Recent Roadway Geometric Design Research for Improved Safety and Operations (2012)

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

59 Overview Similar to intersections, new ways to design interchanges received attention during the last 10 years in an attempt to improve capacity while minimizing the cost of constructing or expanding the interchange. Researchers also revisited charac- teristics of ramp design and ramp terminal design, and they considered the effects of work zones near interchanges. Design Of ramps anD ramp Terminals Chaudhary and Messer (2002) developed guidelines for design- ing freeway on-ramps in which ramp metering is envisioned. Specifically, they looked for design issues in which ramp meters use a queue detector to identify and prevent a queue of vehicles from blocking the upstream intersection. They focused on three design elements: safe stopping distance, storage distance, and acceleration distance from meter to merge point. Combining these three elements, they stated that the desired distance between the cross street and freeway merge point be at least 400 m (1,312 ft) for ramps at which metering is envisioned. Fitzpatrick and Zimmerman (2007) reviewed the Green Book’s process for adjusting acceleration and deceleration lengths on graded ramps. They found that the source of the adjustment factors in the 2004 Green Book was provided in the 1954 Policies on Geometric Highway Design (i.e., the Blue Book, AASHO 1954), in which they first appeared as being based on applying “principles of mechanics to rates of speed change for level grades.” Their reviews of that document and others did not reveal a procedure for deter- mining adjustment factors. They posited that a potential source for an adjustment factor for entrance ramps is the calculation of the distance needed to accelerate from one speed to another on different grades by means of vehicle performance equations available in the literature, and thus they reviewed the literature to develop potential accelera- tion length adjustment factors. They subsequently applied the Green Book methodology for calculating SSD on dif- ferent grades to the equations used to calculate decelera- tion lengths so as to determine deceleration lengths for different grades. The ratio of the deceleration length on a grade to the deceleration length on a level surface formed the basis for their adjustment factors for deceleration. They recommended that actual performance of vehicles on grades and on a level surface should be measured and compared with the suggested adjustment factors to determine the accu- racy of those factors. ramp and interchange spacing Under NCHRP Project 03-88, researchers evaluated and summarized design, operations, safety, and signing consid- erations that influence ramp and interchange spacing deci- sions (Ray et al. 2011). The Green Book contains guidelines on the distance between successive ramp terminals, but they “are acknowledged to be based on operational experi- ence and recommend basing actual spacing on operations and safety procedures derived from applied research.” To provide a better understanding of the impacts of ramp and interchange spacing on safety and operations, research- ers collected and analyzed data from a variety of existing freeway ramps and interchanges, focused on relatively simple, single lane, service ramps and interchanges. The team con- ducted operational and safety assessments of two types of ramp pairs—an entry ramp followed by an exit ramp (EN-EX) and an entry ramp followed by another entry ramp (EN-EN). They then performed simulation modeling, calibrated with field data, of closely spaced pairs of ramps and developed safety performance models. Based on their findings, the team then developed guide- lines to assist practitioners in selecting ramp and interchange spacing values for their particular design context. These guide- lines presented substantial discussions on geometric design, traffic operations, safety, and signing, and the role each of these play in determining ramp and interchange spacing needs. The guidelines made a distinction to separately define “ramp spacing” and “interchange spacing” and recommended ramp spacing values be the primary consideration in freeway and interchange planning and design. Guidelines were pre- sented based on four areas of emphasis: geometric design, traffic operations, signing, and safety. Geometric design principles, as well as site-specific features, dictate minimum lengths needed for ramps and other interchange compo- nents. Traffic volumes can necessitate increased spacing beyond the dimensions needed purely for geometrics. Safety tradeoffs, which have rarely been quantified until recently, can now be considered in project decision making. Finally, signing and other human factors issues should be taken into account at the earliest in the evaluation process when chapter six inTerchanges

60 making choices about ramp and interchange spacing. The guidelines were presented as information that can also be incorporated in future editions or updates of relevant man- uals and other guidance documents. Among the geometric design guidelines are spacing assessments, as shown in Table 20. access management The adequate spacing and design of access to crossroads in the vicinity of freeway ramps are critical to the safety and traffic operations of both the freeway and the crossroad. Rakha et al (2008) conducted research for the Virginia DOT to develop a methodology to evaluate the safety impact of different access road spacing standards. The models they developed were used to compute the crash rate associated with alter- native section spacing, and the authors concluded that the models satisfied statistical requirements and provided rea- sonable crash estimates. Their results indicated that the crash rate decreased by 88% when access road spacing increased from 0 to 300 m. An increase in the minimum spacing from 90 m (300 ft) to 180 m (600 ft) resulted in a 50% reduction in the crash rate. The models were used to develop lookup tables that quantified the impact of access road spacing on the expected number of crashes per unit distance. Those tables revealed a decrease in the crash rate as the access road spacing increases. The researchers also attempted to quantify the safety cost of alternative access road spacing using a weighted average crash cost. The weighted average crash cost was computed based on the observed distribution of crashes in Virginia that were fatal, injury, and property damage crashes. Costs of crashes in each severity category were provided by the Virginia DOT, which the researchers used to compute an average weighted crash cost. This aver- age cost was multiplied by the number of crashes per mile to compute the cost associated with different access spacing scenarios. The researchers developed tables containing the cost data that planners, designers, and policymakers could use in determining their choice of intersection and access spacing for specific freeway ramp locations. managed lanes Fitzpatrick et al. (2003b) conducted an evaluation of managed lane ramp design issues in Texas, with a comparison to then- current practices in national and other states’ guidelines. The 2001 Green Book (AASHTO 2001) specified a 2,000-ft weaving section for a system-to-service interchange. For a direct-connection ramp between a traffic generator and the managed lane, AASHTO recommended a minimum design speed for direct connection ramps of 40 mph, whereas Cali- fornia’s guidelines (Caltrans 2001) called for a minimum of 50 mph. Each state’s guidelines that contained specific discussions on the spacing between successive ramps used approximately 900 to 1,000 ft spacing. They also used computer simulation “to obtain an appre- ciation of the effects on corridor operations when several pairs of ramps are considered. Speed was the primary mea- sure of effectiveness used to evaluate the effects of different ramp spacings, volume levels, and weaving percentages. The research found that a direct connect ramp between a gen- erator and the managed lane facility should be considered when 400 veh/hr is anticipated to access the managed lanes. If a more conservative approach to preserving freeway per- formance is desired, then a direct connect ramp should be considered at 275 veh/hr (which reflected the value when the lowest speeds on the simulated corridor for the scenarios examined were at 45 mph or less).” This finding builds on the recommendations made by Venglar et al. (2002) on weaving distances for managed lane cross-freeway maneuvers, shown in Table 21. Combination Ramp Spacing Dimension (ft) Feasibility Diamond Interchange Entrance–Exit Less than 1,600 Likely not geometrically feasible 1,600 to 2,600 Potentially geometrically feasible Greater than 2,600 Likely geometrically feasible Partial Cloverleaf Entrance–Exit Less than 1,600 Likely not geometrically feasible 1,600 to 1,800 Potentially geometrically feasible Greater than 1,800 Likely geometrically feasible Entrance–Entrance Less than 1,400 Likely not geometrically feasible 1,400 to 1,800 Potentially geometrically feasible Greater than 1,800 Likely geometrically feasible Exit–Exit Less than 900 Likely not geometrically feasible 900 to 1,100 Potentially geometrically feasible Greater than 1,100 Likely geometrically feasible Exit–Entrance (Braided) Less than 1,700 Likely not geometrically feasible 1,700 to 2,300 Potentially geometrically feasible Greater than 2,300 Likely geometrically feasible Source: Ray et al. (2011). TABlE 20 POTENTIAl FEASIBIlITy OF SPACING FOR VARIOUS FREEwAy RAMP COMBINATIONS

61 Fitzpatrick et al. (2007) developed guidance materials on intermediate at-grade access to a buffer-separated managed lane. They determined that compliance with access points was better for those with greater lengths (e.g., 1,500 ft), but that over 7% of observed access maneuvers involved vehicles using the managed lane to pass slower-moving vehicles. They also found “that when presented with the opportunity to enter a managed lane that is located very close to an entrance ramp, drivers will attempt to cross multiple lanes to do so.” Providing sufficient weaving distance for cross-freeway maneuvers was therefore important to facil- itate access to the managed lane. For the design of the at- grade access opening, they recommended the configuration shown in Figure 21. Toll facilities A particular type of managed facility is a tolled facility. Some tolled facilities are separate roadways on unique alignments, whereas others are selected lanes on a concurrent alignment with a general-purpose facility. Each has particular charac- teristics to consider when designing access points, whether they are at-grade openings or full-fledged interchanges. In ITE’s Freeway and Interchange Geometric Design Hand- book (leisch et al. 2005), McDonald describes details of geo- metric design elements for toll plazas. Although a number of the practices listed are influenced by traditional manned toll plazas where tickets and cash are exchanged, there are a variety of examples and principles that are also valid for unmanned electronic toll collection. Toll plazas typically have more lanes than adjacent sec- tions of a freeway and require sufficient merge and diverge tapers to accommodate the added lanes. Similarly, a toll plaza or toll island on a ramp requires enough lanes to serve the anticipated demand, necessitating the addition and/or reduction of lanes on the ramp proper. Table 22 repro- duces the information from the Florida Turnpike cited by McDonald for taper rates at toll plazas with traditional payment collection. Electronic toll collection methods have improved capac- ity at toll plazas, but there is still a need to accommodate the anticipated volume of vehicles using the facility. McDonald provides a detailed procedure for estimating the appropri- ate number of queue lanes and queue length, depending on toll collection method, but he also provides a general rule of thumb from Caltrans to provide 3.5 to 4 toll lanes per approaching freeway lane and a minimum queue storage length between 200 and 250 ft (leisch et al. 2005). alTernaTive inTerchange Designs An FHwA study (Hughes et al. 2010) examined two alter- native interchange designs, reviewing characteristics related to geometric design, access management, traffic control devices, and other features. The two designs included Double Crossover Diamond (DCD) and DlT interchanges. Addi- tional studies have also evaluated these interchange designs and others. Findings from those studies related to geometric design are summarized in this section. Double crossover Diamond/Diverging Diamond The DCD interchange, also called a Diverging Diamond interchange (DDI), is a recent interchange design that is being considered as a viable interchange form to improve traffic flow and reduce congestion. Similar to the design of a conventional diamond interchange, the DCD interchange differs in the way that the left and through movements navigate between the ramp terminals. The purpose of this interchange design is to accommodate left-turning move- ments onto arterials and limited-access highways while eliminating the need for a left-turn bay and signal phase at the signalized ramp terminals. Figure 22 shows the typical movements that are accommodated in a DCD interchange. The highway is connected to the arterial cross street by two on-ramps and two off-ramps in a manner similar to a conventional diamond interchange. However, on the cross street, the traffic moves to the left side of the roadway between the ramp terminals. This allows the vehicles on TABlE 21 wEAVING DISTANCES FOR MANAGED lANE CROSS-FREEwAy MANEUVERS Design Year Volume Level Allow up to 10 mph Mainline Speed Reduction for Managed Lane Weaving Intermediate Ramp (between freeway entrance/exit and managed lanes entrance/exit)? Recommended Minimum Weaving Distance Per Lane (ft) Medium (LOS C or D) Yes No 500Yes 600 No No 700Yes 750 High (LOS E or F) Yes No 600Yes 650 No No 900Yes 950 Source: Venglar et al. (2002). Note: The provided weaving distances are appropriate for freeway vehicle mixes with up to 10% heavy vehicles; higher percentages of heavy vehicles will require increasing the per lane weaving distance. The value used should be based on engineering judgment, although a maximum of an additional 250 ft per lane is suggested.

62 See Detail A See Detail C Shoulder Intermediate Access Managed Lane 4" White Lane Lines 4" Yellow Edge Line General Purpose Lanes 1300' Minimum - 1500' Desired Shoulder or Median All pavement marking materials shall meet the required Departmental Material Specifications as specified in plans. = Direction of travel. Notes: 4' T yp ica l Managed Lane General Purpose Lane 8" White Line 8" White Line Detail A 8" White Line 4' T yp ica l Detail C 3' 3' 12'12' Managed Lane General Purpose Lane 3' 12' 45' Raised Pavement Markers Type II-C-R 8" White Line Detail B 3' 3' 12'12' Managed Lane General Purpose Lane 3' 12' 45' Raised Pavement Markers Type II-C-R 3' See Detail B Raised Pavement Markers Type I-C or Type II-C-R 40' Max. - 20' Desirable1" Gap FIGURE 21 Design of intermediate at-grade access opening for buffer-separated freeway managed lane (Fitzpatrick et al. 2007). Plaza Type Number of Traditional Payment Lanes at Plaza D esirable Taper Rate Mainline Plaza Up to 8 lanes 25: 1 10 to 14 lanes 20: 1 16 or more lanes 15: 1 Ramp Toll Plaza All 20: 1 Source: Leisch et al. (2005). TABlE 22 DESIRABlE TAPER RATES AT TOll PlAZAS the cross street that need to turn left onto the ramps to con- tinue to the on-ramps without conflicting with the opposing through traffic (Hughes et al. 2010). The primary design element of a DCD interchange is the relocation of the left-turn and through movements to the opposite side of the road within the bridge structure. The turn- ing radii used at the crossover junction to displace these movements at an existing installation in Springfield, Missouri,

63 are approximately 300 ft. FHwA advises that consider- ation should be given to designing radii at crossovers with heavy vehicles in mind. On rural locations where the minor street has high-speed limits, the use of reverse curvature has been suggested. This may result in loon-like flare-outs at the ends of the bridge structure, and additional right-of- way may be required to widen the bridge or the underpass structure. Median width is also an important design element for a DCD interchange (Hughes et al. 2010). Greater median width is required for the flaring needed for reverse curves. Designers are advised to obtain minimum median widths from the Green Book and to take into account the installa- tion of post-mounted signs on medians on the bridge deck for safe and effective channelization of traffic. Appropriate offsets for signs should be in accordance with the MUTCD. The report states that driver simulator experiments on the Missouri DCD interchange, which included the use of glare screens, showed no erroneous maneuvers by tested subject drivers. Suggested design practices, based on input from Missouri DOT, include the following: • The minimum crossing angle of the intersection should be 40 degrees. • The radius design should accommodate between 25 and 30 mph. • Superelevation may not be needed because it could detract from any desired traffic calming effect. • lane width should be approximately 15 ft. • Design should accommodate wB-67 trucks. • Adequate lighting should be provided. • Nearside signals should be considered. • DCD interchange designs may only be appropriate where there are high-turning volumes. • Nearby intersections with long cycle lengths should be avoided. • Pedestrian crossings at free-turning movements should be evaluated and pedestrian signals may be needed. • The noses of the median island should extend beyond the off-ramp terminals to improve channelization and prevent erroneous maneuvers. • left- and right-turn bays should be designed to allow for separate signal phases. Bared et al. (2005) used simulation to compare the opera- tional performance of a four-lane DDI with a conventional diamond. They concluded that performances for lower and medium volumes are nearly identical in both designs; how- ever, their results from higher volumes showed that the con- ventional diamond had lower throughput, higher average delay per vehicle, greater stop time, longer queues, and max- imum off-ramp flows as compared with the DDI. Evaluation of a six-lane DDI at three scenarios with very high volume indicated that the left-turn capacity of the DDI was twice that of the conventional diamond. Displaced left-Turn The DlT interchange, also known as the continuous flow interchange, is an innovative interchange design that has sev- eral aspects similar to the at-grade DlT intersection and some aspects similar to the DCD interchange. It is a design treat- ment that has been advocated as promising because it removes the conflict at the main intersection between left-turning and opposing through vehicles (Hughes et al. 2010). The main feature of the DlT interchange design is the left- turn crossovers that are present on the cross-street approaches. In a DlT intersection, the left-turning traffic is relocated at a location several hundred feet upstream of the first sig- nal-controlled ramp terminal of the diamond interchange, shown at the right side of Figure 23. This left-turning traffic (shown as a dashed line) is crossed over the opposing through lanes. The traffic then travels on a new roadway that is situ- ated between the opposing through lanes and a roadway and that carries the right-turning traffic from the ramp. Drivers FIGURE 22 Typical DCD interchange configuration (Hughes et al. 2010).

64 make the left turn onto the ramp from the new roadway after crossing over the freeway, as shown at the top of Figure 23. As with a DlT intersection, the differentiating design ele- ment of a DlT interchange is the left-turn crossover. The DlT lanes typically cross the opposing through traffic at locations that are approximately 400 to 500 ft upstream of the signal- controlled ramp terminals. Geometrically, the left-turn cross- over in a DlT interchange is similar to the design of a left-turn crossover for a DlT intersection. Hughes et al. (2010) cite research into the operation of DlT intersections sponsored by the Maryland State Highway Administration that revealed that the distance between the crossover and the main intersection was dependent on queuing from the main intersection and on costs involved in constructing a left-turn storage area. Radii of the crossover movements range from 150 to 200 ft. The radii of the left-turn movement at the nodes of the interchange are dependent on the turning movement of a design vehicle. Median width affects the interchange footprint and con- sequently the right-of-way acquisition. As with DlT inter- sections, FHwA encourages designers to obtain minimum median widths from the AASHTO Green Book; offset rec- ommendations for post-mounted signs should be accounted for in accordance with MUTCD when determining median width, though the minimum median width for any type of intersection or interchange is 4 ft (Hughes et al. 2010). The authors further state that a wide median is counterproductive at a DlT interchange for the following reasons: • wide medians result in long walking distances for pedes- trians at the interchange. In turn, this results in the need for long pedestrian clearance intervals and potentially increased cycle lengths, which is counterproductive to traffic efficiency. • wide medians necessitate a wide interchange footprint and consequently higher bridge deck construction costs. wOrk ZOne cOnsiDeraTiOns NCHRP Report 581 (Mahoney et al. 2007) discusses a variety of geometric design principles and their applications within work zone traffic control on high-speed highways. In partic- ular, the authors discussed the appropriate design principles for temporary entrance and exit ramps. They stated that a temporary single-lane interchange ramp should have a travel lane of approximately 4.5 m (15 ft), with a 1.8-m (6-ft) right shoulder and minimum 0.6-m (2-ft) left shoulder. However, different cross-sectional arrangements are appropriate when supported by agency experience and in consideration of project-specific factors (e.g., traffic volume, mix, and duration of service). with respect to entrance ramps, they concluded that the fea- sibility of maintaining an entrance during construction often hinges on providing an adequate combination of roadway geometry and traffic control to facilitate merging. Figure 24 illustrates a temporary entrance ramp for a median crossover. The authors concluded that the basic principles and issues associated with permanent ramps also pertain to temporary arrangements. Therefore, acceleration lanes in work zones that meet the design criteria for permanent facilities were desirable. However, providing these lane lengths was often not practical. Thus, they provided several “rules of thumb” as options to guide designers: • Provide at least 90 m [300 ft] of acceleration lane. • Provide at least 70% of the permanent roadway criteria length. • Employ traffic control measures (e.g., STOP, yIElD, and other signs) to mitigate less-than-desirable accel- eration lane lengths. The authors also offered similar guidance for exit ramps, stating that it was desirable for exiting traffic to depart the through lanes at mainline speed and not reduce speed while occupying the mainline through lane. when this is not prac- tical, they recommended that the geometry of the ramp be reviewed to determine if the ramp’s length, horizontal align- ment, and grade allow for gradual deceleration before reaching speed-critical features. summary Of key finDings This section summarizes key findings from the research noted in this chapter. This is an annotated summary; conclu- sions and recommendations are those of the authors of the references cited. interchange ramp Design • The desired distance between the cross street and freeway merge point is at least 400 m (1,312 ft) for ramps at which metering is envisioned (Chaudhary and Messer 2002). • The source of the adjustment factors in the 2004 Green Book was provided in the 1954 AASHTO Blue Book, in which they first appeared as being based on apply- ing “principles of mechanics to rates of speed change for level grades.” Further review did not reveal a pro- cedure for determining adjustment factors. A new pro- cedure contains an alternative set of adjustment factors for acceleration length and deceleration length, the latter FIGURE 23 Detailed view of movements and paths for half of a DLT interchange.

65 of which is based on the ratio of the deceleration length on a grade to the deceleration length on a level surface. Actual performance of vehicles on grades and on a level surface should be measured and compared with the sug- gested adjustment factors to determine the accuracy of those factors (Fitzpatrick and Zimmerman 2007). ramp and interchange spacing • Recent guidelines make a distinction to separately define “ramp spacing” and “interchange spacing” and recom- mend ramp spacing values be the primary consideration in freeway and interchange planning and design (Ray et al. 2011). • Guidelines are presented based on four areas of emphasis: geometric design, traffic operations, signing, and safety. Geometric design principles, as well as site-specific features, dictate minimum lengths needed for ramps and other interchange components. Traffic volumes can necessitate increased spacing beyond the dimensions needed purely for geometrics. Safety tradeoffs, which until recently have rarely been quantified, can now be considered in project decision making. Finally, signing and other human factors considerations should be taken into account at the earliest in the evaluation process when making choices about ramp and interchange spacing (Ray et al. 2011). • Spacing assessments indicate that ramp spacing of less than 900 ft is likely not geometrically feasible. That spacing value increases up to 1,600 ft for entrance–exit ramp pairs (Ray et al. 2011). alternative interchange Designs • Design practices for the DDI (Hughes et al. 2010) include: – The minimum crossing angle of intersection should be 40 degrees. – The radius design should accommodate between 25 and 30 mph. – Superelevation may not be needed because it could detract from any desired traffic calming effect. – lane width should be approximately 15 ft. – Design should accommodate wB-67 trucks. – Adequate lighting should be provided. – Nearside signals should be considered. – DCD interchange designs may only be appropriate where there are high-turning volumes. – Nearby intersections with high cycle lengths should be avoided. – Pedestrians at free-turning movements should be evaluated, and pedestrian signals may be needed. – The noses of the median island should extend beyond the off-ramp terminals to improve channelization and prevent erroneous maneuvers. – left- and right-turn bays should be designed to allow for separate signal phases. • The Displaced left-Turn interchange has functions similar to a DlT at-grade intersection. DlT lanes typi- cally cross the opposing through traffic at locations that are approximately 400 to 500 ft upstream of the signal- controlled ramp terminals. Minimum median widths are preferred for this design (Hughes et al. 2010). FIGURE 24 Temporary interchange entrance ramp for median crossover (Mahoney et al. 2007).