Potential MUTCD Criteria for Selecting the Type of Control for Unsignalized Intersections (2015)

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.

88 CHAPTER 7: OVERVIEW, CONCLUSIONS, AND RECOMMENDATIONS OVERVIEW As stated in the introduction of this report, the intent of this research study was to develop criteria and supporting material for the selection of appropriate right-of-way control at an unsignalized intersection, and for those criteria to be in a format that can be integrated into a future revision of the MUTCD. Within the context of this research, an unsignalized intersection is one where one of the following methods of right-of-way control is used on one or more of the approaches:  No control: Right-of-way is based on the rules of the road where the first to arrive at the intersection has the right of way, and if two vehicles arrive at the same time, a driver yields to the vehicle to the right.  Yield control: YIELD sign(s) are installed on the minor approach or approaches. At a roundabout intersection, YIELD signs are installed on all approaches.  Minor-road stop control: STOP sign(s) are installed on one approach for a three-leg intersection or on two approaches for a four-leg intersection. The STOP sign is normally installed on the minor road but in some cases may be installed on the major road with no control on the minor road.  All-way stop control: STOP signs are installed on all approaches to the intersection. The next level of right-of-way control for an intersection is a traffic control signal, criteria for which were not included in the scope of this research. This project was conducted within seven tasks that included a review of literature and existing policies and guidelines, identification of intersection and traffic characteristics, a critical evaluation of relevant sections of the 2009 MUTCD, an economic analysis to evaluate control alternatives, development of recommendations for revisions to the MUTCD, and the completion of this report. Review of Policies and Guidelines Researchers reviewed the current MUTCD and the supporting material for the guidance found therein. The research team also conducted searches of guidelines and manuals from all 50 states (available online) to review their current policies. In addition, researchers asked practitioners for information on novel approaches they were considering for selecting traffic control at unsignalized intersections. Several states provide guidance in addition to that found in the MUTCD, but in many jurisdictions, the MUTCD (or a particular state’s equivalent) is the prevailing source for guidance. Much of the existing text in the MUTCD has remained largely intact for several decades. Literature Review The research team had a three-pronged approach to reviewing the relevant literature: key reference documents, previous literature that discussed methods for selecting traffic control at

89 unsignalized intersections, and previous literature that discussed methods for selecting traffic control at unsignalized pedestrian crossings. Key reference documents included the Highway Capacity Manual (23), Highway Safety Manual (24), and ITE Manual of Traffic Engineering Studies (25). Literature that described the selection of traffic control included processes that considered delay, traffic volumes, number of lanes, crashes, and other variables. Some processes resulted in regression equations or charts to calculate the variables of interest, while others were based on a point system that described a recommended traffic control for a certain point score. Key intersection and traffic characteristics included in these processes are summarized in Chapter 4. Critical Review of MUTCD Researchers reviewed key sections of MUTCD Chapter 2B to determine which sections could have the most potential benefit from new research to support revised guidance. Numeric and non-numeric criteria for traffic control in Sections 2B.04, 2B.06, and 2B.07 were reviewed, and comments for potential revisions were made as provided in Chapter 5. Based on the activities in the initial phase of the project, the research team, with the guidance of the project panel, conducted a study in the second phase of the project that focused on the following items:  Set a higher priority on investigating when to go from TWSC to AWSC rather than when to go from no control to yield or TWSC.  Develop criteria that reflect urban and rural or speed conditions.  Develop criteria that are sensitive to the number of legs at the intersection.  Consider roundabouts as a geometric design alternative within the evaluation.  Consider a variety of major- and minor-road volume splits and not just when the split is “approximately equal.”  Consider the existing and ongoing revisions to relevant sections of the MUTCD, such as the changes suggested for the reorganization. Economic Analysis Procedure The research team used a procedure for comparing traffic control alternatives based on the relative economic costs and benefits of those alternatives for particular intersection types (three- leg or four-leg), environments (urban or rural), and volumes (varying levels of major- and minor- road volumes). Based on information from a variety of relevant sources, the research team selected user delay, crashes, vehicle operating, and construction as the four costs for consideration in the project. Researchers used microsimulation to measure the effects of delay. A multi-step process for calculating crash costs was adapted from the HSM. Vehicle operating costs were estimated using information from federal sources such as the Environmental Protection Agency and the Energy Information Administration. Roundabout construction costs were estimated from FHWA information.

90 Potential Criteria Table 49 summarizes potential criteria for AWSC as identified from the literature review, from the review of policies and guidelines, or as part of the economic analysis. Table 50 summarizes criteria identified from the reviews for no control, yield control, or minor-road stop control. Table 49. Potential Criteria for AWSC. Criteria All-Way Stop Number of Crashes Susceptible to Correction by Intersection Control (e.g., Right-Turn, Left-Turn, or Right- Angle Crashes)  5 or more within 12 months (1)  4-leg: 5 or more within 12 months, 6 or more within 36 months, proposed crash warrant criteria for signals (49)  3-leg: 4 or more within 12 months, 5 or more within 36 months, proposed crash warrant criteria for signals (49)  5 or more preventable crashes within 12 months (21) Peak-Hour Entering Volume  500 veh/hr major approach and 210 veh/hr minor approach (and 300 veh/hr major approach and 300 veh/hr minor approach) based on average delay or average queue length (54)  Examples of volumes based on control delay determined using Highway Capacity Software with consideration of percent left turns (56) are shown below: Major % Left Minor % Left Minor % Left Minor % Left Minor 200 5 50 10 50 15 50 20 50 300 5 140 10 100 15 90 20 80 400 5 NR 10 NR 15 105 20 90 Major and minor are veh/hr/approach. NR = not recommended.  500 units/hr/major approach and 300 units/hr/minor approach, values from simulation conducted as part of this research project associated with more than 35 sec/minor-road vehicle Entering Volume per Day  When combined ADT of 7,500 to 50,000 exist, then conduct all-way stop evaluation (16)  Highly desirable for intersection roadways to have closely balanced ADTs, e.g., the volume of at least one minor approach is not less than 70% of higher volume of major approach (18)  4,000 vpd, major 65% of total (21)  17,000 (3-leg) or 21,000 (4-leg) vpd (values calculated using the peak-hour simulation numbers associated with more than 35 sec/minor-road delay and adjusting to a daily value) 8 hr  300 veh/hr entering from major and 200 units (veh and ped) from minor (1)  210 veh/hr entering from major and 140 units (veh and ped) from minor (i.e., minimum vehicular volume warrant is 70%) when 85th percentile major street exceeds 40 mph (1)  430 units/hr/major approach and 260 units/hr/minor approach (values were calculated using the peak-hour simulation numbers associated with more than 35 sec/minor-road delay and the top 8 hourly factors used in the economic analysis and then averaged for the 8 hr) Delay (Minimum)  30 sec/minor-street veh (1)  35 sec/minor-street veh, suggested based on HCM (23) Exhibit 19-1, lowest control delay (sec/veh) for LOS E (when v/c <=1.0) Other  “All-way stop conversion in urban areas is not limited to a certain range of entering volumes that follow current MUTCD warrants and is no less effective when approach volumes are unbalanced as when they are equal on all approaches.” (34)  Need to control left-turn conflicts, vehicle/ pedestrian conflicts, or multi-way stop control would improve traffic operations (1)  Sight distance  Engineering study

91 Table 50. Potential Criteria for No Control, Yield Control, or Minor-Road Stop Control. Criteria No Control Yield Control Minor-Road Stop Number of Crashes Susceptible to Correction by Intersection Control  Box (52) recommended fewer than 2 crashes in 1 year or 4 crashes in 3 years  Maryland MUTCD Table 2B-1a (10) provides guidelines for conversion from stop to yield control and recommends 2 or less reported crashes in 1 year or 4 or less in 3 years  Box (52) recommended fewer than 2 crashes in 1 year or 4 crashes in 3 years  Maryland MUTCD Table 2B-1a (10) provides guidelines for conversion from stop to yield control and recommends 2 or less reported crashes in 1 year or 4 or less in 3 years  Fewer than 3 crashes in 2 years from NCHRP Report 320 (36)  3 or more preventable crashes within 12 months (21)  3 or more within 12 months or 5 or more within 2 years (1)  4-leg: 3 or more within 12 months, 6 or more within 36 months, suggested with consideration of the proposed crash warrant criteria for signals (49)  3-leg: 3 or more within 12 months, 5 or more within 36 months, suggested with consideration of the proposed crash warrant criteria for signals (49)  3 or more preventable crashes within 12 months (21) Peak-Hour Entering Volume  Maximum 80 units/hr (rounded calculation from the 1,000 units/day value using 7.8 percent, which is the peak-hour factor used in the economic analysis)  Box (52) recommended less than 100 total entering volume during peak hour  Maximum 140 units/hr (rounded calculation from the 1,800 units/day value using 7.8 percent, which is the peak- hour factor used in the economic analysis)  No volume criteria identified Entering Volume per Day  1983 study in rural Michigan (41) found no statistical difference for stop-controlled and no- control intersections with major street volumes less than 1,000 vpd  1,500 vpd from Nitzel et al. (51)  Maximum 1,800 units/day from NCHRP Report 320 (36)  Maryland MUTCD Table 2B-1a (10) provides guidelines for conversion from stop to yield control: high priority: major 2,000 vpd and minor less than 200 vpd or medium priority: major 2,000–3,000 vpd and minor 200–500 vpd  1,500 to 3,000 vpd from Nitzel et al. (51)  1,000 vpd (21)  Exceeds 6,000 veh/day (1)  2,000 vpd (21) 8 hr  No volume criteria identified  No volume criteria identified  No volume criteria identified Delay  No delay criteria identified  No delay criteria identified  No delay criteria identified Other  Adequate sight distance  Angle of intersection (72)  Adequate sight distance  Merge conditions  Angle of intersection (72)  Sight distance Note: Underlined text represents values recommended by the authors; the underlining was added to aid in reading the table.

92 CONCLUSIONS Caution Regarding Warrants As criteria were being identified, examined, and considered for inclusion in the proposed new MUTCD language, a caution regarding the term “warrant” was acknowledged. The criteria eventually selected for the MUTCD do not necessarily represent the minimum requirements for installing AWSC. The criteria identify a condition that merits detailed analysis and consideration of right-of-way control through a traffic engineering study. Such a study should review the traffic data, pedestrian and bicycle volume, intersection geometrics, expected operational characteristics, and expected safety impacts to conclude whether AWSC is justified. AWSC offers operational benefits and safety benefits under certain traffic conditions. The intent of a traffic engineering study is to document that the AWSC will improve the safety and efficiency of the intersection. How and whether the above thoughts need to be integrated into the MUTCD are questions that FHWA or members of NCUTCD may need to consider. Use of Findings from Economic Analysis A portion of the research efforts focused on an economic analysis to determine when AWSC or roundabout geometric design should be considered based on cost considerations. In general, the findings from the economic analysis are:  For roundabouts: always install a roundabout when the benefits of installation are greater than the construction costs. For the scenarios tested in this study, such as construction costs of $250,000 and $500,000, a roundabout was always justified for the volume levels studied because the delay and crash savings were greater than the construction costs.  For all-way stop control: do not use AWSC except at rural four-leg intersections with a two- lane major highway (one lane in each direction) when you should use AWSC at all intersections with a major-road volume of 400 units/hr and greater and a minor-road approach volume of 100 units/hr/approach and greater. The research team members do not support implementation of these findings (i.e., do not use all- way Stop on urban intersections—except as an interim measure for a signal—or use in most situations for four-leg intersections on rural two-lane highways) at this time for several reasons, including: 1. The economic analysis is based on several assumptions and calculations. Some of the assumptions are based on data from only one state (e.g., the CMFs are based on North Carolina data). The variability of other input values, such as the predicted number of crashes or the average cost of time, is not known and could have a sizable impact on the results. It is important to recognize that the findings were developed using several assumptions. Each of these assumptions could have an associated range of reasonable values, such that significantly different results could be produced when all the values’ ranges or variability are considered.

93 2. The research results provide specific criteria for the selection of a specific type of unsignalized control, but these results do not provide an appreciation of the range of reasonable values—a limitation frequently present within a decision-making process that is based on specific criteria. Additional MUTCD language should include a caution that the criteria should be only one consideration given that a wide range of factors may impact the decision-making process. 3. The user costs for safety considerations have significantly increased in recent years. As indicated in the analysis procedure, the cost of a single fatal accident is over $9 million. This is significantly higher than safety costs that have been used in the past, and the cost of a single fatality can be much greater than other costs associated with an alternative means of providing right-of-way control. As a result, the analysis procedure used in this study may be inconsistent with the procedures used in the past to develop criteria that are currently in the MUTCD and other guidelines. Accordingly, the application of these criteria may lead to implementation results that are inconsistent with existing decision-making criteria in the MUTCD. 4. This research project has demonstrated a “disconnect” between the economic analysis approach used in this research and the existing peak-hour signal warrant in the MUTCD in that an all-way stop is not warranted until volumes are much greater than those that would warrant a traffic control signal. The team suggests that another research project should use a similar basis to examine all types of intersection traffic control so that the relationships between AWSC and the various signal warrants are consistent. 5. If the peak-hour signal warrant is used as the basis to warrant AWSC, the disparity in volumes may result in extensive delay to the major traffic volumes during other times of the day. This raises concerns that the economic analysis is not the best approach for determining AWSC warrants. The consideration of peak-hour, 4-hr, or even 8-hr traffic conditions can justify intersection traffic control that may not be needed the remaining hours of the day. The traffic control, while beneficial during the limited time frame, creates a disparity for the other hours of the day, imposing unneeded traffic regulation and intersection delay. While the findings from the economic analysis are based on thorough research, because of the inconsistencies identified above, the differences in basis between these criteria and those that are currently in the MUTCD mean that the criteria developed from the economic analysis may not be ready for inclusion in the MUTCD until such time as the existing MUTCD criteria and warrants for traffic signals can also be reevaluated in a manner that considers the impacts of user safety costs in the same manner that this research project did. Only through the use of consistent decision-making criteria can practitioners correctly determine the most appropriate means of providing right-of-way control at an intersection. Use of Findings from Traffic Simulation The research team considered other approaches for determining a minimum volume recommendation for AWSC. Delays in excess of 35 sec/veh on the minor-road approach have been suggested as a tipping point due to it representing LOS E in Exhibit 19-1 of the HCM (23). If so, the volumes when the VISSIM runs are above 35 sec/minor-road vehicle are 500 units/hr/major approach and 300 units/hr/minor approach. Using the hourly volume distribution determined in this project, those peak-hour volumes would equate to 8-hr volumes of

94 430 units/hr/major approach and 260 units/hr/minor approach. (The assumed hourly volume distribution is 7.8 percent for 3 hr and 6.1 percent for 5 hr.) RECOMMENDATIONS Suggested Language for Next Edition of MUTCD Using information available from reviews of existing literature, policies, guidelines, and findings from the economic analysis along with the engineering judgment of the research team and panel, recommendations were developed and are summarized in Table 51. The language proposed for the next edition of the MUTCD for unsignalized intersections developed at the conclusion of this research is provided in the appendix. The proposed language includes introductory general considerations, discusses alternatives to changing right-of-way control, and steps through the various forms of unsignalized control from least restrictive to most restrictive, beginning with no control and concluding with AWSC. Supplemental notes are provided to suggested additions to the current text, which show the reader the source(s) of the material and/or the research team’s reasoning for proposing the text. Future Research Needs This research project demonstrated a “disconnect” between the AWSC results produced by the economic analysis approach and the existing signal warrant in the MUTCD in that AWSC is not warranted until volumes are much greater than those that would warrant a traffic control signal. A research project is needed that would examine all types of intersection traffic control from a similar basis so that the relationships between AWSC and the various signal warrants are consistent. The future research project should also consider the effects of left-turning vehicles along with appropriate consideration of pedestrian and bicycle travel. How should multimodal traffic needs, such as those of pedestrians and bicyclists, be considered within intersection traffic control evaluations? Should delay, perhaps using values available in the HCM (23), set the threshold values, or should values be a function of the setting of the intersection, for example, nearness to school or rural versus urban area. The consideration of peak-hour, 4-hr, or even 8-hr traffic conditions can justify intersection traffic control that may not be needed the remaining hours of the day. The traffic control, while beneficial during the limited time frame, creates a disparity for the other hours of the day, imposing unneeded traffic regulation and intersection delay. The relationship of traffic volumes and intersection distribution of those volumes versus the disadvantages of the traffic control during the remaining hours needs further study. The intersection traffic control impacts under lower traffic volumes are especially onerous for AWSC because it imposes unnecessary stops on the major traffic flows. The same disadvantages have been noted relative to the traffic signal warrants based on limited time periods. It is recommended that further research be directed at variations in traffic volumes, including specific consideration of turning vehicles versus advantages and disadvantages of that intersection traffic control for both AWSC and traffic signal control.

95 Table 51. Recommended Criteria for Unsignalized Intersection Control. Criteria No Control Yield Control Minor-Road Stop All-Way Stop Number of Crashes Susceptible to Correction by Intersection Control No crash criteria Two or fewer reported crashes in a yeara 4-leg: 3 or more within 12 months, 6 or more within 36 monthsb 3-leg: 3 or more within 12 months, 5 or more within 36 monthsb 4-leg: 5 or more within 12 months, 6 or more within 36 monthsb 3-leg: 4 or more within 12 months, 5 or more within 36 monthsb Peak-Hour Entering Volume Maximum 80 units/hrc Maximum 140 units/hrc No volume criteria No volume criteria Entering Volume per Day Maximum 1,000 units/ dayd Maximum 1,800 units/daye No volume criteria No volume criteria 8 hr No volume criteria No volume criteria No volume criteria 1. The vehicular volume entering the intersection from the major street approaches (total of both approaches) averages at least 300 units per hour for any 8 hr of an average day; and 2. The combined vehicular, pedestrian, and bicycle volume entering the intersection from the minor-street approaches (total of both approaches) averages at least 200 units per hour for the same 8 hr; but 3. If the 85th percentile approach speed of the major-street traffic exceeds 40 mph, the minimum vehicular volume warrants are 70 percent of the values provided in Items 1 and 2.f Delay No delay criteria No delay criteria No delay criteria 35 sec/vehg Other Adequate sight distance One-lane approaches Angle of intersectionh Adequate sight distance One-lane approaches Angle of intersectionh Sight distance Sight distance Engineering study a Maryland MUTCD Table 2B-1a (10) provides guidelines for conversion from stop to yield control. b Selected with consideration of the proposed crash warrant criteria for signals, NCHRP Project 07-18 (49). c Rounded calculation from the 1,000 and 1,800 units/day value using 7.8 percent, which is the peak-hour factor used in the economic analysis. d Value selected because a 1983 study in rural Michigan (41) found no statistical difference for stop-controlled and no-control intersections with major-street volumes less than 1,000 vpd, and the 1,000 value is less than the value selected for YIELD sign control (1,800). e From NCHRP Report 320 (36). f Values currently in 2009 MUTCD with changes of vehicular volume to units. g Selected based on HCM (23) Exhibit 19-1, lowest control delay (sec/veh) for LOS E (when v/c ≤ 1.0). hAs recommended in the Handbook for Designing Roadways for the Aging Population (72).