Government agencies, mostly at the state, county and local levels, own and operate approximately four million miles of roads and streets in the U.S. Through road ownership and construction, statutes, traffic regulation and enforcement, government agencies create the physical and legal driving environment. Not all drivers respond in the same way to the same driving environment and these differences extend to speed selection. An understanding of speed issues requires some basic insights of agency functions and human behavior in a regulated environment.

Before a new road or street is built, plans are prepared by engineers. Except for minor streets and roads, the design process considers speed by adopting a designated design speed. The road is then constructed in accordance with the design plans and put into service for the motoring public, which is regulated by a set of laws and regulations, including an established speed limit. A more detailed description of these processes and their relationships is provided below.

PLANNING AND DESIGN OF ROADS AND STREETS

Roadway geometry is one of the informal information sources that drivers interpret when selecting a speed. However, roadway geometry can provide deceptive, conflicting and confusing speed cues. This is unfortunate but true and because the existing network is so extensive, substantial modifications are often impractical.

The Green Book is the dominant reference publication for geometric design in the U.S. and its application involves selecting a "design speed." The definition of design speed has changed over time but the way it is used in the design process has not. The Green Book guidance on designating design speed does not address speed limits. The Green Book recommends that topography, anticipated operating speed, adjacent land use and functional classification be considered and as high a design speed as practical be selected, except for local streets. The selected (designated) design speed is used explicitly to determine:

• minimum sight distances
• minimum horizontal curve radii
• maximum grade
• minimum rate of vertical curvature
• minimum width of selected roadway cross section features
• superelevation rate
• maximum gradient
• roadside clearance
• freeway nose taper
• acceleration/deceleration lane length

Chapter 4 provides background for many of these speed-based design criteria. Some of these criteria, such as sight distance and vertical curvature, are related.

The Green Book provides minimum or limiting values for the design criteria. It also recommends "above-minimum design values should be used, where practical." The underlying rationale for this guidance is that above-minimum features will safely accommodate a condition beyond the assumed parameters. Historically, this was thought to be "conservative" and consistent with other engineering disciplines that use factors of safety to intentionally "over design" critical components. The direct effect of utilizing above-minimum design elements is that it affords drivers greater comfort to travel at higher speeds, and thereby leads to an "inferred" design speed greater than the designated design speed that may be inappropriate for surrounding conditions. A definition for "inferred design speed" is included earlier in this publication and will be referenced in the next paragraph. An example is provided on the following page to complement the definition and clarify the term’s meaning.

Frequently, roads and streets designed for a particular speed appear suitable for much higher speeds. Drivers read the road, not the design plans. Some roadway segments, such as a long straight section, look the same regardless of designated design speed. When these features are combined with over-designed speed sensitive features (i.e., from using above-minimum values as recommended by highway geometric design policy) the visible cues on appropriate speed may be in sharp contrast to the designated design speed. What was contemplated by the designer as a factor of safety (with respect to the designated design speed) is often negated by driver speed choice. The differing perceptions of speed by designer and driver are especially problematic where a single minimum-value feature (e.g., minimum radius curve) is located within a segment that generally has an inferred design speed much higher than the designated design speed. This condition is contrary to the design objective of "design consistency." However, the Green Book includes no specific design criteria for design consistency and the condition described technically complies with the geometric design policy.

Highway geometry is not the only cue drivers are receiving. Landscape, development, perceived conflicts, and intersections also contribute to the driver’s selection of speed. Just as selecting a design that is higher than necessary is inappropriate, it is also inappropriate to select a lower design speed if other cues will lead drivers to select a speed that is too high for the design geometry. We do not want the highway geometry to be a surprise to the driver.

Design policy recognizes that a wide range of performance levels exist amongst the large and diverse population of drivers. The need to integrate human performance and engineering design gives rise to another speed conundrum. Green Book criteria attempt to safely accommodate below-average driver proficiencies. Many drivers routinely drive at speeds faster than the speed limit and faster than other drivers. Overconfident drivers may come to believe they possess superior abilities and skills that enable them to operate safely at higher speeds.

Figure 10 illustrates an example relationship between speed, geometric design and differing levels of driver performance. Minimum stopping sight distance design criteria are based on several assumptions, as were previously discussed in chapter 4, including 2.5 seconds of driver perception-reaction time. The solid line in figure 10 is a plot of sight distance versus speed using a 2.5-second reaction time. The sight distance values coincide with Green Book stopping sight distance criteria, which are also listed in table 2. The broken-line plot is based on a reaction time of 1.5 seconds. Drivers do not use formulae to select their speeds; consequently some drivers drive faster because they feel capable of reacting quickly. However, the probability of dealing successfully with unexpected circumstances is reduced with higher speeds.

Figure 10. Sight distances corresponding to different reaction times.

The 85th percentile speed is used extensively in the field of traffic engineering and safety. Since the majority of drivers are considered reasonable and should be accommodated, some numerical definition for this segment of the driver population is needed. Over time, the 85th percentile driver (or speed) has been used to characterize reasonable and prudent behavior.

For a variety a reasons, including those summarized in this chapter, policy-compliant designs often appear adequate for speeds far above the designated design speed. The speed profile in figure 11 is an example of a case where the 85th percentile speed and posted speed exceed the designated design speed. These data were collected along an existing multi-lane highway in central Pennsylvania. The geometric features are in keeping with Green Book guidance recommending higher-than-minimum values for design features (e.g., vertical curve lengths, horizontal curve radii). The inferred design speed is substantially higher than the designated design speed. When the facility was first opened to traffic, the posted speed of 40 mph was "mismatched" with the geometric features. Based on a review of operating speeds, the speed limit was changed to 50 mph. Many researchers have noted that the design speed process has no systematic check for consistency between operating and design speeds. (13,14,15)

Figure 11. Speed profile with inferred design speed substantially exceeding designated design speed.

Even though the Green Book is the principal reference for geometric design, for two basic reasons there are many street and highway segments that do not conform to its current provisions:

• Road design methods have changed substantially over time but when design methods change existing roads are not necessarily updated,
• Some agencies with road and street design responsibility, especially municipal governments, may use different methods.

Many city streets and some of the first freeways were designed prior to development of the design speed concept. Many designs were created without explicit consideration of vehicle speeds. AASHO, which was the predecessor to AASHTO, published its first design-related policies in the 1940s. The original AASHO policy used a design speed but it did not apply to urban streets. Over the years, AASHO, and then AASHTO, periodically updated and expanded its geometric design policy. As the policy became more complete (e.g., included procedures for urban streets) it reached its current status as the principal reference for designing U.S. roads and streets. When geometric design policy is updated, it is often not applied retroactively. In other words, facilities that were designed under an earlier policy are not automatically revised to meet the most current policy.

The use of AASHTO policy is only mandatory in the design of roads that are part of the National Highway System, which includes all Interstate highways and other selected arterial highways. For other roads and streets, state DOTs and local government units (i.e. counties, cities, boroughs, towns), have the discretion to develop and apply the geometric design process and criteria they consider most appropriate. State DOTs use design processes and criteria that are the same or similar to AASHTO’s. Local agency practice varies widely, with larger agencies typically adopting policies similar to their state DOTs, and therefore those of AASHTO. Local governments with limited roadway responsibilities (i.e., limited mileage and no arterials) may use very simple guidelines, such as those incorporated into subdivision regulations. These guidelines typically assume low, but unspecified, speeds. Minimum or typical values are provided for features such as curve radii, grades and cross section features, without reference to a design speed.

The previous discussion focused on the case where inferred design speeds exceed the designated design speed. Cases where the opposite is true can also occur—see Appendix A which provides an example of how to calculate the inferred design speed of horizontal and vertical curvature. In such cases, design speed-related criteria are less than the minimum value associated with a particular designated design speed. For example, a radius-superelevation combination on a horizontal curve may be less than the minimum radius for the designated design speed. Use of geometric design criteria that do not meet the minimum value based on a designated design speed may occur for a variety of reasons. First, the roadway alignment may have been constructed using geometric design criteria that have since been revised. As noted earlier, design policy is typically not applied retroactively and therefore the below-minimum design feature is not re-constructed to comply with the most recent design policy. Traffic control devices (e.g., warning signs with speed advisory plaque) are often placed near below-minimum design features to alert drivers of the need to reduce their operating speed. On planned or reconstructed roadway facilities, particularly those on the National Highway System, use of below-minimum design criteria are often constructed only after granting a design exception. For more information about design exceptions, the reader is referred to NCHRP Synthesis 316 (25).

To summarize, government agencies exercise control over the planning and detailed design of roads and streets, either by performing these functions or approving the work of consultants and third parties, such as developers. The current design process includes consideration of speed. However, the design speed process and AASHTO guidance often lead to inferred design speeds far in excess of the designated design speed. Further, many roads and streets were not designed for any particular anticipated speed, either because they predate the design speed method or the jurisdictional agency uses a different approach. The net result is that the geometric features of roads and streets have not been designed to encourage a predetermined operating speed. In fact, many aspects of the geometric design process are suspected of encouraging operating speeds higher than the design speed and higher than desirable.

ESTABLISHING SPEED LIMITS AND ADVISORY SPEEDS

Speed limits are set in one of two ways: (1) determined specifically for a particular road or segment on the basis of an engineering study and displayed as a posted speed on a regulatory sign, or (2) a statutory speed limit that applies in the absence of a posted speed. Statutory speed limits are set forth in state laws. Often, more than one statutory speed limit applies within a state. For example, one speed limit may apply everyplace except within urban areas for which a different and lower limit applies.

Figure 12. Speed Limit Sign

The authority to establish speed limits varies by jurisdiction and can include elected bodies, state DOTs and local transportation, public works and police agencies. Posted speeds must be reasonable. Speed limits are only meaningful if the majority of motorists comply voluntarily and that occurs only if a speed limit is reasonable. An engineering study made in accordance with established traffic engineering practices should be conducted as part of setting a posted speed (MUTCD 2B.13); this is required in many states. A universal process for conducting these studies does not exist. As shown in table 3, different agencies often consider different factors. Different studies with similar circumstances may sometimes lead to a different speed limit. However, any such difference should be relatively minor since the 85th percentile speed is a dominant factor in establishing posted speeds. The (MUTCD) (16) indicates that posted speeds "should be within 5 mph of the 85th-percentile speed of free-flowing traffic."

Table 3. Primary factors used to establish posted speeds

Source: Ref (17)

State or local officials often receive citizen requests for speed limit reductions because of perceived excessive speeds. However, research has repeatedly shown that changes in posted speeds have little effect on operating speeds. (17)

Advisory speed plaques can be displayed on any warning sign to inform the road user of the recommended speed for the warning condition ahead. The placement of warning signs should be based on an engineering study or engineering judgment. Since advisory speed plaques often apply to a particular feature, the signage may combine information regarding the feature and associated advisory speed. An advisory speed plaque is shown on the left side of Figure 13. The warning sign on the right side of Figure 13 combines the right curve warning symbol with an advisory speed.

Figure 13. Speed Limit Sign

Advisory speeds for horizontal curves are often determined by driving a vehicle equipped with a ball-bank indicator through the curve several times at different speeds and applying the Green Book guidelines shown in . These guidelines are based on relatively old values for driver comfort. It should be noted that the table 4 friction values are less than the maximum side friction factors used in design and discussed in chapter 4 (Horizontal Curvature and Superelevation).

Table 4. Recommended advisory speeds with corresponding ball bank reading.

Source: Ref (1)

Advisory speeds are displayed on warning signs in speed values that are multiples of 5 mph. For further information from the 2009 MUTCD, see Advisory Speed Posting in the next chapter.

ENFORCING SPEED LIMITS

Traffic laws, including speed limits, are enforced by police agencies at the state, county and municipal levels. It is extremely rare and generally considered counterproductive to cite drivers operating slightly over the speed limit. Since exceeding the speed limit is so common, it is not practical to issue a ticket to each and every offending driver. Flagrant violators (i.e., drivers operating at very high speeds) pose the greatest risk and are generally the focus of enforcement. Police exercise discretion in deciding at what speed and circumstances a citation will be issued. Police have no specific knowledge of designated and inferred design speeds. Individual vehicle speeds are assessed on the basis of the speed limit, prevailing operating speeds, and environmental conditions.

Decisions on when and where to enforce speed limits directly affect driver speed selection. Visible and active enforcement reduces operating speeds but the effect diminishes as the distance and time from the enforcement increases. Although several automated enforcement programs have been instituted in the U.S., speed enforcement is still accomplished primarily by sworn police officers. The cost and availability of assigning police officers to this function limits the frequency, coverage, and effectiveness of speed enforcement.

SPEED BASED TECHNICAL PROCESSES

As discussed in previous sections of this chapter, there are a variety of decisions related to traffic speed. These decisions are made by different groups and people, within different levels of government and at different times. In this section, speed-related developments and decisions are reviewed over the life cycle of a road or street. The patterns described are common but not accurate for each and every facility.

The conceptual relationship shown in figure 14 between designated design speed, operating speed, and speed limit can be thought of as ideal. The concept depicted is not found in any authoritative publication or manual on traffic speed, but conforms to the general engineering approach. Under this concept, public transportation agencies establish a speed limit less than the designated design speed. Drivers, with rare exceptions, heed the limits. In the life cycle of transportation facilities, design comes first. Through the design process, concepts are converted to operational guidance and then to detailed decisions for implementation. Figure 15 illustrates an interpretation of how these idealized relationships are pursued through operational design guidance. A design speed equal to or higher than the speed limit is usually (but not always) designated when the speed limit is known during the design process. Design speeds are sometimes designated by determining the speed limit and adding a specified value, such as 5 or 10 mph, although the Green Book (1) guidance on design speed does not address speed limits. The relationship between designated design speed and speed limit varies. Speed limits are not necessarily known during the design process. In addition, established speed limits are subject to review and revision.

Figure 14. Conceptually ideal speed relationships.

Figure 15. Typical speed relationships contemplated by design process.

Designated design speed is explicitly determined during the design process. Inferred design speed is determined implicitly (but typically not calculated) as a result of geometric design decisions. As noted previously, the designated and inferred design speeds often differ because designers are encouraged to exceed minimum values for geometric design features that are determined based on the design speed. The result is that many design features meet criteria for design speeds far greater than the designated design speed.

After a road is open to traffic, actual operating speeds may be higher than anticipated as shown in figure 16. This often happens when the designated design speed is less than the desired speed, a common occurrence for streets and roads with low and moderate designated design speeds. The desired speed is defined as the speed, under free-flow conditions, that drivers choose to travel when not constrained by roadway design features (18). The MUTCD (16) recommends that posted speeds be within 5 mph of the 85th percentile operating (free-flowing) speed. If the speed limit is changed to reflect actual operating speeds, the relationship between design speed and posted speed will be altered. Figure 17 illustrates a resulting scenario of a speed limit being higher than the designated design speed.

Figure 16. Speed relationships that sometimes develop on low and moderate design speeds.

Figure 17. Relationships that can result when speed limit is increased on basis of actual operation speeds.

The undesirable conditions shown in figures 16 and 17 occur more frequently on certain roadway types.  There is a concern that raising the speed limit, as shown in figure 17, will lead to even higher operating speeds and thus contribute to a cycle of speed escalation and reduced levels of safety.

Some people have difficulty understanding how it can be acceptable for a speed limit to exceed the designated design speed. First, geometric design criteria above minimum values for a designated design speed are often used. Secondly, geometric design criteria have been developed using a variety of premises. For example, horizontal curve criteria are based on "comfort" levels, the sensory response of vehicle occupants to unbalanced lateral acceleration. The values were developed in the 1940s. Lastly, the criteria are based on assumed conditions that are improbably poor (e.g., actual conditions will likely be better). Consequently, driving at the design speed under favorable environmental conditions provides a margin of safety that can compensate for unexpected developments, below-average driver performance and less favorable road conditions.

Over the life cycle of a road, different groups and individuals make speed-related decisions. These groups have different roles and often use different information in their decision processes. A set of speed-related references are listed below.

Each of these references provides information, direction or guidance on a particular facet of a complex subject. None provides an overall or unifying approach to managing speed.