guide. Report No. FHWA-RD-00-067. Federal Highway Administration. June 2000. Available online at http://www.tfhrc.gov/safety/00068.pdf (last accessed May 20, 2004).

Shebeeb, O. “Safety and Efficiency for Exclusive Left-Turn Lanes at Signalized Intersections.” ITE Journal. July 1995.

Stamatiadis, N., K. R. Agent, and A. Bizakis. “Guidelines for Left-Turn Phasing Treatment.” Transportation Research Record 1605. TRB, National Research Council, Washington, DC. 1997.

Staplin, L., K. Gish, L. Decina, K. Lococo, D. Harkey, M. Tarweneh, R. Lyles, D. Mace, and P. Garvey. Synthesis of Human Factors Research on Older Drivers and Highway Safety: Volume 2. Federal Highway Administration, Report No. FHWA-RD-97-095. Washington, D.C. October 1997.

Staplin, L., K. Lococo, and S. Byington. Older Driver Highway Design Handbook. Report No. FHWA-RD-97-135. Federal Highway Administration. Washington, D.C. January 1998. Traffic Safety Facts 2002, National Highway Traffic Safety Administration, Available online at http://www-nrd.nhtsa.dot.gov/pdf/nrd-30/NCSA/TSFAnn/TSF2002Final.pdf (last accessed May 20, 2004).

Transportation Research Board. Access Management Manual. TRB, National Research Council, Washington, DC. 2003.

Transportation Research Board. Special Report 254, Managing Speed: Review of Current Practice for Setting and Enforcing Speed Limits. TRB, National Research Council, Washington, DC. 1998.

TranSafety, Inc., “Study Reports on the Effectiveness of Photo-Radar and Speed Display Boards.” USroads Web site. 1998. Available online at http://www.usroads.com/journals/p/rilj/9805/ri980504.htm (last accessed May 20, 2004).

Twomey, J. M., Heckman, M. L., Hayward, J. C., and R. Zuk. “Accidents and Safety Associated with Interchanges.” Transportation Research Record 1385. TRB, National Research Council, Washington, DC. 1993.

Walker, R. A. “Coordination of Basic Intersection Design Elements: An Overview.” Transportation Research Record 1385. TRB, National Research Council, Washington, DC. 1993. Pp. 51–59. Z

ador, P., H. Stein, S. Shapiro, and P. Tarnoff. “Effect of Signal Timing on Traffic Flow and Crashes at Signalized Intersections.” Transportation Research Record 1010. TRB, National Research Council, Washington, DC. 1985.