Development of Left-Turn Lane Warrants for Unsignalized Intersections (2013)

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.

131 CHAPTER 6 COMPARISON OF PROCEDURES OVERVIEW OF IDENTIFIED PROCEDURES Several states include warrants for the installation of left-turn lanes. These warrants are primarily based on: • Left-turn volume (or percent of left turns), • Opposing volume, and • Approach volume. Many warrants also suggest consideration of: • Crashes, • Sight distance, • Operations (capacity analysis), or • Functional classification of the major roadway. Table 76 and Table 77 list characteristics of the warrants or guidelines included in the state manuals along with other procedures suggested in the literature. These tables also summarize the input data required, what is compared in the method, and the research team’s observations and comments. Table 76 focuses on those procedures where additional evaluations were conducted as part of this research project. Findings from those evaluations are compared within this chapter. Table 77 provides information on other procedures mentioned in state manuals or in the literature.

132 Table 76. Characteristics of procedures evaluated. Base Procedure Current Use Input Data How Evaluated Comments Conflict Avoidance (Harmelink) Several states use a derivation of the Harmelink procedure. Approach volume per hour. Percent left turns. Opposing volume. Minimum approach volume to avoid conflict is compared to actual approach volume. Has long history of use. Updating of assumptions has been suggested by several authors. Methodology not appropriate for four-lane conditions. Benefit-Cost Ratio Not specifically referenced for left-turn treatments to date. If site-specific data are required, could result in high level of input data. Using a procedure developed, say from this research project, would require volume, number of legs, and urban or rural determination. Computes a benefit to cost ratio; left turn warranted if ratio is greater than 1.0 (or value determined by user). A B/C of 1.0 indicates benefits are greater than costs. Considers delay savings, crash savings, and cost of treatment. Crashes Several state manuals note that crashes are to be considered. Crashes. Engineering judgment is typically used to decide if safety is a concern. Development of turn-lane AMFs provides greater opportunity for additional consideration of crashes. Also available are equations to predict number of crashes, which requires volume, number of legs at the intersection, urban or rural, and, for rural, number of lanes per approach for the calculation. Can be a component of benefit- cost ratio. Delay Reduction A few states include criteria based on delay. Volumes. Volume curves are developed for comparison. Procedure generally warrants left-turn lanes at higher volumes as compared to other approaches. Does not consider safety. Can be a component of benefit- cost ratio. Cost Mentioned in a few manuals. Crash costs. Construction costs. General guidance. Can be a component of benefit- cost ratio. Minimum Volume(s) Some states include minimum volumes. Number of left- turning vehicle. Number of turning vehicles compared to threshold value. Simple and easy to use.

133 Table 77. Characteristics of other procedures identified during the course of this study. Base Procedure Current Use Input Data How Evaluated Comments Capacity Evaluation Capacity evaluations are suggested in several manuals. Volumes. Roadway characteristics, etc. Capacity evaluation conducted to determine needed treatments for desired level of service. Capacity evaluations can include extensive efforts. Speed Differential Generally not mentioned or used. Volumes. Perhaps arrival patterns. Is anticipated speed difference greater than threshold value? Need speed prediction for existing or anticipated conditions. Policy/ Functional Classification Considered in several state manuals. Road characteristics (e.g., speed limit). If characteristic is present, then left- turn lane is warranted. Design Consistency Noted in several state manuals. Design characteristics in the corridor. General guidance. Sight Distance Considered in several state manuals. Speed. Engineering analysis of available sight distance. Catalyst (Major, Spot, or Development) Mentioned in a few state manuals. Any of the above, plus reason left- turn lane is being considered. Any of the above, plus consideration of reason left-turn lane is being considered. Should different criteria be used for different funding sources? CONFLICT AVOIDANCE (HARMELINK) PROCEDURE The current AASHTO guideline for installing a left-turn lane at an unsignalized intersection is based on Harmelink’s volume warrants. Harmelink’s warrants are derived from queuing models considering the probability of a through vehicle arriving behind a stopped left-turn vehicle. Limitations A number of limitations of Harmelink’s have been identified over the years. Following is a review of those identified limitations. The same concept of limiting the probability of a through vehicle arriving behind a stopped left- turn vehicle is used for both two-lane and four-lane roadways in Harmelink’s warrant development. Even though different values of maximum allowable probability are used for two- lane and four-lane roadways, this approach ignores the fact that lane changes can take place for through vehicles on the multilane approach to avoid being stopped behind a stopped left-turning vehicle. The queuing model-based probability may not directly correlate with the delay and safety impact experienced by the through vehicles on four-lane roadways. Therefore, revised Harmelink-based warrants were not generated for four-lane roadways as part of NCHRP Project 3-91.

134 Harmelink’s warrants were developed based on allowable maximum probability of 0.020, 0.015, and 0.010 for operating speeds of 40, 50, and 60 mph, respectively, on two-lane highways. These values are subjective in nature and do not directly indicate stops, delay, speed change, or level- of-service degradation in any quantitative manner. Harmelink assumed arrival time for both left-turn and through vehicles to be negative exponentially distributed, an assumption that is only valid for isolated intersections under low demand conditions. It has been reported by Kikuchi and Chakroborty (19) that there are problems in Harmelink’s formulation. An inconsistency in the definitions of arrival and service rate in the queuing model could become a problem when through arrival demand becomes high. The total number of possibilities of making a left turn is overestimated because the formulation uses average gap values in the opposing flow that include unusable residual gaps. Harmelink’s warrants were developed based on dated model parameter values. Researchers have suggested that changes are needed for the time drivers use to complete a left turn, the time to clear the left-turn lane, and the critical gap. Harmelink’s warrants are also based on “typical” drivers and geometry and cannot be used for special geometries and for safety evaluation of left turns at unsignalized intersections by special driver groups such as older drivers. Comparison Between Harmelink Assumptions and Field Study Findings The Harmelink procedure uses critical gap, time to clear the advancing lane, and time to clear the opposing lane(s), along with the probability of a through vehicle arriving behind a stopped, left- turning vehicle, to determine the warrants. Gap and clearance time values were gathered in the field studies conducted as part of this project (see Chapter 4). Table 78 compares the critical gap findings from the field studies to the values used by Harmelink. Table 79 shows the time-to-clear values from this field study and those assumed in the Harmelink procedure. Table 78. Values for critical gap. Conditions Harmelink Current Study Crossing One Lane (11 to 12 ft) 5.0 sec Raff/Hart: 5.12 sec Logit 50th: 4.60 sec Logit 85th: 6.26 sec Crossing One Wide Lane (17 to 20 ft) Raff/Hart: 5.43 sec Logit 50th: 5.14 sec Logit 85th: 6.62 sec Crossing Either One or Two Lanes (20.5 to 23.5 ft) 6.0 sec Raff/Hart: 5.35 sec Logit 50th: 4.34 sec Logit 85th: 6.51 sec Crossing Two Lanes (24 to 27 ft) Raff/Hart: 5.70 sec Logit 50th: 4.79 sec Logit 85th: 7.26 sec

135 Table 79. Values for time to clear. Variable Conditions Harmelink Average (Average + Standard Deviation) Values for: Current Study (Subdivided by Number of Lanes) Current Study (Grouped by Typical Crossing Width) Clear Approach Lane, Average Overall 1.9 sec, based on 150 vehicles 1.1 (1.6) sec, based on 2945 vehicles Clear Opposing Lane(s) Two-lane highway or 11- to 12-ft crossing width 3.0 sec 2.5 (3.6) sec, based on 1181 vehicles 2.1 (2.6) sec, based on 277 vehicles and crossing width of 11 or 12 ft Four-lane highway or 22- to 27-ft crossing width 4.0 sec 2.7 (3.5) sec, based on 1764 vehicles 2.8 (3.6) sec, based on 1792 vehicles and crossing width of 22 to 27 ft Mixed No value given 2.6 (3.4) sec, based on 2945 vehicles 2.4 (3.1) sec, based on 1792 vehicles and crossing width of 17 to 21 ft Turning Time Equation Based on Field Study Findings Overall 4.9 to 5.9 sec TT = 8.500324 + 0.017133 × CW2 – 0.511235 × CW + 0.079616 × log(LGT) + 0.376136 × log(TAH) + 0.115794 × log(TIQ) – 0.045157 × PSL – 0.525688 × NOL2 Where: TT = turning time (sec), LGT = accepted lag or gap time (sec), TAH = time at the head of the queue (sec), TIQ = time in the queue (sec), CW = crossing width (ft), CW_SQ = square of the crossing width (ft), PSL = posted speed limit (mph), and NOL2 = include indicator variable when number of lanes = 2. Critical Gap Several recent research projects have determined the critical gap for use in intersection sight distance calculations and unsignalized intersection capacity analysis. As reported by Harwood et al. (89), Kyte et al. recommended a critical gap value of 4.2 sec for left turns from the major road

136 by passenger cars for inclusion in the unsignalized intersection analysis procedures of the Highway Capacity Manual (90). A heavy-vehicle adjustment of 1.0 sec for two-lane highways and 2.0 sec for multilane highways was also recommended. It is reasonable that design policies should be more conservative than operational criteria. Using a higher critical gap value accepted by 85 percent of the drivers, rather than the gap accepted by only 50 percent of the drivers, should result in a more conservative, design-oriented approach. With that philosophy, the authors of the 1996 intersection sight distance guidelines recommended a 5.5-sec gap value for use in intersection sight distance (89). This gap value is to be increased to 6.5 sec for single-unit trucks and 7.5 sec for combination trucks. Also, an additional 0.5 sec for cars and 0.7 sec for trucks should be added when crossing an additional opposing lane. Harmelink’s assumptions of 5.0 sec and 6.0 for critical gap on two-lane and four-lane highways, respectively, are less than or similar to the average values identified in this research. If a more conservative gap value for use in design is desired, then the critical gap value should be increased to the following: • 6.25 to 6.50 sec for two-lane highways and • 6.50 to 7.25 sec for four-lane highways. Time to Clear This research project found differences in the time to clear the approach lane and the time to clear the opposing lane as compared to Harmelink’s assumptions (see Table 79). Both the average time to clear the approach lane and the average time to clear the opposing lane were smaller (or faster). The average plus standard deviation value was also smaller than Harmelink’s assumptions in most situations. The only condition when the average plus standard deviation value was larger was for the two-lane highway sites when using the data for all two-lane highway sites (3.6 sec as compared to 3.0 sec). Changes to Warrants The left-turn lane warrants would shift downward (i.e., left-turn lanes would be warranted at lower volumes) for two-lane highways if the following are assumed: • Critical gap of 6.25 sec, • Time to clear the advancing lane of 1.6 sec, • Time to clear the opposing lane of 3.6 sec, and • No change to the assumed probability of a through vehicle arriving behind a stopped left- turning vehicle. Figure 38 illustrates the change from the Green Book warrants to the warrants that would be generated based on the field study findings for a 60-mph two-lane highway.

137 Figure 38. Example of change in Green Book (GB) left-turn lane warrants if findings from field study (Field) are used in the Harmelink procedure for a 60-mph two-lane highway. BENEFIT-COST RATIO Economic criteria for left-turn lanes were established by determining the level of peak-hour major-road volume and peak-hour conflicting left-turn volume that result in a B/C ratio equal to 1.0 and 2.0. A range of crash costs and a range of construction costs were evaluated. The research team recommends that the mid-range crash cost and the moderate construction cost identified as part of this research be considered in developing the final left-turn lane warrant recommendations. A benefit-cost range of 1.0 to 2.0 can also provide consideration of potential variability in the assumptions. A benefit-cost of 1.0 represents the point when the calculations find benefits outweigh costs. Using a ratio of 2.0 may represent a more practical application of the benefit-cost evaluation. The left-turn lane warrants developed using the moderate construction cost and mid-range crash cost along with a benefit-cost of 1.0 and 2.0 are listed in Table 73 and also shown in Figure 35 and Figure 36. For rural two-lane highways, a left-turn lane is warranted for as low as five left-turning vehicles in 1 hour when opposed by 50 major-road vehicles per hour. For rural four-lane highways, a left- turn lane is warranted when five left-turn vehicles oppose 50 veh/hr/ln for a four-leg intersection or 75 veh/hr/ln for a three-leg intersection. For urban four-leg intersections, a left-turn lane is to be considered when turning across 50 veh/hr/ln in the peak hour. Urban three-leg intersections had the biggest range of warrant recommendations as shown in Figure 35. The suggested warrants are: • Use B/C = 1.0 for urban and suburban arterials (see Table 80 and Figure 39). • Use B/C = 1.0 for rural two-lane highways to warrant a bypass lane (see Table 81 and Figure 40). 0 50 100 150 200 250 300 350 400 0 200 400 600 800 1000 A dv an ci ng V ol um e (v eh /h r) Opposing Volume (veh/hr) 60 mph GB, 10% LT GB, 20% LT GB, 30% LT Field, 10% LT Field, 20% LT Field, 30% LT

138 • Use the results based on B/C = 2.0 for rural two-lane highways to warrant a left-turn lane (see Table 81 and Figure 40). • Use B/C = 1.0 for rural four-lane highways to warrant a left-turn lane (see Table 82 and Figure 41). Table 80. Suggested left-turn lane warrants based on results from benefit-cost evaluations for urban and suburban arterials. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 450 50 10 300 50 15 250 50 20 200 50 25 200 50 30 150 50 35 150 50 40 150 50 45 150 < 50 50 or More 100 < 50 (a) Three Legs (b) Four Legs Figure 39. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on urban and suburban arterials. 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r/l n) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Three Legs Left-turn lane warranted Left-turn lane not warranted 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Four Legs Left-turn lane warranted Left-turn lane not warranted

139 Table 81. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for rural two-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 50 200 50 150 10 50 100 < 50 50 15 < 50 100 < 50 50 20 < 50 50 < 50 < 50 25 < 50 50 < 50 < 50 30 < 50 50 < 50 < 50 35 < 50 50 < 50 < 50 40 < 50 50 < 50 < 50 45 < 50 50 < 50 < 50 50 or More < 50 50 < 50 < 50 (a) Three Legs (b) Four Legs Figure 40. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for intersections on rural two-lane highways. Bypass lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Two Lanes on Major Left-turn treatment not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Two Lanes on Major Left-turn treatment not warranted Bypass lane warranted Bypass lane warranted Left-turn lane warranted

140 Table 82. Suggested left-turn lane warrants based on results from benefit-cost evaluations for rural four-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 75 50 10 75 25 15 50 25 20 50 25 25 50 < 25 30 50 < 25 35 50 < 25 40 50 < 25 45 50 < 25 50 or More 50 < 25 (a) Three Legs (b) Four Legs Figure 41. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on rural four-lane highways. Figure 42 shows the suggested LTL installation guidelines from this research as compared to the guidelines from the departments of transportation (DOTs) of Georgia (GA), New Mexico (NM), and South Dakota (SD). Figure 43 is a similar graph that shows the suggested left-turn bypass lane guidelines from this research as compared to the guidelines from the Georgia DOT. These guidelines are for rural two-lane highways with speeds of 55 mph or greater. Locations on the figure that are above and/or to the right of a particular threshold are conditions in which a left- turn lane, bypass lane, or deceleration lane are warranted. The figures show that the suggested guidelines from this research provide for the installation of a left-turn lane for lower volumes of opposing traffic than the DOT guidelines in most cases. New Mexico has the lowest left-turn volume threshold (5 veh/hr) among the states, for the installation 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted

141 of a deceleration lane; this threshold largely agrees with that of Georgia for a right-hand bypass lane. For combinations of very low opposing volumes and turning volumes, these two states have more permissive guidelines than those developed in this research; for other conditions, this research provides the lowest installation thresholds. Comparisons of guidelines for rural four- lane highways and for suburban/urban arterials produced similar results. Figure 42. Comparison of suggested left-turn lane warrants for rural two-lane highways. 0 5 10 15 20 25 30 35 40 45 50 0 100 200 300 400 500 600 700 Le ft- Tu rn V ol um e (v eh /h r) Opposing Volume (veh/hr) Rural, Two Lanes on Major 3-91 LTL - 3 Legs 3-91 LTL - 4 Legs Green Book (60) GA - LTL (55) NM - LT Decel (55) SD - LTL (>55)

142 Figure 43. Comparison of suggested bypass lane warrants for rural two-lane highways. MINIMUM VOLUME/ENGINEERING JUDGMENT Minimum volume criteria are considered in different criteria recommended by researchers and used in some states. For example, NCHRP Report 348 criteria by Koepke and Levinson provide two methods for determining the need for left-turn lanes (15). The first method is shown in Figure 6; a left-turn lane is warranted if more than 30 vehicles are turning left for 30- to 35-mph roadways (25 vehicles for 40- to 45-mph roadways). However, Koepke and Levinson state that in most cases, left-turn lanes should be provided where there are more than 12 left turns per peak hour. The benefit-cost ratio identified scenarios for rural highways when a left-turn lane is justified with as few as 5 veh/hr turning left when turning across as few as 50 veh/hr/ln. The research team recommends that the values in Table 80 and Table 81 be used rather than a minimum volume. The characteristics of the field study sites along with a selection of the case study sites included in the Design Guide (4) were used in the benefit-cost and Harmelink procedures. The results were reviewed to identify whether they generated logical and appropriate results. In the engineering judgment of the research team, the results from the benefit-cost approach are reasonable. 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Opposing Volume (veh/hr) Rural, Two Lanes on Major 3-91 Bypass - 3 Legs 3-91 Bypass - 4 Legs GA - Bypass (55)

143 LEFT-TURN LANE STORAGE The left-turn lane should be sufficiently long to store the number of vehicles likely to accumulate during a critical period; the definition of that critical period can vary depending on the traffic conditions at the site. Regardless of the specific critical period, the storage length should be sufficient to avoid the possibility of the left-turning queue spilling over into the through lane. According to the Green Book (5), at unsignalized intersections, the storage length, exclusive of taper, may be based on the number of turning vehicles likely to arrive in an average 2-minute period within the peak hour. Space for at least two passenger cars should be provided; with over 10 percent truck traffic, provisions should be made for at least one car and one truck. Table 83 shows the recommended spacing by percent truck included in the TRB Access Management Manual (91). Table 83. Queue storage length per vehicle (91). Trucks (Percent) Assumed Queue Storage Length (ft) per Vehicle in Queue ≤ 5 25 10 30 15 35 The 2-minute waiting time suggested in the Green Book may need to be changed to some other interval that depends largely on the opportunities for completing the left-turn maneuver. These intervals, in turn, depend on the volume of opposing traffic, which the Green Book does not address. For additional information on storage length, the Green Book refers the reader to the Highway Capacity Manual (92). The first equation shown in Table 84 can be used to determine the design length for left-turn storage as described by the Green Book. Many states use the Green Book method, a method based on work done by Harmelink (1) (see Figure 44), or a method based on work by Jack E. Leisch and Associates (93) (see Figure 45) to describe their recommended storage lengths in their design guidelines. Others recommend that the designer assume that the intersection is signalized with a two-phase signal using a 40- to 60-sec cycle length, and then use the Highway Capacity Manual methodology to determine the expected storage length. The authors of NCHRP Report 348 (15) state that the required storage length of a left-turn lane depends upon the likely left-turn volumes during the peak 15 minutes of the design hour, which is typically but not always the morning or evening peak hour. The length for a stop-controlled lane should be adequate 95 percent of the time and can be estimated by using the cumulative Poisson distribution.

144 Table 84. Equations used to determine storage length. Equation in TRB Access Management Manual ܮ ൌ ௏ே೎ ݇ݏ (72) Where: L = design length for left-turn storage (ft); V = estimated left-turn volume, vehicles per hour (veh/hr); Nc = number of cycles per hour (for the Green Book unsignalized procedure, this would be 30 [V/N is the average number of turning vehicles per cycle]); k = factor that is the length of the longest queue (design queue length) divided by the average queue length (a value of 2.0 is commonly used for major arterials, and a value of 1.5 to 1.8 might be considered for an approach on a minor street or on a collector where capacity will not be critical) (for the Green Book procedure this would be 1.0); and s = average length per vehicle, including the space between vehicles, generally assumed to be 25 ft (adjustments are available in several documents for trucks and buses, such as the TRB Access Management Manual [see Table 83]). Equations Used in NCHRP Report 457 Equations also used to generate values in Table 85 ܲሺ݊ ൐ ܰሻ ൌ ቀ௩௖ቁ ሺேାଵሻ (73) ܿ ൌ ௏೚௘షೇ೚೟೎/యలబబଵି௘షೇ೚೟೑/యలబబ (74) ܰ ൌ ௟௡ሾ௉ሺ௡வேሻሿ௟௡ہ௩/௖ۂ െ 1 (75) ܵܮ ൌ ܰ ൈ ܸܮ ൌ ቄ௟௡ሾ௉ሺ௡வேሻሿ௟௡ሾ௩/௖ሿ െ 1ቅ ൈ ܸܮ (76) Where: P(n>N) = probability of bay overflow; v = left-turn vehicle volume (veh/hr); N = number of vehicle storage positions; c = movement capacity (veh/hr); Vo = major-road volume conflicting with the minor movement, assumed to be equal to one-half of the two-way major-road volume (veh/hr); SL = storage length (ft); tc = critical gap (sec); tf = follow-up gap (sec); and VL = average length per vehicle, including the space between vehicles, generally assumed to be 25 ft (adjustments are available in several documents for trucks and buses such as the TRB Access Management Manual [see Table 83]).

145 Table 85. Recommended storage lengths from Access Management Manual equation and NCHRP Report 457 equations with revised critical gap. Left-Turn Volume (veh/hr) Storage Length, Rounded Up to Nearest 25-ft Increment (ft) Storage Lengths from Other Manuals for Comparison Storage Lengths Calculated from Equations b Documented in NCHRP Report 457 Using Revised Critical Gaps and 0.005 Probability of Overflow Green Book Procedure (k = 1)a Equation (k = 2)a Opposing Volume (veh/hr) 200 400 600 800 1000 Critical Gap = 5.0 sec, Follow-Up Gap = 2.2 sec (Represents the 50th Percentile Critical Gap Found in Field Studies) 40 75 75 50 50 50 50 50 60 50 100 50 50 50 50 50 80 75 150 50 50 50 50 50 100 100 175 50 50 50 50 75 120 100 200 50 50 50 75 75 140 125 250 50 50 50 75 75 160 150 275 50 50 75 75 100 180 150 300 50 50 75 75 100 200 175 350 50 75 75 100 125 220 200 375 50 75 75 100 125 240 200 400 75 75 100 125 150 260 225 450 75 75 100 125 175 280 250 475 75 75 100 125 175 300 250 500 75 100 125 150 200 Critical Gap = 6.25 sec, Follow-Up Gap = 2.2 sec (Represents the 85th Percentile Critical Gap Found in Field Studies, 85th Percentile Is Preferred for Design) 40 75 75 50 50 50 50 50 60 50 100 50 50 50 50 50 80 75 150 50 50 50 50 75 100 100 175 50 50 50 75 75 120 100 200 50 50 75 75 100 140 125 250 50 50 75 100 125 160 150 275 50 75 75 100 150 180 150 300 50 75 75 125 150 200 175 350 50 75 100 125 200 220 200 375 75 75 100 150 225 240 200 400 75 75 125 150 275 260 225 450 75 100 125 175 325 280 250 475 75 100 125 200 400 300 250 500 75 100 150 225 525 a, b See Table 84 for equations. This table assumes 25 ft per vehicle spacing. Table 83 provides other suggested spacing lengths based on percent trucks.

146 Source: Harmelink, M., “Volume Warrants for Left-Turn Storage Lanes at Unsignalized Grade Intersections,” in Highway Research Record 211, Figure 1, p. 8. Copyright, National Academy of Sciences, Washington, D.C., 1967. Reproduced with permission of the Transportation Research Board. Figure 44. Storage length recommendations based on work by Harmelink (1).

147 Source: TRB Committee on Access Management, Access Management Manual, Figure 10-7, p. 174. Copyright, National Academy of Sciences, Washington, D.C., 2003. Reproduced with permission of the Transportation Research Board. Figure 45. Storage length recommendations based on work by Leisch (93) as presented in the TRB Access Management Manual (91).

148 NCHRP Report 457 (12) developed suggested storage length values using the equations shown in Table 84. The NCHRP Report 457 procedure was similar to Harmelink’s work regarding storage length of left-turn bays at unsignalized intersections. The storage length equation is a function of movement capacity, which is dependent upon assumed critical gap and follow-up gap. Critical gap is defined by the Highway Capacity Manual as the minimum time interval in the major-street traffic stream that allows intersection entry for one minor-street vehicle. Thus, the driver’s critical gap is the minimum gap that would be acceptable. The time between the departure of one vehicle from the minor street and the departure of the next vehicle using the same major-street gap, under a condition of continuous queuing on the minor street, is called the follow-up time. NCHRP Report 457 used a smaller critical gap (4.1 sec as recommended in the Highway Capacity Manual compared to the 5.0 or 6.0 sec used by Harmelink for two-lane and four-lane highways, respectively), which resulted in shorter values than those generated by Harmelink. The follow-up gap was assumed to be 2.2 sec as recommended in the Highway Capacity Manual. The assumptions made regarding critical gap and the resulting capacity for the movement used in these procedures can have a significant effect on the calculated storage length recommendations as demonstrated by several researchers (94, 62, 12). It is generally recognized that a storage area should adequately store the turn demand a large percentage of the time (e.g., 95 percent or more, which means that the demand would exceed the storage length less than or equal to 5 percent of the time). A 0.5 percent limit was used for the major-road left-turn bay lengths in NCHRP Report 457 based on the recommendation of Harmelink. This smaller limit reflects the greater potential for severe consequences when a bay overflows on an unstopped, major-road approach. The critical and follow-up gaps were assumed to equal 4.1 and 2.2 sec, respectively. Figure 46 shows the graphical representation of the storage length guidelines presented in NCHRP Report 457; the Internet version of that document also provides an interactive spreadsheet tool in which the designer can input the specific volume and gap variables to receive a recommended storage length for those conditions. NCHRP Report 457 assumed a 25-ft minimum storage length. Harmelink used larger values for critical gap (5.0 sec for two-lane highways and 6.0 sec for four- lane highways). When those gaps are used within the approach presented by Bonneson and Fontaine in NCHRP Report 457, storage lengths similar to those suggested by Harmelink are obtained. When the critical gap of 5.0 and 6.25 sec determined in NCHRP Project 3-91’s field studies are used, the storage lengths shown in Table 85 are generated. Each of the sources emphasizes that the appropriate storage length is dependent on both the volume of turning traffic and the volume of opposing traffic. If volume data are not available, for urban and suburban streets with lower speeds (e.g., less than 40 mph), it is recommended that the minimum storage length be at least 50 ft to accommodate two cars; for high-speed and rural locations, a minimum storage length of 100 ft is recommended.

149 Source: Bonneson, J., and M. Fontaine, Engineering Study Guide for Evaluating Intersection Improvements, NCHRP Report 457, Figure 2-7, p. 24. Copyright, National Academy of Sciences, Washington, D.C., 2001. Reproduced with permission of the author. Figure 46. Recommended storage lengths for left-turn lanes at uncontrolled approaches using a 25-ft minimum storage length along with a critical gap of 4.1 sec (12). 0 20 40 60 80 100 120 140 160 0 50 100 150 200 250 300 S to ra ge L en gt h (ft ) Turn Movement Volume (veh/h) Adequate 99.5% of the time. Conf licting volume =1050 veh/h 850 650 450 250

151 CHAPTER 7 SUMMARY AND CONCLUSIONS SUMMARY Left-turn movements at intersections, including driveways—especially movements that are made from lanes that are shared with through traffic—cause delays and adversely impact safety. The left-turn problem has been given relatively little attention at unsignalized intersections. Although updated warrants have been developed in some jurisdictions for when to provide left-turn lanes, many agencies still use research from the mid-1960s. Current conditions require a broader assessment of when to provide left-turn accommodations. Technical warrants are an important element of the decision-making process. Objective The key objectives for this NCHRP project were to: • Develop an objective and clear process for the selection of left-turn accommodations at unsignalized intersections and • Provide guidance on the design of these accommodations. The findings presented in the sections below provide the recommended warrants for left-turn lanes based on the research conducted in this project. A companion document, the Design Guide on Left-Turn Accommodations at Unsignalized Intersections (4), provides guidance on the design of left-turn lanes. Other objectives for the project included: • Review left-turn lane installation guidelines available in the literature or state manuals, • Review the design guidance provided in the literature or state manuals, • Conduct a legal review, • Conduct interviews to determine the state of the practice regarding left-turn lane installations at unsignalized intersections, • Compare different methods used or suggested for determining when to install a left-turn lane, • Update the assumptions used in the Harmelink procedure, and • Generate warrants based on an economic procedure that considers both delay savings and benefits resulting from a reduction in crashes. Literature Review A review of the literature was performed using many sources, including research reports, state and federal design manuals, and handbooks. Although many procedures are currently in use by various organizations to determine the need for left-turn lanes, several are either very similar or identical. Most states’ criteria are based on the Green Book values or the NCHRP Report 279 values—both of which are based on M. Harmelink’s work.

152 Legal Review The legal review addressed the following question: When a government seeks to fulfill a broad public objective such as safety and, in this project, left-turn accommodation, who should bear the costs—the developer who would be adding traffic to the roadway network or the general public? Unless the facts of the case are clear cut, the outcome is unpredictable even within specific jurisdictions. Results will be more predictable if agencies have the documented authority to manage access. This authority could be achieved in a number of different ways (state or municipal codes for access management, administrative rules, etc.). Otherwise, there are too many hazy cases and multiple factors so that predicting the outcome is impossible. Interviews To help investigate the implementation of left-turn accommodations at unsignalized intersections, researchers conducted interviews of representatives from state DOTs, county governments, city governments, and consultants. The 25 questions in the interview were structured into planning, design, legal/policy/finance, and potential future applications. All interview participants indicated that left-turn treatments are provided at unsignalized intersections. In addition to left-turn lanes, all state DOTs, all cities, and a majority of the counties in the survey indicated they use two-way, left-turn lanes. Several participants noted that TWLTLs are typically used in areas where there may be poor access control. Several state DOTs and two counties indicated they consider bypass or shoulder widening, but they noted that these techniques are typically used when there are limited options. One state DOT noted that roundabouts are another treatment considered for dealing with left-turn movements. The state DOT process for determining where left-turn treatments should be installed at unsignalized locations, who is involved, and what affects the decision may vary depending upon the circumstances involved and whether the question relates to: • A developer seeking access onto the roadway system for a new development or major redevelopment, • The state DOT reevaluating the roadway condition as part of an improvement project, • A county- or city-initiated project that is being coordinated with the state DOT, or • A problem location identified by citizen complaints or its crash rate. Two of the state DOTs reflected in the survey had two interview participants. Based on responses from these two state DOTs, it appears practices may vary depending on the area of the state or which DOT unit is involved. A majority of the interview participants at all levels were not aware of the two Supreme Court decisions relating to essential nexus or rough proportionality—Nollan v. California Coastal Commission and Dolan v. City of Tigard. Five state DOT and two county respondents were aware of the Supreme Court decisions. All of the interview participants who expressed an awareness of the Supreme Court decisions indicated they were aware of no ramifications of these court cases to their jurisdictions’ decisions related to left-turn accommodations. Those

153 participants indicated that their policies are consistent with these decisions and already reflected rough proportionality, rational nexus, etc. Comparison of Existing Procedures The use of the Harmelink criteria requires approaching volume, opposing volume, left-turn percentage, and speed of the roadway. Those states that have adopted other criteria generally have fewer required input values; for example, one state only requires the anticipated left-turn volume, one state only requires the anticipated minor-road volume, and another state has criteria based on ADT and left-turning volume by posted speed. In addition to having fewer demands for input values, the newer procedures also result in left-turn lanes being warranted at lower volumes than the criteria currently in the Green Book. While most states use procedures that are based on Harmelink, a number of limitations of Harmelink’s procedure have been identified over the years. Harmelink’s warrants are developed based on allowable maximum probability values that are subjective in nature and do not directly indicate stops, delay, speed change, or level-of-service degradation in any quantitative manner. Harmelink assumed arrival time for both left-turn and through vehicles to be negative exponentially distributed, an assumption valid only for isolated intersections under low demand conditions. Driver Behavior Study This project’s driver behavior study used videotaped recordings of vehicle and pedestrian operations to obtain needed data for calibrating the simulation model and updating the Harmelink procedure. Left-turn movements were studied at 30 sites that were located in College Station/Bryan, Texas; Houston, Texas; Staten Island, New York; and Phoenix, Arizona. The sites were selected based on a variety of intersection arrangements and geometric characteristics, including: • Number of lanes on the major road—two or four lanes; • Presence of a left-turn lane—yes or no; • Signal coordination—location is near enough to a signal to be affected or far enough from a signal to result in random arrival; and • Approach speed range—low or high speed, with posted speed limits between 25 and 40 mph being defined as low speed and posted speed limits of 45 mph or more being defined as high speed. The reduction process began with reviewing the site video and obtaining turning movement counts using 5-minute intervals. The goal for each site was to obtain data for a minimum of 100 left-turning vehicles whose drivers had to make a decision based on the available gaps in the opposing traffic. The behaviors for a total of 3570 vehicles were collected from the field studies. Heavy trucks only represented a small portion of the data collected (39 vehicles), and since their operations are known to be slower than passenger cars, they were excluded from the evaluations. Of the remaining vehicles, 2945 vehicles started from a stopped position. A vehicle was included as starting from a stopped position if the vehicle spent at least 0.25 sec between arriving at the front of the queue and starting the left-turn maneuver.

154 The most influential variables on the amount of time used to clear the intersection are crossing width and posted speed limit. As an example, a change from a crossing width of 11 ft to 27 ft resulted in an additional 2.24 sec in total turning time. The posted speed limit variable is associated with a decrease in turning time for the higher speeds. For turning time, an additional 1.81 sec is subtracted when the roadway has a 65-mph posted speed as compared to a 25-mph posted speed. The posted speed limit range represented in the dataset was 25 to 65 mph. The relationship between clearance time and the accepted lag or gap time was in the direction expected although it did not have as large an influence as initially thought. The preliminary thought was that drivers will notably drive faster (i.e., lower turning time) when accepting a small gap. The evaluation only found a small increase in clearance time for larger gaps. For a 20-sec gap as compared to a 5-sec gap, the increase in turning time was about 0.05 sec. Critical gap is defined as the time interval between two opposing vehicles that is necessary for a left-turning vehicle to safely complete a left-turn maneuver. Two methods were used to determine critical gap: logistic regression and Raff/Hart. Logistic regression is appropriate when the dependent variable is binary or dichotomous (e.g., either the acceptance or rejection of a gap). The relationship between posted speed limits from the field studies was similar to the finding from a simulator study reported in the literature—smaller gaps are accepted at higher posted speed sites. The difference is on the order of 1 to 1.5 sec. Gap acceptance values increase as the crossing width increases, although only by a small amount (less than 1 sec between the one-lane group and the two-lane or very-wide-one-lane group). Updating Harmelink The findings from the field studies were used to update Harmelink assumptions for two-lane rural highways. The time-to-clear values were smaller at the 30 field study sites, and the critical gap was larger, resulting in a shift of the left-turn warrants downward. In other words, left-turn lanes would be warranted at lower volumes. Because of several concerns with the Harmelink procedure, including the lack of a clear relationship between the assumptions in the model and delay or safety on a highway, the research team recommends that the results from the benefit- cost ratio be used as the basis for left-turn lane warrants. Updating Harmelink Storage Lengths Harmelink also provided left-turn lane storage lengths in his research. Storage lengths were developed in this project using the method presented in NCHRP Report 457 along with critical gap values found in this field study (5.0 sec and 6.25 sec, representing 50th and 85th percentiles, respectively). These lengths were compared to lengths generated using other methods, such as the method discussed in the Green Book. Economic Analysis for Existing Sites Economic analysis can provide a useful method for combining traffic operations and safety benefits of left-turn lanes to identify situations in which left-turn lanes are and are not justified economically. A benefit-cost approach was used in this project to determine when a left-turn lane would be justified. The steps included simulation to determine delay savings from installing a

155 left-turn lane, crash costs, crash reduction savings determined from safety performance functions and accident modification factors available in the Highway Safety Manual, and construction costs. The comparison identifies the benefits when the left turns at an existing driveway or intersection are provided a left-turn lane. The total average delay for when a left-turn lane is present is subtracted from the total average delay when a left-turn lane is not present. This difference represents the total average delay savings per vehicle at the intersection on the major roadway. The estimated reduction in delay provided by a left-turn lane on a two-lane highway ranged from 0 to 4 sec/veh when the major- road volume ranged from 400 to 800 veh/hr/ln. The delay reduction for four-lane highways generally ranged between 0 and 2 sec/veh for volumes up to 800 veh/hr/ln on the major road and left-turning vehicle volume of 100 veh/hr or less. Delays were much higher when the left-turning volume was 140 veh/hr or greater and the major-road volume was 600 veh/hr/ln or greater. The predicted average crash frequency for an intersection can be determined from equations in the Highway Safety Manual. These equations, called safety performance functions, are regression models for estimating the predicted average crash frequency of individual roadway segments or intersections for a set of specific base conditions. As discussed in the HSM, each SPF in the predictive method was developed with observed crash data for a set of similar sites. The SPFs, like all regression models, estimate the mean value of a dependent variable as a function of a set of independent variables. In the SPFs developed for the HSM, the dependent variable is the crash frequency for a roadway segment or intersection under base conditions, and the independent variables are the AADTs of the roadway segment or intersection legs (and, for roadway segments, the length of the roadway segment). For rural conditions, different SPFs are provided for two- and four-lane highways and for three- and four-leg intersections. For urban and suburban arterials, prediction equations are provided for three-leg and four-leg intersections. Separate urban and suburban prediction equations are not provided based on the number of lanes on the major-road approach. The accident modification factor for left-turn lanes is available from the Highway Safety Manual. For this evaluation, the intersections were assumed to have a stop sign on the minor approaches and that only one of the major-road approaches would be treated with a left-turn lane. The AMFs for both the rural two-lane and four-lane highway scenarios are 0.56 for three-leg intersections and 0.72 for four-leg intersections. The AMFs for the urban and suburban scenarios are 0.67 for three-leg intersections and 0.73 for four-leg intersections. In 2008, a memo was released by the U.S. Department of Transportation regarding the treatment of the economic value of a statistical life in developmental analyses. The memo “raises to $5.8 million the value of a statistical life to be used by analysts in the Department of Transportation when assessing the benefit of preventing fatalities.” The cost per crash requires knowing the distribution of crash severity, which was available in the Highway Safety Manual. Also needed is a conversion of the cost per person to a cost per crash. The number of individuals killed or injured in a crash is not readily available; however, values found in the literature or other projects were reasonable for use in this evaluation. The Highway Safety Manual also provides crash cost estimates by crash severity. These values were converted to 2009 dollars, and

156 then the typical crash costs by number of legs and rural or urban area were determined. The HSM values represent costs per crash, while the value of a statistical life represents cost per person and had to be converted to cost per crash. The crash cost values in the Highway Safety Manual are assumed to already have accounted for the number of persons typically involved in a crash along with the distribution of injuries within a crash. The HSM values, however, do not account for the higher value being placed on a statistical life from the 2008 memo. A comparison is made between the two approaches as part of this project. The typical crash cost per number of legs and rural or urban roadway based on the 2008 U.S. DOT memo, which states that the economic value of a statistical life is $5.8 million, is: • $214,000 for three-leg rural roadways, • $198,000 for four-leg rural roadways, • $167,000 for three-leg urban and suburban roadways, and • $180,000 for four-leg urban and suburban roadways. The typical crash cost per number of legs and rural or urban roadway based on HSM data is: • $129,000 for three-leg rural roadways, • $135,000 for four-leg rural roadways, • $113,000 for three-leg urban and suburban roadways, and • $121,000 for four-leg urban and suburban roadways. Typical construction costs for left-turn lanes were identified from several sources including previous literature and state department of transportation websites. A reasonable range for the cost of constructing a left-turn lane appears to be $100,000 to $375,000, with an average value of $250,000. The above values were used in the benefit-cost evaluation to identify volume conditions when the installation of a left-turn lane at unsignalized intersections and major driveways would be cost-effective. Plots and tables were developed that indicate combinations of major-road traffic and left-turn lane volume where a left-turn lane would be recommended. Because the Highway Safety Manual prediction equations are not a function of operating speed or speed limit, the warrants are not separated by speed. The warrants are a function of the number of legs at the intersection along with whether it is in a rural or urban/suburban area. Warrants were developed using the following: • A range of values for the economic value of a statistical life, • Crash costs based on values in the Highway Safety Manual, • A range of construction costs, and • Benefit-cost ratio of 1.0 and 2.0. The research team suggests a benefit-cost ratio of 1.0 along with the mid-range economic value of a statistical life and moderate construction cost to identify the warrants for a minimum left- turn treatment. For urban and suburban areas and multilane rural highways, that is a left-turn lane. For two-lane rural highways, that is a bypass lane. The benefit-cost ratio of 2.0 has been argued as being a more practical value to offset the potential variability in other assumptions. The warrants based on a benefit-cost ratio of 2.0 were selected for a left-turn lane on rural two-

157 lane highways. These values were similar to the warrants that resulted when the lower crash costs based on older Highway Safety Manual values were used. Economic Analysis for New Sites There is an inherent difference between adding a left-turn lane at an existing site and constructing a new left-turn lane at a proposed development. While private property enjoys the right of access to the general system of public roadways, this is not an unlimited right. The right of access must be balanced with the needs of and potential harm to the general traveling public. In order to preserve mobility and provide safety for the traveling public, many transportation agencies have established regulations and programs to manage access to their roadway network. Agencies may require that steps be taken by a developer to mitigate projected traffic operations and/or safety impacts. An example of mitigation would be providing a left-turn lane to remove the traffic turning left into the site from the through arterial lanes. Many transportation agencies have the authority to require a developer to pay for this mitigation as long as there is a rational nexus between the projected impacts of the development and the needed improvements. In this manner, taxpayers do not have to pay for an improvement that may benefit predominantly one property owner. The findings from this study can be used to estimate the additional costs to a site if a new development causes left-turn volumes. The information available from this project reflects certain conditions, such as no cross minor-road volumes for a four-leg intersection. Simulation for a specific site would be needed to determine the expected increase in delay for the given intersection characteristics. The Highway Safety Manual can provide information regarding crashes. CONCLUSIONS Left-turn lanes can reduce the potential for collisions and improve capacity by removing stopped vehicles from the main travel lane. The conclusions and recommendations developed based on this research are: • Left-turn lane warrants were developed using an economic analysis procedure for: o Rural two-lane highways (see Table 81and Figure 40), o Rural four-lane highways (see Table 82 and Figure 41), and o Urban and suburban roadways (see Table 80 and Figure 39). • The methodology presented in this report could also be used if a transportation agency has available local values for delay reductions due to the installation of a left-turn lane, crash frequency or crash predictions, crash reduction factors, crash costs, and/or construction costs. • Methodology is also presented in this report that can be used to estimate the impacts due to a new development. Impacts discussed include increased delay due to the new development along with predicted safety consequences. The example provided is for conditions assumed to generate the delay value used in this project. Simulation for a specific site would be needed to determine the expected increase in delay for the given intersection characteristics. The Highway Safety Manual provides safety performance functions for predicting the number of crashes. • A legal review conducted as part of this research addressed the following question: When a government seeks to fulfill a broad public objective such as safety and, in this project,

158 left-turn accommodation, who should bear the costs—the developer who would be adding traffic to the roadway network or the general public? Unless the facts of the case are clear cut, the outcome is unpredictable even within specific jurisdictions. Results will be more predictable if agencies have the documented authority to manage access.

A-1 APPENDIX A REVISED TEXT ON LEFT-TURN LANE WARRANTS FOR THE AASHTO GREEN BOOK This appendix presents revised text on left-turn lane warrants for consideration by AASHTO for inclusion in the AASHTO A Policy on Geometric Design of Highways and Streets. Material suggested for removal is shown with a strikethrough line. Double underlines indicate material to be added. Figures or tables beginning with the number 9 are from the AASHTO Green Book, while those beginning with the letter A reflect recommended additions to the text. AASHTO GREEN BOOK (2004), RURAL AND URBAN ARTERIALS (URBAN), PAGES 488 TO 490 Operational and Control Measures for Left-Turn Maneuvers Vehicles turning left into cross streets or all mid-block locations may cause substantial delays to through traffic and may contribute to crashes, thus diminishing arterial effectiveness. There is a popular belief that the effects of such left-turn movements can be eliminated simply by placing “No Left Turn” signs. In fact, motorists that desire to turn left do not just disappear, but reach their destinations by alternative routes. Thus, prohibition of left turns at some locations may create or increase operational or safety problems at other locations. Effective control of turning movements lies in discovering or anticipating the extent of the problem and in providing for the movements through a combination of measures including selective prohibition or turns, geometric design, and traffic control. It is difficult to discuss these factors independently, and no firm rules are applicable to all situations. Several principles and methods that, if properly considered and applied, will lead to appropriate designs are outlined as follows: 1 The capability for motorists to reach their desired destinations must be provided. Left turns should not be prohibited unless alternative routings are available. 2 As a general rule, the fewer the number of left turns at any location, the less the interference with other traffic. Thus, for a given total number of left turns within a given length of highway, it may be better to encourage a few left turns at each of several locations than to concentrate the turns at a single location. 3 Separate signal phases for left-turn movements reduce the amount of green time available for other movements at the intersection. Multiphase signals are therefore advantageous only if traffic operation and safety are improved sufficiently to offset the loss in green time. This determination should be made on a case-by-case basis. 4 Where selective prohibition of left turns is necessary, there are operational advantages in concentrating left turns at intersections where the volume of cross traffic is low so that a

A-2 large fraction of the signal time is available for the green phase on the arterial. Where two arterial streets intersect, there may be advantage in requiring left-turning vehicles to bypass the main intersection. For instance, one manner in which the left-turn maneuver from one arterial to another can be accomplished is to require the motorist to turn left from the first arterial a block in advance of the main intersection, then proceed one block, turn right, proceed another block, and turn left. Where such techniques are used, clear guide signing is essential. 5 It is sometimes advantageous to route left-turning traffic around a block, through a series of right turns after passing through the main intersection, rather than permitting a direct left-turn maneuver. However, this approach has disadvantages as well. Traffic volumes are increased because the left-turning vehicle now must pass through the intersection twice. In addition, the distance of travel by the vehicle that desires to turn left is increased, and the increased right-turn volumes may have an impact on the operation of three other intersections. This approach to left-turn maneuvers should generally be limited to locations where the left-turn volumes are small and the provision of a separate left-turn lane is not practical. 6 The desirability of exclusive left-turn lanes cannot be overemphasized. Such lanes may consist of separate left-turn lanes in the median or continuous center lanes used exclusively for left turns from both directions. Economic analysis was used to identify situations for an unsignalized intersection in which left-turn lanes are and are not justified based on delay savings and crash reductions. The warrants are presented in the Intersection Chapter. Multiphase signal control is very inefficient if turning traffic and through traffic both use the same lane. Where turning traffic is light, a left-turn lane may eliminate the need for the left-turn signal phasing because the storage of left-turning vehicles will not affect through traffic. Traffic safety is greatly enhanced if turning vehicles can be stored separately from lanes used by through vehicles. 7 Multiphase signal control is very inefficient if turning traffic and through traffic both use the same lane. Where turning traffic is light, a left-turn lane may eliminate the need for the left-turn signal phasing because the storage of left-turning vehicles will not affect through traffic. Traffic safety is greatly enhanced if turning vehicles can be stored separately from lanes used by through vehicles. 78 With a separate left-turn phase, dual left-turn lanes can accommodate up to about 180 percent of the volume that can be served by a single left-turn lane with the same available green time, depending on the width of the cross street and the radius of turn. Desirably, the turning radius for a dual left-turn lane is 27 m (90 ft). Thus, where sufficient right-of-way space for a long-radius turn, and a wide cross street are available, the installation of dual left-turn lanes may be a practical design to serve a heavy left-turn movement. Exhibit 7-15 shows an example of dual left-turn lanes at an intersection on an urban arterial. Further guidance concerning the design of dual left-turn lanes is presented in Chapter 9 and in the HCM (2).

A-3 89 Grade separations or other special treatments for left-turn movements are sometimes appropriate, as discussed in Chapter 10. In summary, left-turn demands should be accommodated as near as practical to the point of which the motorist desires to turn left. Shifting the left-turn maneuvers away from this point of desire may lead to secondary problems. Nevertheless, if the point at which motorists desire to turn left is highly objectionable from the standpoint of design, traffic control, or safety, regulatory measures may be employed to move those left turns to a location that is more suitable. Only in exceptional cases should such maneuvers be shifted more than two blocks from the point of desire. Where left turns are permitted from an arterial street, the intersection design should incorporate left-turn storage lanes unless it is impractical to provide them. AASHTO GREEN BOOK (2004), INTERSECTION CHAPTER, PAGES 488 TO 490 GENERAL INTERSECTION TYPES General Design Considerations General types of intersections and terminology are indicated in Exhibits 9-73 and 9-74. The geometric forms are the three-leg, four-leg, and multileg intersections. Further classification includes such variations as unchannelized, flared, and channelized intersections. Details and specific adaptations of each general type are demonstrated in the section of this chapter on “Types and Examples of Intersections.” Many factors enter into the choice of type of intersection and the extent of design of a given type, but the principal controls are the design-hour traffic volume, the character or composition of traffic, and the design speed. The character of traffic and design speed affect many details of design, but in choosing the type of intersection they are not as significant as the traffic volume. Of particular significance are the actual and relative volumes of traffic involved in various turning and through movements. When designing an intersection, left-turning traffic should be removed from the through lanes, whenever practical. Therefore, provisions for left turns (i.e., left-turn lanes) have widespread application. Ideally, left-turn lanes should be provided at driveways and street intersections along major arterial and collector roads wherever left turns are permitted. In some cases or at certain locations, providing for indirect left turns (jughandles, U-turn lanes, and diagonal roadways) may be appropriate to improve safety and preserve capacity. The provision of left-turn lanes has been found to reduce crash rates anywhere from 20 to 65 7 to 48 percent (Highway Safety Manual) (1). Left-turn facilities should be established on roadways where traffic volumes are high enough or safety considerations are sufficient to warrant them. They are often needed to ensure adequate service levels for the intersections and the various turning movements. Guidelines for when left-turn lanes should be provided are set forth in Table A-1 for urban and suburban roadways and Tables A-2 and A-3 for rural highways (and Figure A-1, A-2, and A-3 for urban and suburban roadways, rural two-lane highways, and rural four-lane highways, respectively). Sseveral documents for both signalized and unsignalized intersections provide guidance on left-turn lanes (11, 12, 13, NCHRP 3-91 Design Guide on Left-Turn Accomodations

A-4 at Unsignalized Intersections, FHWA Signalized Intersections: Informational Guide, FHWA-HRT-04-091). These guidelines key the need for left-turn lanes to (a) the number of arterial lanes, (b) design, and operating speeds, (c) left-turn volumes, and (d) opposing traffic volumes. Table A-1. Suggested left-turn lane warrants based on results from benefit-cost evaluations for urban and suburban ar ter ials. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 450 50 10 300 50 15 250 50 20 200 50 25 200 50 30 150 50 35 150 50 40 150 50 45 150 < 50 50 or More 100 < 50 (a) Three Legs (b) Four Legs Figure A-1. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on urban and suburban arterials. 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r/l n) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Three Legs Left-turn lane warranted Left-turn lane not warranted 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Four Legs Left-turn lane warranted Left-turn lane not warranted

A-5 Table A-2. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for rural two-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 50 200 50 150 10 50 100 < 50 50 15 < 50 100 < 50 50 20 < 50 50 < 50 < 50 25 < 50 50 < 50 < 50 30 < 50 50 < 50 < 50 35 < 50 50 < 50 < 50 40 < 50 50 < 50 < 50 45 < 50 50 < 50 < 50 50 or More < 50 50 < 50 < 50 (a) Three Legs (b) Four Legs Figure A-2. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for intersections on rural two-lane highways. Bypass lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Two Lanes on Major Left-turn treatment not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Two Lanes on Major Left-turn treatment not warranted Bypass lane warranted Bypass lane warranted Left-turn lane warranted

A-6 Table A-3. Suggested left-turn lane warrants based on results from benefit-cost evaluations for rural four-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 75 50 10 75 25 15 50 25 20 50 25 25 50 < 25 30 50 < 25 35 50 < 25 40 50 < 25 45 50 < 25 50 or More 50 < 25 (a) Three Legs (b) Four Legs Figure A-3. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on rural four-lane highways. The HCM (6) indicates that exclusive left-turn lanes at signalized intersections should be installed as follows: • Where fully protected, left-turn phasing is to be provided; • Where space permits, left-turn lanes should be considered when left-turn volumes exceed 100 veh/hr (left-turn lanes may be provided for lower volumes as well as on the basis of the judged need and state of local practice, or both); and • Where left-turn volumes exceed 300 veh/hr, a double left-turn lane should be considered. 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted

A-7 Exhibit 9-75 is a guide to traffic volumes where left-turn facilities should be considered on two- lane highways. For the volumes shown, left turns and right turns from the minor street can be equal to, but not greater than, the left turns from the major street. Additional information on left-turn lanes, including their suggested lengths, can be found in published sources (2, 11, 13, NCHRP Project 3-91 Design Guide on Left-Turn Accommodations at Unsignalized Intersections). In the case of double left-turn lanes, a capacity analysis of the intersection should be performed to determine what traffic controls are needed in order for it to function properly. Local conditions and the cost of right-of-way often influence the type of intersection selected as well as many of the design details. Limited sight distance, for example, may make it desirable to control traffic by yield signs, stop signs, or traffic signals when the traffic densities are less than those ordinarily considered appropriate for such control. The alignment and grade of the intersecting roads and the angle of intersection may make it advisable to channelize or use auxiliary pavement areas, regardless of the traffic densities. In general, traffic service, highway design designation, physical conditions, and cost of right-of way are considered jointly in choosing the type of intersection. Metric US Customary Opposing Advancing volume (veh/h) Opposing Advancing Volume (veh/h) volume 5% 10% 20% 30% volume 5% 10% 20% 30% (veh/h) left turns left turns left turns left turns (veh/h) left turns left turns left turns left turns 60-km/h operating speed 40-mph operating speed 800 330 240 180 160 800 330 240 180 160 600 410 305 225 200 600 410 305 225 200 400 510 380 275 245 400 510 380 275 245 200 640 470 350 305 200 640 470 350 305 100 720 515 390 340 100 720 515 390 340 80-km/h operating speed 50-mph operating speed 800 280 210 165 135 800 280 210 165 135 600 350 260 195 170 600 350 260 195 170 400 430 320 240 210 400 430 320 240 210 200 550 400 300 270 200 550 400 300 270 100 615 445 335 295 100 615 445 335 295 100-km/h operating speed 60-mph operating speed 800 230 170 125 115 800 230 170 125 115 600 290 210 160 140 600 290 210 160 140 400 365 270 200 175 400 365 270 200 175 200 450 330 250 215 200 450 330 250 215 100 505 370 275 240 100 505 370 275 240 Exhibit 9-75. Guide for Left-Turn Lanes on Two-Lane Highways (6)

A-8 For the general benefit of through-traffic movements, the number of crossroads, intersecting roads, or intersecting streets should be minimized. Where intersections are closely spaced on a two-way facility, it is seldom practical to provide signals for completely coordinated traffic movements at reasonable speeds in opposing directions on that facility. At the same time the resultant road or street patterns should permit travel on roadways other than the predominant highway without too much inconvenience. Traffic analysis is needed to determine whether the road or street pattern, left open across the predominate highway, is adequate to serve normal traffic plus the traffic diverted from any terminated road or street. The functional classification of the road, the patterns of traffic movement at the intersections, and the volume of traffic on each approach, including pedestrians, during one or more peak periods of the day are indicative of the type of traffic control devices necessary, the roadway widths needed (including auxiliary lanes), and, where applicable, the degree of channelization needed to expedite the movement of all traffic. The differing arrangement of islands and the shape and length of auxiliary lanes depend on whether signal control is provided. The composition and character of traffic are a design control. Movements involving large trucks need larger intersection areas and flatter approach grades than those needed at intersections where traffic consists predominantly of passenger cars. Bus stops located near an intersection may further modify the arrangement. Approach speeds of traffic also have a bearing on the geometric design as well as on control devices and markings. The number and locations of the approach roadways and their angles of intersection are major controls for the intersection geometric pattern, the location of islands, and the types of control devices. Intersections preferably should be limited to no more than four approach legs. Two or more crossroads intersecting an arterial highway in close proximity should be combined into a single crossing. The distance between intersections influences the degree of channelization at any one particular intersection. For example, where intersections are closely spaced, turn restrictions may be imposed at some intersections and pedestrian crossings may be prohibited at others. This makes some channelizing islands and auxiliary pavement areas unnecessary, or it may be appropriate to introduce continuous auxiliary lanes between two or more intersecting roads or streets to handle a buildup and weaving of traffic. Where crossroads are widely spaced, each intersection should accommodate all crossing, turning, and pedestrian movements.

B-1 APPENDIX B REVISED TEXT ON LEFT-TURN LANE WARRANTS FOR THE TRB ACCESS MANAGEMENT MANUAL Following are sections from the TRB Access Management Manual. Material suggested for removal is shown with a strikethrough line. Double underlines indicate material to be added. Figures or tables beginning with the number 10 are from the TRB Access Management Manual, while those beginning with the letter B reflect recommended additions to the text. ACCESS MANAGEMENT MANUAL (2003), CHAPTER 10, ACCESS DESIGN, AUXILIARY LANES (PAGES 171 TO 177) AUXILIARY LANES An auxiliary lane (left-turn and right-turn bays) is the most effective means of limiting the speed differential between a turning vehicle and following through traffic to a safe level. This section addresses considerations in the decision of when to provide a left-turn lane; in the design of auxiliary lanes, including maneuver distance, queue storage, determination of turn bay length, and providing for dual left turns; and in the design of right-turn storage. Guidance for when to provide left-lane lanes was developed in NCHRP Project 3-91, Left-Turn Accommodations at Unsignalized Intersections. The minimum physical length of a right-turn or left-turn bay, including the taper, consists of the maneuver distance plus the queue storage (distance d2, d3) (Figure 10-5). (The distance to maneuver laterally and decelerate to a stop, d2, is the same for left-turn bays as for right-turn bays, because the initial speed and the speed at which the turning vehicle clears the through- traffic lane are the same.) Additional details on the design of left-turn and right-turn bays are provided in Transportation and Land Development (3) and the participant’s notebook from National Highway Institute Course 133078 (1). Warrants for Left-Turn Lanes Economic analysis was used to identify situations in which left-turn lanes are and are not justified at an unsignalized intersection based on delay savings and crash reductions (reference to NCHRP Project 03-91). Guidelines for when left-turn lanes should be provided are set forth in Table B-1 and Figure B-1 for urban and suburban roadways, Table B-2 and Figure B-2 for rural two-lane highways, and Table B-3 and Figure B-3 for rural four-lane highways. The Highway Capacity Manual (reference) indicates that exclusive left-turn lanes at signalized intersections should be installed as follows: • Where fully protected, left-turn phasing is to be provided; • Where space permits, left-turn lanes should be considered when left-turn volumes exceed 100 veh/hr (left-turn lanes may be provided for lower volumes as well as on the basis of the judged need and state of local practice, or both); and

B-2 • Where left-turn volumes exceed 300 veh/hr, a double left-turn lane should be considered. Table B-1. Suggested left-turn lane warrants based on results from benefit-cost evaluations for urban and suburban ar ter ials. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Urban and Suburban Arterial Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 450 50 10 300 50 15 250 50 20 200 50 25 200 50 30 150 50 35 150 50 40 150 50 45 150 < 50 50 or More 100 < 50 (a) Three Legs (b) Four Legs Figure B-1. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on urban and suburban arterials. 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r/l n) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Three Legs Left-turn lane warranted Left-turn lane not warranted 0 5 10 15 20 25 30 35 40 45 50 0 50 100 150 200 250 300 350 400 450 Le ft- Tu rn V ol um e (v eh /h r) Major Arterial Volume (veh/hr/ln) Urban and Suburban Arterial, Four Legs Left-turn lane warranted Left-turn lane not warranted

B-3 Table B-2. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for rural two-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Three-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Bypass Lane Four-Leg Intersection, Major Two- Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 50 200 50 150 10 50 100 < 50 50 15 < 50 100 < 50 50 20 < 50 50 < 50 < 50 25 < 50 50 < 50 < 50 30 < 50 50 < 50 < 50 35 < 50 50 < 50 < 50 40 < 50 50 < 50 < 50 45 < 50 50 < 50 < 50 50 or More < 50 50 < 50 < 50 (a) Three Legs (b) Four Legs Figure B-2. Suggested left-turn treatment warrants based on results from benefit-cost evaluations for intersections on rural two-lane highways. Bypass lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Two Lanes on Major Left-turn treatment not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Two Lanes on Major Left-turn treatment not warranted Bypass lane warranted Bypass lane warranted Left-turn lane warranted

B-4 Table B-3. Suggested left-turn lane warrants based on results from benefit-cost evaluations for rural four-lane highways. Left-Turn Lane Peak-Hour Volume (veh/hr) Three-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane Four-Leg Intersection, Major Four-Lane Highway Peak-Hour Volume (veh/hr/ln) That Warrants a Left-Turn Lane 5 75 50 10 75 25 15 50 25 20 50 25 25 50 < 25 30 50 < 25 35 50 < 25 40 50 < 25 45 50 < 25 50 or More 50 < 25 (a) Three Legs (b) Four Legs Figure B-3. Suggested left-turn lane warrants based on results from benefit-cost evaluations for intersections on rural four-lane highways. Maneuver Distance Table 10-2 presents the estimated distances that would be needed by drivers to maneuver from the through lane into a turn bay and brake to a stop. These distances provide for a 10-mph (15-km/h) difference in speed for the turning vehicle when it clears the through lane. Shorter length will result in a speed differential greater than 10 mph (15 km/h). 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Three Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted 0 5 10 15 20 25 0 50 100 150 200 250 Le ft- Tu rn V ol um e (v eh /h r) Major Highway Volume (veh/hr/ln) Rural, Four Legs, Four Lanes on Major Left-turn lane not warranted Left-turn lane warranted

B-5 Table 10-2 Desirable Maneuver Distances (d2 in Figure 10-5) Speed Distancea mph km/h ft m 20 30 40 50 60 70 30 50 65 80 95 110 70 160 275 425 605 820 20 45 85 130 185 245 Note: 1. assumes a turning vehicle has “cleared the through lane” when it has moved laterally about 9 ft (3 m) so that a following through vehicle can pass without encroaching upon the adjacent traffic lane. 2. The speed differential between the turning vehicle and following through vehicles is 10 mph (15 km/h) when the turning vehicle “clears the through traffic lane.” 3. 5.8 ft/s2 deceleration while moving from the through lane into the turn lane; 6.5 ft/s2 average deceleration after completing lateral shift into the turn lane. a Rounded to 5 ft (5 m) Queue Storage A turn bay needs to be of adequate length to store vehicles waiting to complete the turn (distance d, in Figure 10-5). The number of left-turning vehicles arriving at an intersection will vary from one short time interval to another. Figure 10-6 illustrates this characteristic for a series of 2-min intervals. For this example, the average number of left turns per cycle is two vehicles, where the number of left turns on any given cycle ranges from none to four. It is important for turn bays on roadways of a high functional classification to be of sufficient length to store all arriving vehicles most of the time. On roadways of lesser importance, a lower likelihood of storing all arriving vehicles may be acceptable. The left-turn storage length at a signalized intersection will be affected by: • The left-turn flow rate for the design hour; • Signal timing, including the cycle length (number of cycles per hour), duration of a separate left-turn phase (left-turn arrow, if left turns are permitted on “green” or if left turns are restricted to “left on green arrow only”), and the number of left-turn lanes; • The probability the storage length will be sufficient to store all left-turning vehicles on an acceptable percentage of the cycles, such as at least 99.5% or 95% of the time for a at the intersection on a major arterial or a collector; and • The percentage of large vehicles.

B-6 Figure 10-7 presents an easy to use nomograph for estimating minimum left-turn storage. The same procedure can be used for left turns at unsignalized intersections by assuming a time interval (in lieu of cycle length). AASHTO (6, 1990, p. 829) suggests using a 2-min interval, which is the practice of the Florida Department of Transportation.

B-7

B-8 The following equation is another easy method for estimating left-turn queue storage length: (1) Where L = design length for left-turn storage (ft); V = estimated left-turn volume [vehicles per hour (veh/hr)]; N = number of cycles per hour (for the Green Book unsignalized procedure this would be 30) (V/N is the average number of turning vehicles per cycle); k = factor that is the length of the longest queue (design queue length) divided by average queue length (a value of 2.0 is commonly used for major arterials and a value of 1.5 to 1.8 might be considered for an approach on a minor street or on a collector where capacity will not be critical) (for the Green Book procedure this would be 1); and s = average length per vehicle, including the space between vehicles, generally assumed to be 25 ft (7.6 m) (if percentage of trucks and buses is known, minimum queue storage can be adjusted by using Table 10-3). Table 10-3 Queue Storage Length Adjustments for Trucks Percent Trucks Adjustments ft m ≤ 5 10 15 25 30 35 7.6 9.0 10.7 Bonneson et al. suggest guidelines for determining if the bay storage length for an unstopped approach is adequate in NCHRP Report 457. They use 4.1 sec for critical gap and 2.2 sec for follow-up time. Using their approach along with the critical gaps identified as part of NCHRP Project 3-91 results in the storage length recommendations listed in Table B-4. The approach assumes a 0.005 probability of overflow. Table 10-4 presents a comparison of the queue lengths obtained from Leisch’s nomograph (cf. Figure 10-7), and Equation 1, and the updated Bonneson approach. The nomographs developed by Harmelink in 1967 (7), or variations thereof, are often used to estimate queue storage length at unsignalized access connections. These curves and their variations (6, 11) have been widely used. The updated Bonneson approach considers the critical gaps observed in more recent research and allows sensitivity to the opposing volume quantity. However, recent information indicates that the time intervals for perception–reaction time plus the maneuver time to complete the left-turn maneuver are consistently larger than that used in developing the curves in 1967. Hence, left-turn bays would be warranted at lower volumes and storage lengths would need to be larger. A discussion of this problem is included in the participant’s notes for National Highway Institute Course 133078 (1) and Transportation and Land Development (3).

B-9 Table B-4. Recommended Storage Lengths for Arterials from Access Management Manual Equation and NCHRP Report 457 Equations with Revised Critical Gap. Left-Turn Volume (veh/hr) Storage Length, Rounded Up to Nearest 25-ft Increment (ft) Storage Lengths from Other Manuals for Comparison Storage Lengths Calculated from Equations b Documented in NCHRP Report 457 Using Revised Critical Gaps and 0.005 Probability of Overflow Green Book Procedure (k=1)a Equation (k=2)a Opposing Volume (veh/hr) 200 400 600 800 1000 Critical Gap = 5.0 sec, Follow-Up Gap = 2.2 sec (Represents the 50th Percentile Critical Gap Found in Field Studies) 40 75 75 50 50 50 50 50 60 50 100 50 50 50 50 50 80 75 150 50 50 50 50 50 100 100 175 50 50 50 50 75 120 100 200 50 50 50 75 75 140 125 250 50 50 50 75 75 160 150 275 50 50 75 75 100 180 150 300 50 50 75 75 100 200 175 350 50 75 75 100 125 220 200 375 50 75 75 100 125 240 200 400 75 75 100 125 150 260 225 450 75 75 100 125 175 280 250 475 75 75 100 125 175 300 250 500 75 100 125 150 200 Critical Gap = 6.25 sec, Follow-Up Gap = 2.2 sec (Represents the 85th Percentile Critical Gap Found in Field Studies, 85th Percentile Is Preferred for Design) 40 75 75 50 50 50 50 50 60 50 100 50 50 50 50 50 80 75 150 50 50 50 50 75 100 100 175 50 50 50 75 75 120 100 200 50 50 75 75 100 140 125 250 50 50 75 100 125 160 150 275 50 75 75 100 150 180 150 300 50 75 75 125 150 200 175 350 50 75 100 125 200 220 200 375 75 75 100 150 225 240 200 400 75 75 125 150 275 260 225 450 75 100 125 175 325 280 250 475 75 100 125 200 400 300 250 500 75 100 150 225 525 a, b See Table B-5 for equations. This table assumes 25 ft per vehicle spacing. Table 10-3 provides other suggested spacing lengths based on percent trucks.

B-10 Table B-5. Equations Used to Determine Storage Length. Equation in TRB Access Management Manual (1) Where L = design length for left-turn storage (ft); V = estimated left-turn volume, vehicles per hour (veh/hr); Nc = number of cycles per hour (for the Green Book unsignalized procedure this would be 30) (V/N is the average number of turning vehicles per cycle); k = factor that is the length of the longest queue (design queue length) divided by average queue length (a value of 2.0 is commonly used for major arterials, and a value of 1.5 to 1.8 might be considered for an approach on a minor street or on a collector where capacity will not be critical) (for the Green Book procedure this would be 1); and s = average length per vehicle, including the space between vehicles, generally assumed to be 25 ft (adjustments are available in several documents for trucks and buses such as the TRB Access Management Manual—see Table 10-3). Equations Used in NCHRP Report 457 Equations also used to generate values in Table B-4 Where P(n>N) = probability of bay overflow; v = left-turn vehicle volume (veh/hr); N = number of vehicle storage positions; c = movement capacity (veh/hr); Vo = major-road volume conflicting with the minor movement, assumed to be equal to one-half of the two-way major-road volume (veh/hr); SL = storage length (ft); tc = critical gap (sec); tf = follow-up gap (sec); and VL = average length per vehicle, including the space between vehicles, generally assumed to be 25 ft (adjustments are available in several documents for trucks and buses such as the TRB Access Management Manual—see Table 10-3).

B-11 Table 10-4 Comparison of Queue Storage Method Volume (veh/hr) Cycle Length (s) Storage (ft) Desirable Minimum Nomograph 240 240 60 120 200 400 150 300 L=(V/N)(2.0)(25) 240 240 60 120 200 400 L=(V/N)(1.5)(25) 240 240 60 120 150 300 Method Volume (veh/hr) Cycle Length (s) Opposing Volume (veh/hr) Storage (ft) Desirable Minimum Nomograph 240 240 60 120 200 400 150 300 L=(V/N)(2.0)(25) 240 240 60 120 200 400 L=(V/N)(1.5)(25) 240 240 60 120 150 300 Updated Bonneson Approach with Critical Gap = 5.0 sec 240 NA 200 400 600 800 1000 75 75 100 125 150 Updated Bonneson Approach with Critical Gap = 6.25 sec 240 NA 200 400 600 800 1000 75 75 125 150 275 The nomograph for four-lane roadways is presented in Figure 10-8. This nomograph is used by reading horizontally for the opposing volume Vo and vertically for the turning volume VL. The minimum storage length S is where the vertical and horizontal lines intersect. For example, for Vo = 750 and VL = 50, the minimum storage length is 75 ft. A vehicle making a left turn from a two-lane roadway blocks all following advancing vehicles. Therefore, a set of several nomographs is needed for two-lane roadways for different combinations of speed, opposing volume, advancing volume, and percent left turns. Table 10-5 presents in a simplified form minimum left-turn queue storage on two-lane roadways.

B-12 When applying the storage length from Figure 10-8 and Table 10-5, as well as adaptations of the Harmelink nomographs, keep in mind the following. Recent research (12) indicates that the perception–reaction time needed by most drivers is considerably longer than that used by Harmelink. The time to complete a left turn is also found to be longer (6). Current practice indicates that a minimum storage length is at least 100 ft (30 m) in suburban/urban areas and 50 ft (15 m) in rural areas (1, 3). These minimum lengths would apply when methods such as the nomograph or Equation 1 result in shorter lengths or when there are no traffic estimates upon which to determine queue length. Where dual left-turn lanes are provided, the storage length may be estimated by dividing the single-lane queue storage length by 1.8 and rounding to the next highest 25-ft interval.

B-13 Table 10-5 Minimum Left-Turn Queue Storage Lengths (ft) on Two-Lane Roadways Opposing Volume (veh/hr) Advancing Volume (veh/hr) 200 400 600 800 Speed = 40 mph 200 400 600 800 M M M M M 75a 75a 100 75a 100 125 150 100 125 150 Speed = 50 mph 200 400 600 800 M M M 75a 75a 75a 100 100 100 100 125 150 100 125 150 Speed = 60 mph 200 400 600 800 M M M 75a 75 75 100 125 100 100 150 175 125 150 175 Note: 1. AASHTO (6) as adapted from Harmelink (7) 2. Table values are for left-turn volumes that are 10% of the advancing volume. 3. Queue storage lengths rounded to 25-ft intervals are not highly sensitive to the percentage of left turns. With 20% of the advancing volume turning left, the minimum left-turn storage would typically be 25 ft (one car length, including the gap between vehicles) longer than the table volumes when the sum of advancing and opposing volumes is less than 1200 veh/hr. 4. M indicates that a 50-ft minimum storage in rural areas and a 100-ft minimum in suburban/urban areas would be required. 75a = ≥ 75-ft storage required in rural areas; a 100-ft minimum should be used in suburban/urban areas. Turn Bay Length The design length of a turn bay may be controlled by either off-peak conditions or peak period conditions. Both need to be calculated, with the longer of the two generally the more desirable to be used as the design length. The following steps are suggested for determining the design length. Determine the length for the peak period as follows: a. Obtain the maneuver distance from Table 10-2. b. Estimate the queue storage length: 1. For a signalized intersection use Figure 10-7 or Equation 1. 2. For an unsignalized intersection on an undivided, multilane roadway use Figure 10-8 or Equation 1 using 30 intervals of 2 min per hour Table B-3. 3. For an unsignalized intersection on a two-lane roadway use curves such as that developed by Harmelink (7) or Table 10-5 (interpolating if necessary) or by Equation 1 using 30 intervals of 2 min per hour. Table B-4. c. Add the maneuver and the storage length to determine the total turn bay length. 1. Repeat Step b(1) for off-peak conditions (speed and volumes).

B-14 2. Obtain the minimum turn bay length by selecting the longer of the peak or off-peak length, including taper. 3. Estimate the queue length in the through-traffic lanes. Typically, this needs to be done only for the peak period. If the queue in the through lane [Step 6 c(3)] is longer than that from Step 5 c(2), entry into the turn bay will be blocked. For vehicles to enter the turn bay under these conditions, the length of the full width of the auxiliary lane would need to be longer than the longest queue in the through traffic lane. For example: • Off-peak conditions – 40-mph progression speed – 60-s cycle, 60 cycles per hour – 85-veh/hr left turn – 85 veh/hr/60 cycles per hour = 1.42 vehicles per cycle • Off-peak bay length – Maneuver (Table 10-2, 275 ft) – Queue storage = (1.42 vehicles per cycle ×2 × 25 ft) =71 ft; use 75 ft – Turn bay length = 275 + 75 = 350 ft • Peak-period conditions – 30 mph – 120-s cycle, 30 cycles per hour – 220-veh/hr left turn – 220 veh/hr/ 30 cycles per hour = 7.33 vehicles per cycle • Peak-period bay length – Maneuver (Table 10-2) = 160 ft – Queue storage = (7.33 veh/hr × 2 × 25 ft) = 366.5 ft; use 370 ft – Turn bay length = 160 +370 = 530 ft • Design length of turn bays, including taper, is 530 ft (peak-period requirement of 530 ft is longer than off-peak requirement of 350 ft). Additional Resources J.A. Bonneson and M.D. Fontaine (2001). Engineering Study Guide for Evaluating Intersection Improvements. NCHRP Report 457. Transportation Research Board, Washington, D.C. http://onlinepubs.trb.org/onlinepubs/nchrp/esg/esg.pdf. Accessed August 2, 2010. K. Fitzpatrick, M. A. Brewer, J. S. Gluck, W.L. Eisele, Y. Zhang, H. S. Levinson, W.von Zharen, M. R. Lorenz, V. Iragavarapu, E. S. Park. (2010). Development of Left-Turn Lane Warrants for Unsignalized Intersections. NCHRP Report [to be determined]. Transportation Research Board, Washington, D.C.

B-15 ACCESS MANAGEMENT MANUAL (2003), APPENDIX A, ACCESS MANAGEMENT TECHNIQUES (PAGES 294 TO 297) TURN BAYS AND TURN LANES Isolated Left-Turn Bay on Undivided Roadways Description The isolated left-turn bay provides an auxiliary lane to remove left-turning vehicles from the through-traffic lane on an undivided roadway (Figure A-7). Applications The technique can be applied to • Two-lane and undivided four-lane roads, • High-speed roadways, • Areas with a high left-turn volume, and • Areas with low left-turn volume and high through volume., and • Signalized and unsignalized locations. Special Considerations • The purpose of the isolated left-turn bay is to provide a protected area for left turning vehicles, thereby increasing intersection safety and reducing delay for through traffic. • Painted channelization is commonly practiced on high-speed roads. • Reflectorized pavement buttons often supplement the painted channelization to improve nighttime visibility. • It is desirable to retain the shoulder wherever possible, especially on higher speed (> 45 mph) roadways. • Lane width. The width of the left-turn lane is commonly the same as that of a through- traffic lane [e.g., 12 ft (3.6 m)] (1, 2). Narrower lanes may be appropriate on low-speed roadways where right-of-way is limited. The separator between the left-turn lane and the opposing traffic lane commonly consists of double yellow 4-in. (100-mm) lines. Depending on roadway width, the painted separator may range from about 1 ft (300 mm) to 4 ft (1.2 m). Where a raised separator is used, visibility and aesthetics can be enhanced by using a width that is adequate for appropriate landscaping. At its widest point, the median width is the sum of the width of the separation outside edge to outside edge) plus the width of the left-turn lane. • At a four-way intersection, the lane may need to be sufficiently wide to store a crossing vehicle clear of both traffic streams. A width of 30 ft (9.1 m) has been suggested for design for passenger cars (3, 4). This generally ensures that passenger car drivers will stop within the median area and clear the edge of the traffic lanes in front and behind the stopped vehicle. In cases where adequate width cannot be provided, alternatives may be used to offset the crossroad connections so that a crossing maneuver is made by a right turn followed by a left turn or to design the turn bays as a directional opening for opposing left turns from the major roadway only.

B-16 • Bay length. The redirection taper length (L1) allows drivers to make a gradual lateral movement. The redirection taper ratio is a function of speed and may range from 15:1 at speeds less than 30 mph to 70:1 or more at 70 mph (Figure A-8). • A short bay taper length (L2) helps to clearly identify the beginning of an auxiliary lane. The turn bay length (L4) is the sum of the bay taper length (L2) and the length of the full- width segment (L3). This length permits the driver of a left-turning vehicle to leave the through traffic lane at an acceptable speed differential with follow-through traffic and comfortably brake to a stop. This length also permits storage of vehicles waiting to complete the left turn. In developed areas, storage for four vehicles is often considered to be the minimum; storage for two vehicles is considered adequate for rural locations (4, Chapter 3 addresses issues in detail design of left-turn lanes, and course notes and presentation comprehensive summaries of various practices are always included). Where speed or turn volume varies throughout the day, the condition that results is the longest sum of maneuver distance plus storage. This will be the minimum bay length. See Chapter 8 for more information on the functional intersection area; Chapter 10 addresses turn bay length. Advantages • The left-turning vehicle is able to clear the through traffic lane at an acceptable speed. • Rear-end and left-turn collisions are greatly reduced (4–7). • Crash rates may be reduced by 25% (4). • Capacity is increased (4). Disadvantages • The technique may require reconstruction of a considerable length of roadway to attain the additional pavement width. • Achieving the turn bay by paint striping only results in loss of the shoulder. • A transition by through traffic is required. Examples • NCHRP Project 03-91 produced guidance to identify whether a left-turn lane should be provided at an unsignalized location. See Chapter 10 for more information and the warrants (reference NCHRP Project 03-91). • Colorado DOT has established that a warrant (8) for a left-turn deceleration lane is met when the peak-hour ingress volume exceeds that in Table A-4. See Chapter 5 for a description of the Colorado DOT access management categories. • Oregon DOT criteria (1) are shown in Table A-5. • Many jurisdictions use the AASHTO (9) guidelines for left-turn lanes. These guidelines are a simplified tabular version of Harmelink’s curves (6). Recent investigators have shown that the time drivers take to execute a left turn is longer than that assumed by Harmelink. Thus, volume warrants for left-turn bays would be less than that indicated by Harmelink’s curves, and the storage length would be longer. • Delaware uses a more extensive table for opposing, advancing, and left-turn volumes.

B-17 • Colorado redirection taper ratios (L1) (8) are Posted Speed (mph) Taper Ratio 30 15:1 40 30:1 45 45:1 50 50:1 60 60:1 70 70:1 • See Figure A-9 for examples from TX-105 in Grimes County and US-97 near Redmond, Oregon. References 1. Standard Drawing TM539. Oregon Department of Transportation, Salem, May 2001. 2. Design Manual—Roadway. New Jersey Department of Transportation, Trenton, 2002. www.state.nj.us/transportation/cpm/RoadwayDesignManualMetric/interactivedownload.htm. 3. Stover, V.G., and F.J. Koepke. Transportation and Land Development, 2nd ed. Institute of Transportation Engineers, Washington, D.C., 2002. 4. S&K Transportation Consultants, Inc. Access Management, Location and Design. Participant notebook for NHI Course 133078. National Highway Institute, FHWA, Arlington, Virginia, April 2000. 5. Agent, K.R. Warrants for Left-Turn Lanes. Transportation Quarterly, Eno Transportation Foundation, Washington, D.C., Jan. 1983. 6. Harmelink, M.D. Volume Warrants for Left-Turn Storage Lanes at Unsignalized Grade Intersections. In Highway Research Record 211, HRB, National Research Council, Washington, D.C., 1967, pp. 1–18. 7. Pline, J.L. NCHRP Synthesis of Highway Practice 225: Left-Turn Treatments at Intersections.TRB, National Research Council, Washington, D.C., 1996. 8. State Highway Access Code. Colorado Department of Transportation, Denver, effective Aug. 31, 1998. www.dot.state.co.us/BusinessCenter/Permits/Access/601_1_AccessCode_May2002.pdf. 9. A Policy on Geometric Design of Highways and Streets. AASHTO, Washington, D.C., 1984, 1990, 1994. Additional Resources Left-Turn Lane Criteria. Oregon Department of Transportation, Salem, June 22, 1996. Guidelines for Left-Turn Lanes. Committee 4A-22 Final Report. Institute of Transportation Engineers, Washington, D.C., Sept. 1991.

B-18 Gluck, J., H.S. Levinson, and V. Stover. NCHRP Report 420: Impacts of Access Management Techniques. TRB, National Research Council, Washington, D.C., 1999. Geometric Design Standards. Ontario Ministry of Transportation, Toronto, Canada. ACCESS MANAGEMENT MANUAL (2003), APPENDIX A, ACCESS MANAGEMENT TECHNIQUES, TURN BAYS AND TURN LANES (PAGES 298 TO 300) Shoulder Bypass at Three-Way (T) Intersection Description The shoulder bypass allows a through vehicle to bypass a left-turning vehicle that is stopped in the traffic lane (Figure A-10). Applications The technique can be applied • To three-way intersections on minor roadways, • To locations where space is not available for an isolated left-turn bay, and • As a temporary solution until an isolated left-turn bay can be constructed. Special Considerations • The technique is appropriate for locations with low left-turn volumes. • It is also appropriate for low-volume roadways. • The bypass lane is typically the same width as a regular traffic lane. • Approach and departure tapers are typically much shorter than the redirection tapers used for isolated left-turn bays. • A narrow shoulder of earth and grass is commonly provided on the outside of the bypass lane. Advantages • Safety is improved, because the bypass helps reduce the potential for rear end collisions. • Delays to through traffic are reduced. It is inexpensive to implement, especially if a paved shoulder >10 ft wide already exists. • The bypass takes less space than an isolated left-turn bay. Disadvantages • The bypass requires through traffic to transition around vehicles stopped in the through lane. • It is less safe than an isolated left-turn bay. • Driver expectancy is violated. • Earthwork may be necessary to add the bypass. Additional right-of-way may be needed for the widening and for modifying the drainage.

B-19 Examples • NCHRP Project 03-91 produced guidance to identify whether a left-turn shoulder bypass should be provided at an unsignalized location on rural two-lane and four-lane roadways. See Chapter 10 for more information and the warrants (reference NCHRP Project 03-91). • In Becker County, Minnesota, shoulder bypasses are used for T-intersections for major roads with an annual average daily traffic (AADT) > 500 and for minor roads with an AADT > 250 (D. Heyer, Detroit Lakes, Minnesota, personal communication, Feb. 22, 1999). • In Minnesota DOT warrants, the technique is applied for T-intersections that do not meet left-turn warrants (1, Section 5-3-01.02; 2). The bypass is designed with a 54-m-long approach and departure tapers and a 75-m full- width section of which 15 m are beyond the far edge of the intersecting road. • Connecticut DOT uses a shoulder bypass lane where space is inadequate for its standard left-turn lane design (discussions with Minnesota DOT personnel during National Highway Institute Course 133078, April 2000). • The Texas DOT uses the shoulder bypass on a case-by-case basis. However, there are no specific warrants for the technique (correspondence with Texas DOT District 21 traffic engineer). • Delaware DOT warrants (2) are as follows: Volume (AADT) Highway Left-Turn Technique Applied <2,000 >200 Bypass lane 2,001–4,000 201–400 Bypass lane >400 Consider left-turn lane >4,000 50–400 Bypass lane >400 Consider left-turn lane • For Michigan DOT warrants, see Figure A-11 (2). • The schema for Michigan DOT design is shown in Figure A-12. • Figure A-13 shows an example of the technique used at an intersection with a road serving a resort and farms in Becker County, Minnesota. Speeds on the road exceed 60 mph. References 1. Road Design Manual. Minnesota Department of Transportation, St. Paul, Minnesota, Dec. 24, 1996. 2. Note on Warrant for Passing Flares at Driveways. Traffic and Safety Division, Michigan Department of Transportation, Lansing, Michigan, Nov. 30, 1988.

B-20 ACCESS MANAGEMENT MANUAL, APPENDIX A, ACCESS MANAGEMENT TECHNIQUES, TURN BAYS AND TURN LANES (PAGES 303 TO 304) Left-Turn Bay at Median Opening Description The left-turn bay at a median opening provides for left-turn deceleration and storage at a signalized or an unsignalized median opening (Figure A-16). Application It can be applied to all median openings, both signalized and unsignalized. Special Considerations • Issues involving design of a left-turn bay (such as taper, maneuver distance, and storage) are addressed in Chapter 10. See also Access Management, Location and Design for warrants and guidelines for left-turn bays (1, Chapter 3) and for design guidelines (1, Chapter 4). • The width of the left-turn lane and the divider separating the left turns from the opposing traffic will depend upon the available median width. Below are examples of the combinations of lane and separation widths that might be considered for different median widths, as adapted from the National Highway Institute (1): Median Left-Turn Width (ft) Lane Width (ft) Divider 16 11 or 12 4 ft raised 14 11 3 ft raised 14 10 4 ft raised 12 11 4 in double yellow lines 12 10 2 ft slightly raised • In urban and suburban areas, a short taper does not restrict lateral movement into the left- turn lane at the slow speeds encountered during peak periods. • Short tapers on horizontal curves to the left help keep drivers of through vehicles from inadvertently entering the left-turn bay. • A turn bay of appropriate length allows left-turning vehicles to clear the through lane without interfering unduly with following through traffic and then decelerate to a stop before reaching the end of a queue of vehicles waiting to complete the left turn. • Left-turn lanes are normally the same width as a traffic lane. However, lanes as narrow as 9 ft (2.75 m) have been used in urban areas where right-of-way is limited and speeds are very slow [e.g., ≤30 mph (≤50 km/h)]. Advantages • The technique provides refuge for drivers who are making left turns, thereby minimizing impacts on through traffic (2). • It allows left-turning vehicles to leave the through lane with an acceptable speed differential with through traffic (1, 3). • Crash rates at unsignalized median openings may be reduced by 50% to more than 75% (2–6) and by about 20% to more than 55% at signalized locations (1, Chapter 5; 2; 4).

B-21 • Capacity can be increased by 25% or more (1, Chapter 2; 2; 5). • Delay to through traffic is reduced (2). Disadvantages • The left-turn bay cannot be used on medians that are too narrow to provide a left-turn lane and a separator between the left-turn lane and the opposing traffic lane. • Proximity of the bay to any other median openings may limit the length of the turn lane or require closure of a less important opening. Examples • NCHRP Project 03-91 produced guidance to identify whether a left-turn lane should be provided at an unsignalized location. See Chapter 10 for more information and the warrants (reference NCHRP Project 03-91). • Vancouver, British Columbia, in Canada has a program to provide existing intersections with left-turn bays. • The DOTs of Colorado, Florida, and Oregon call for left-turn bays at all median openings on divided highways. • Florida DOT practice is to add turn bays as part of resurfacing projects. In some resurfacing projects, the Florida DOT also converts unsignalized full median openings in suburban and urban areas to directional left-turn and U-turn openings. References 1. S&K Transportation Consultants, Inc. Access Management, Location and Design. Participant notebook for NHI Course 133078. National Highway Institute, FHWA, Arlington, Virginia, April 2000. 2. Gluck, J., H.S. Levinson, and V. Stover. NCHRP Report 420: Impacts of Access Management Techniques. TRB, National Research Council, Washington, D.C., 1999. 3. Stover, V.G., and F.J. Koepke. Transportation and Land Development, 2nd ed. Institute of Transportation Engineers, Washington, D.C., 2002. 4. Hagenauer, G.F., J. Upchurch, D. Warren, and M.J. Rosenbaum. Intersections. In Synthesis of Safety Research Related to Traffic Control and Roadway Elements, Vol. 1, Report FHWA-TS-82-232, FHWA, U.S. Department of Transportation, Dec. 1982. 5. Rudberg, D.H. Arterial Streets: Important Components of Vancouver’s Urban Transportation System. ITE Journal, Oct. 1988. 6. Wilson, J.E. Simple Types of Intersection Improvements. In Special Report 93: Improved Street Utilization through Traffic Engineering, Highway Research Board, National Research Council, Washington, D.C., 1967. Additional Resources A Policy on Geometric Design of Highways and Streets. AASHTO, Washington, D.C., 1984, 1990, 1994.

B-22 Oppenlander, J.C., and J.E. Oppenlander. Storage Requirements for Signalized Intersection Approaches. ITE Journal, Feb. 1996. Oppenlander, J.C., and J.E. Oppenlander. Complete Tables for Storage Requirements for Signalized Intersection Approaches. Supplement to ITE Journal, Feb. 1996. Oppenlander, J.C., and J.E. Oppenlander. Storage Lengths for Left-Turn Lanes with Separate Phase Control. ITE Journal, July 1994. Koepke, F.J., and H.S. Levinson. NCHRP Report 348: Access Management Guidelines for Activity Centers. TRB, National Research Council, Washington, D.C., 1992.

C-1 APPENDIX C STATE WARRANTS/GUIDELINES FOR LEFT-TURN LANES State Material on Left-Turn Guidelines or Warrants Alabama Currently offline Alaska http://www.dot.state.ak.us /stwddes/dcsprecon/assets /pdf/preconhwy/ch11/cha pter11.pdf Alaska Highway Preconstruction Manual January 2005 1190. Driveway Standards 1190.5 Control Dimensions 11. Speed Change Lane and Left-Turn Lanes: On high-speed (50 mph or over) or high-volume arterial roadways, speed change lanes may be required for the acceleration or deceleration of vehicles entering or leaving the public roadway from or to a higher-volume traffic generation (greater than or equal to 100 vehicles per hour) or attracting development. Use Figure 4-3 of NCHRP 279 Intersection Channelization Design Guide as a guideline for the right-turn treatments. On a one-way street, the above criteria also apply to the left through lane. For guidelines on the need for left- turn lanes on a main street or road at a driveway, refer to Exhibit 9-75 in AASHTO A Policy on the Geometric Design of Highways and Streets 2001. 1150.2. Urban Arterials 1150.2.1. General Design Considerations Design of urban arterials shall conform to recommendations in the Green Book. 1150.2.2. Medians Median openings generally permit cross traffic and left turns which conflict with the through traffic on the arterial. These conflicts result in delays and accident exposure, which you can minimize by providing as few median openings as possible. However, keep in mind that restriction of mid-block left turns often substantially increases the number of U-turns at the adjacent intersection with median openings. A similar situation exists where a minor street intersects the arterial and does not provide a median opening. Generally, provide median openings only if the volume of cross- or left-turn traffic is relatively large, such as at another arterial or major collector street or, in some cases, at an access point to a major traffic generator, such as a regional shopping center or industrial plant. Because the openings are at major traffic points, assume that at some time, if not immediately, these median opening locations will be signalized. Additionally, where signalized intersections are 0.5 mile or less apart, efficiency and safety require interconnected or synchronized signals to achieve smooth traffic flow along the arterial. Arizona http://www.azdot.gov/hig hways/Roadway_Enginee ring/Roadway_Design/Gu idelines/Manuals/PDF/Ro adwayDesignGuidelines.p df Roadway Design Guidelines July 2008 408.10 — Left-turn Channelization A) General: A left-turn lane serves to expedite the flow of through traffic, to control the movement of turning traffic, and to improve the safety and capacity of the intersection. Traffic Engineering Group will analyze the traffic movements and other factors at an intersection to determine the need for a separate left-turn lane(s) and establish the vehicle storage requirements for the lane(s).

C-2 State Material on Left-Turn Guidelines or Warrants Arkansas Not available online California http://www.dot.ca.gov/hq/ oppd/hdm/pdf/english/chp 0400.pdf Highway Design Manual Chapter 400: Intersections at Grade July 2009 405.2 Left-turn Channelization (1) General. The purpose of a left-turn lane is to expedite the movement of through traffic, control the movement of turning traffic, increase the capacity of the intersection, and improve safety characteristics. The District Traffic Branch normally establishes the need for left-turn lanes. See “Guidelines for Reconstruction of Intersections,” August 1985, published by the California Division of Transportation Operations. Colorado http://www.dot.state.co.us /DesignSupport/Design% 20Guide%2005/DG05%2 0Ch%2009%20Intersectio ns.pdf http://www.dot.state.co.us /DesignSupport/Design% 20Guide%2005/Index.ht m Design Guide Chapter 9—Intersections April 2006 9.18.6.1 Median Left-Turn Lane Warrants To facilitate flow where the intersection is unsignalized, the following guidelines are suggested: • Left-turn lanes should be considered at all median crossovers on divided, high- speed highways. • Left-turn lanes should be provided at all uncontrolled approaches of primary, high- speed rural highway intersections with other arterials and collectors. • Left-turn lanes should be provided on stopped or secondary approaches based on analysis of the capacity and operations of the unsignalized intersection. Connecticut http://www.conndot.ct.go v/publications/hdm/Chapt er%2011.pdf ConnDOT Highway Design Manual Chapter 11: Intersections At-Grade December 2006 In general, exclusive left-turn lanes should be provided for at-grade intersections as follows: 1. on all divided urban and rural highways with a median wide enough to allow a left- turn lane (this applies to intersections with public roads and to major traffic generators); 2. for all approaches at arterial/arterial intersection; 3. at any unsignalized intersection on a two-lane urban or rural highway which satisfies the criteria in Figures 11-5B, 11-5C, 11-5D, 11-5E or 11-5F [appears to be based on NCHRP Report 279/Harmelink, also includes graphs for 55 and 45 mph, and does not include the four-lane undivided graph]; 4. at any intersection where a capacity analysis determines a left-turn lane is necessary to meet the level-of-service criteria; 5. at any intersection included or expected to be within an interconnected signal system where the presence of standing left-turning vehicles would disrupt the progression of platooned traffic; or 6. at any intersection where the crash experience, traffic operations, sight distance restrictions or engineering judgment indicates a significant problem related to left- turning vehicles.

C-3 State Material on Left-Turn Guidelines or Warrants Delaware http://www.deldot.gov/inf ormation/pubs_forms/man uals/road_design/pdf/07_i ntersections.pdf DelDOT Road Design Manual Chapter 7—Intersections November 2006 For unsignalized intersections, left-turn lanes should be provided: • At all median openings on high-speed divided highways. • On approaches where sight distance is limited. • At non-stopping approaches of rural arterials and collectors. • At other approaches where required based on capacity and operational analysis. There may be other needs, primarily safety, for left-turn lanes at other locations than mentioned in these general guidelines. For two-lane roadways, a high-volume, intersecting minor roadway or entrance may also create the need to separate movements with auxiliary turning lanes. Figure 7-12 [Green Book values] is a tabulated guide to traffic volumes where left- turn lanes should be considered on two-lane highways. For the values shown, left turns and right turns from the minor street can be equal to, but not greater than, the left turns from the major street. Florida http://www.dot.state.fl.us/ rddesign/FloridaGreenboo k/2007/2007FloridaGreen book.pdf Manual on Uniform Minimum Standards for Design, Construction, and Maintenance for Streets and Highways Did not find discussion on left-turn installation guidelines or warrants Florida http://www.dot.state.fl.us/ planning/systems/sm/acc man/pdfs/driveway2008.p df Driveway Information Guide September 2008 Section 3.6: Left Turn Lanes Serving Driveways on Multilane and 2 Lane Roadways On a multilane roadway with a median Whenever a driveway is directly served by a median opening, a left turn lane should be available. This provides for the safest left turns into the driveway. On a two-lane roadway Exclusive left turn lanes should be considered at any location serving the public, especially on curves and where speeds are 45 mph and higher. The AASHTO Green Book contains guidance on this issue. However, the guidelines were developed based on delay rather than crash avoidance. Safety is the main reason behind exclusive left turn lanes.

C-4 State Material on Left-Turn Guidelines or Warrants Georgia http://www.dot.ga.gov/doi ngbusiness/PoliciesManua ls/roads/Documents/Desig nPolicies/DrivewayFull.p df Regulations for Driveway and Encroachment Control March 2004 4I-1-2 Minimum Requirements for Left Turn Lanes Left turn lanes must be constructed at no cost to the Department if the daily site generated Left Turn Volumes (LTV) based on ITE Trip Generation (assuming a reasonable distribution of entry volumes) meet or exceed the values shown in Table 4-7a Condition 1. If the LTVs are below the requirements for Condition 1, the applicant may be required to construct a Right Hand Passing Lane (see Appendix F) if they meet the criteria in Table 4-7b Condition 2. The District Access Management Engineer will use engineering judgment to determine if the field conditions would allow construction of the Right Hand Passing Lane. Passing lane sections fall under the criteria for two or more lanes. Table 4-7a. Georgia’s Regulations for Driveway and Encroachment Control Left Turn Requirements, Condition 1. Condition 1: Left Turn Requirements—Full Construction Posted Speed 2 Lane Routes More than 2 Lanes on Main Road <6000 ADT ≥6000 ADT <10000 ADT ≥10000 ADT 35 mph or less 40 to 50 mph ≥55 mph 300 * 250 * 200 * 200 * 175 * 150 * 400 * 325 * 250 * 300 * 250 * 200 * * LTV a day Table 4-7b. Georgia’s Regulations for Driveway and Encroachment Control Left Turn Requirements, Condition 2. Condition 2: Left Turn Requirements w/R ght Hand Passing L ne Option Posted Speed 2 Lane Routes Only <4000 ADT ≥4000 ADT 35 mph or less 40 to 50 mph ≥55 mph 200 LTV a day 100 LTV a day 75 LTV a day 125 LTV a day 75 LTV a day 50 LTV a day In the event the District Access Management Engineer determines that field conditions or other factors indicate that it would be in the best interest of the Department to waive the left turn lane requirement, the District Access Management Engineer must document the recommendations using the form in Appendix E. The recommendations shall be approved by the District Engineer and be attached to the Permit. The District Access Management Engineer may also require the addition of a Left Turn lane, even when the conditions in Table 4-7 are not met, if roadway geometry or field conditions indicate that the safety of the traveling public would be improved. The recommendation must be documented and approved by the District Engineer for inclusion with the Permit. Hawaii http://hawaii.gov/dot/high ways Did not find a link for a design manual type of document Idaho http://itd.idaho.gov/manua ls/Online_Manuals/Desig n/Design_Manual.htm Roadway Design Manual July 2008 Did not find information on left-turn lane warrants

C-5 State Material on Left-Turn Guidelines or Warrants Illinois http://dot.state.il.us/desen v/BDE%20Manual/BDE/ pdf/chap36.pdf Bureau of Design and Environment Manual Chapter 36—Intersections December 2002 36-3.01(b) Left-Turn Lanes The accommodation of left turns is often the critical factor in proper intersection design. Left-turn lanes can significantly improve both the level of service and intersection safety. Always use an exclusive left-turn lane at all intersections on divided urban and rural highways with a median wide enough to accommodate a left- turn lane, regardless of traffic volumes. Consider using an exclusive left-turn lane for the following: • at any unsignalized intersection on a two-lane urban or rural highway which satisfies the criteria in Figures 36-3C, D, E, F, or G [appears to be based on Harmelink/NCHRP Report 279]; • at any signalized intersection where the left-turning volume is equal to or greater than 75 veh/hr for a single turn lane or 300 veh/hr for a dual turn lane; • any intersection where a capacity analysis determines a left-turn lane is necessary to meet the level-of-service criteria, including dual left-turn lanes; • for uniformity of intersection design along the highway if other intersections have left-turn lanes (i.e., to satisfy driver expectancy); or • any intersection where the crash experience, traffic operations, sight distance restrictions (e.g., intersection beyond a crest vertical curve), or engineering judgment indicates a significant conflict related to left-turning vehicles.

C-6 State Material on Left-Turn Guidelines or Warrants Indiana http://www.state.in.us/dot /div/contracts/standards/d m/english/Part5Vol1/ECh 46/ch46.htm Indiana Design Manual, Road Design, Intersections At-Grade December 3, 2008 Document Revised January 2010. Updated URL: http://www.in.gov/do t/div/contracts/standa rds/dm/Part5/Ch46/ch 46.htm The Indiana Design Manual provides information on warrants for left-turn lanes (Section 46-4.01(02)) and passing blisters (Section 46-4.03). An exclusive left-turn lane should be provided as follows: • at each intersection on an arterial, where practical; • at each intersection on a divided urban or rural highway with a median wide enough to accommodate a left-turn lane, provided that adequate spacing exists between intersections; • at an unsignalized intersection on a two-lane urban or rural highway which satisfies the criteria shown in [a table that contains the Green Book values]; • at an intersection where a capacity analysis determines that a left-turn lane is necessary to meet the level-of-service criteria, including multiple left-turn lanes; • at a signalized intersection where the design-hour left-turning volume is 60 veh/h or more for a single turn lane, or where a capacity analysis determines the need for a left-turn lane; • for uniformity of intersection design along the highway if other intersections have left-turn lanes in order to satisfy driver expectancy; • at an intersection where the accident experience, traffic operations, sight distance restrictions (e.g., intersection beyond a crest vertical curve), or engineering judgment indicates a significant conflict related to left-turning vehicles; or • at a median opening where there is a high volume of left turns, or where vehicular speeds are 50 mph or higher. The Indiana Design Manual provides information on warrants for passing blisters. Passing blisters are used to relieve congestion due to left-turning vehicles. Appendix B includes a copy of a layout for a passing blister. Indiana states the following is to be reviewed to determine the need for a passing blister: • Traffic volume. A passing blister may be provided at the intersection of a public road or street with a two-lane state highway with a design-year AADT of 5000 or greater. For a two-lane state highway with a design-year of less than 5000, a passing blister should be used only if one or more of the following occurs: o There is an existing passing blister. o There are 20 or more left-turning vehicles during the design hour. o Accident reports or site evidence, such as skid marks in the through lane displaying emergency braking, indicate potential problems with left-turning vehicles. o The shoulder indicates heavy use (e.g., dropped shoulder, severe pavement distress). • The decision on whether to use either a channelized left-turn lane or a passing blister should be based on accident history, right-of-way availability, through- and turning-traffic volumes, design speed, and available sight distance. A channelized left-turn lane should be provided if the left-turning volume is high enough that a left-turn lane is warranted.

C-7 State Material on Left-Turn Guidelines or Warrants Iowa ftp://165.206.203.34/desig n/dmanual/00_START HERE_TOC.pdf updated URL: http://www.iowadot.gov/d esign/dmanual/manual.ht ml Design Manual, Section 6C-5 November 25, 2008 Left Turn Lanes Left turn lanes provide storage in the median for left-turning vehicles, or when warranted, deceleration outside of the through traffic lanes for left-turning vehicles. All Type “A” and high volume Type “B” entrances should have left turn lanes provided, see Section 3E-2 of this manual. If a left turn deceleration is not warranted, a left turn storage lane should be provided. Normally, left turn lanes are designed as parallel lanes. Left Turn Deceleration Lane Warrants The basic guidelines for when left turn deceleration lanes are warranted involve mainline turning and approach volume, and intersection location. Turning and approach volume A left turn deceleration lane may be warranted if left turning traffic flow rate is greater than 30 vehicles per hour measured over a minimum of 15 minutes and either: a. approach volume is greater than 400 vehicles per hour, or b. approach truck traffic volume is greater than 40 vehicles per hour. Intersection location Intersection location may warrant a left turn deceleration lane even if turning and approach volumes do not. To improve operational efficiency, left turn deceleration lanes should be considered for intersections located within approximately 5 miles (8 kilometers) of an urban area with a population of 20,000 or greater. Other locations where left turn deceleration lanes may be judged to be warranted by the PMT include schools, main entrances for towns, shopping areas, housing developments, attraction locations such as recreational areas, and locations that would have special users such as truck traffic or campers. Special attention should be given to intersections serving locations that attract elderly drivers such as drug stores, grocery stores, retirement developments, medical facilities, nursing homes, etc. Entrance Types Type “A” entrance. An entrance developed to carry sporadic or continuous heavy concentrations of traffic. Generally, a Type “A” entrance carries in excess of 150 vehicles per hour. An entrance of this type would normally consist of multiple approach lanes and may incorporate a median. Possible examples include racetracks, large industrial plants, shopping centers, subdivisions, or amusement parks. Type “B” entrance. An entrance developed to serve moderate traffic volumes. Generally, a Type “B” entrance carries at least 20 vehicles per hour but less than 150 vehicles per hour. An entrance of this type would normally consist of one inbound and one outbound traffic lane. Possible examples include service stations, small businesses, drive-in banks, or light industrial plants. Type “C” entrance. An entrance developed to serve light traffic volumes. Generally, a Type “C” entrance carries less than 20 vehicles per hour. An entrance of this type would not normally accommodate simultaneous inbound and outbound vehicles. Possible examples include residential, farm or field entrances. Kansas Not available online Kentucky http://transportation.ky.go v/design/designmanual/ch apters/12Chapter%20090 0%20AS%20PRINTED% 202006.pdf KYTC Highway Design, Intersection Chapter January 2006 Unsignalized Intersection: • Left-turn lanes should be provided at: o Median openings on divided roadways but not at median crossovers on freeways and interstates o All non-stopping approaches of rural arterials and collectors o All other approaches where required on the basis of capacity, safety, and operational analysis • Left-turn lanes should be considered where sight distance is limited.

C-8 State Material on Left-Turn Guidelines or Warrants Louisiana http://www.dotd.louisiana .gov/highways/project_de vel/design/road_design/ro ad_design_manual/Road_ Design_Manual_(Full_Te xt).pdf Roadway Design Procedures and Details July 2002 Document updated January 2009. Updated URL: http://www.dotd.louisiana .gov/highways/project_de vel/design/road_design/do cuments.aspx On new four-lane highways, left turn lanes are usually provided at all intersecting side roads that have dedicated right-of-way. Right turn lanes, left turn lanes on divided roadways (both depressed and raised median sections), and dedicated left turn lanes on five-lane urban roadways are considered based on: __ traffic volumes __ turning movements __ reduced accident potential __ and increased operational efficiency Maine Did not find any discussion on left-turn warrants Maryland http://www.sha.state.md.u s/businesswithsha/bizStds Specs.asp?id=B157+B159 Book of Standards for Highway and Incidental Structures Did not find any discussion on left-turn warrants Massachusetts http://www.vhb.com/mhd Guide/mhd_GuideBook.a sp http://www.vhb.com/mhd Guide/pdf/CH_6.pdf Highway Design Guidebook Chapter 6: Intersection Design 2006 Edition 6.7.3.3 General Criteria for Right-Turn and Left-Turn Lanes Criteria for considering installation of left-turn lanes are summarized in Exhibit 6-23 [Green Book values along with criteria for 30 mph]. Considerable flexibility should be exercised in considering left-turn lanes. Typically, they involve little impact to the setting, while generally yielding large benefits in safety and user convenience. Left-turn lanes may be desirable in many situations with volumes well below those stated. These include to destinations of special interest (shopping, major institutions, etc.), or for locations with marginal sight distance on the main road or a consistent occurrence of rear-end crashes. Michigan http://mdotwas1.mdot.stat e.mi.us/public/design/engl ishroadmanual/ Road Design Manual Did not find any discussion on left-turn warrants

C-9 State Material on Left-Turn Guidelines or Warrants Minnesota http://www.dot.state.mn.u s/design/rdm/english/5e.p df Road Design Manual Chapter 5: At-Grade Intersections June 2000 5-3.01.01 Turn Lane Policy at Urban Intersections Because of the operational and safety benefits associated with right and left-turn lanes, it is Mn/DOT’s policy that, in urban areas, they be considered wherever construction is economically feasible taking into account amount of right of way needed, type of terrain, and environmentally or culturally sensitive areas. For new construction/reconstruction projects on divided highways, left-turn and right- turn lanes should be considered at all locations where a paved crossover will be constructed. For preservation projects, left-turn lanes should, if feasible, be provided: 1. At all public road median crossovers. 2. At non-public access locations generating high traffic volumes. 3. At locations where accident records confirm the existence of an excessive hazard. 4. At locations determined by the District Traffic Engineer in consideration of accidents, capacity and traffic volumes. 5. Where a median opening is planned or exists, and its continued existence is justified, a left-turn lane may be added regardless of what the access point serves. 5-4.01 Turn Lanes As with urban highways, the degree of access control greatly influences the accident rate and efficiency of traffic operation on rural highways. Therefore, designers should try to close any unjustified or potentially dangerous access points. However, if an access point is to remain open, adding a turn lane will enhance the operation and safety. 1. In addition to the policies listed below, left-turn and right-turn lanes should be constructed at locations determined by the District Traffic Engineer in consideration of accidents, capacity, and traffic volumes. 2. Where a median opening is planned, or already exists and its continued existence is justified, a left-turn lane may be added regardless of what the access point serves. 3. Turn lanes should be considered at every public road intersection along a stretch of highway if most intersections on the stretch meet the warrants. If most intersections have turn lanes, motorists will come to expect all intersections to have them. 5-4.01.01 Policy on Multi-Lane Highways 1. Right-turn and left-turn lanes should be standard features at all public access points. 2. Right-turn and left-turn lanes are also warranted if the access point serves an industrial, commercial, or any substantial trip-generating land use, or if the access point serves more than three residential units. 5-4.01.02 Policy on two-lane Rural Highways 1. Right-turn lanes should be considered when the projected ADT is over 1500, the design speed is 45 mph or higher, and the following: a. At all public road access points. b. If industrial, commercial, or substantial trip generating land use is to be served, or c. If the access serves more than 10 residential units. 2. Left turn lanes should be provided when the access is to a public road, an industrial tract or a commercial center. 3. The designer, in conjunction with the District Traffic Engineer, may select either a channelized or a painted left-turn lane. The selection will be based on a number of factors including accident history, traffic volume, comparative costs, availability of right of way, environmental impacts, and physical features such as sight distance.

C-10 State Material on Left-Turn Guidelines or Warrants Minnesota http://www.oim.dot.state. mn.us/access/index.html Access Management, Chapter 3 Updated URL: http://www.dot.state.mn.u s/accessmanagement/pdfc hapters/chapter3.pdf Turn lanes are to be provided at public street connections and driveways in accordance with the Mn/DOT Road Design Manual, Section 5-3, and the guidance provided in the Access Management Manual. Guidance provided includes the following: • Warrant 7: Crash History — At high-volume driveways (>100 trips per day) and all public street connections that demonstrate a history of crashes of the type suitable to correction by a turn lane or turn-lane treatment (typically three or more correctable crashes in one year), or where adequate trial of other remedies has failed to reduce the crash frequency. • Warrant 8: Corridor Crash Experience — On highway corridors that demonstrate a history of similar crash types suitable to correction by providing corridor-wide consistency in turn-lane use. • Warrant 9: Vehicular Volume Warrant — At high-volume driveways (>100 trips per day) and all public street connections on high-speed highways (posted speed ≥ 45 mph) that satisfy the criteria in Table below. Minnesota Access Management Manual, Warrant for Left-Turn Lanes. Two-Lane Highway ADT Four-Lane Highway ADT Cross Street or Driveway ADT Turn Lane Requirement 1500 to 2999 3000 to 3999 4000 to 4999 5000 to 6499 ≥ 6500 ≥ 6500 3000 to 5999 6000 to 7999 8000 to 9999 10,000 to 12,999 ≥ 13,000 ≥ 13,000 > 1500 > 1200 >1000 >800 100 to 400 ≥ 400 Left-turn lane warranted Left-turn lane warranted Left-turn lane warranted Left-turn lane warranted Left-turn or bypass lane Left-turn lane warranted Highway AADT one year after opening. Posted speed 45 mph or greater. Mississippi http://www.gomdot.com/ Divisions/Highways/Reso urces/RoadwayDesign/pdf /Manual/2001/chapter06.p df Roadway Design Manual, Chapter 6 At-Grade Intersections/Interchanges 2001 6-4.02 Warrants for Left-Turn Lanes The accommodation of left turns is often the critical factor in proper intersection design. Left-turn lanes influence both the level of service and intersection safety. Exclusive left turn lanes should be provided: 1. at all median openings (crossovers) on divided urban and rural highways with a median wide enough to accommodate a left-turn lane; 2. at any unsignalized intersection on a two-lane highway which satisfies the criteria in Figures 6-48, 6-4C or 6-4D [appears to be based on Harmelink/NCHRP Report 279]; 3. at any unsignalized intersection on a four-lane undivided highway which satisfies the criteria in Figure 6-4E; 4. at any intersection where a capacity analysis determines a left-turn lane is necessary to meet the level-of-service criteria; or 5. at any intersection where the accident experience, existing traffic operations, adverse geometrics (e.g., restricted sight distance) or engineering judgment indicate a significant hazard related to left-turning vehicles.

C-11 State Material on Left-Turn Guidelines or Warrants Missouri http://epg.modot.mo.gov/i ndex.php?title=940.9_Au xiliary_Acceleration_and _Turning_Lanes Engineering Policy Guide August 2007 Dedicated left- and right-turn lanes are to be provided in situations where traffic volumes and speeds are relatively high and conflicts are likely to develop at public road intersections and driveways between through and turning traffic. Auxiliary lanes are an asset in promoting safety and improved traffic flow in such situations. Some major applications and considerations for the design of auxiliary lanes are as follows: • Installing a right-turn acceleration lane. These lanes allow entering vehicles (those that have turned right from a driveway or minor public road onto the major route) to accelerate before entering the through-traffic flow. Acceleration lanes are to be considered on roadway segments, intersections and driveways with high traffic volumes where speed differential could result in unacceptable conflicts and/or delay. Acceleration lanes may also be appropriate where crash experience indicates a problem with right turning, entering vehicles. The right-turn acceleration lane is to be sufficiently long to allow safe and efficient merge maneuvers. The design length, tapers and other features of right-turn acceleration lanes are to be guided by a traffic study. • Installing auxiliary left-turn lanes. Such lanes, installed in the roadway center, are intended to remove turning vehicles from the through traffic flow. This reduces the frequency of rear-end collisions at locations where there is considerable left- turn ingress activity, such as major driveways and minor public road intersections. Left-turn lane warrants are shown in the following figures [not included]. To use the figures, peak-hour traffic counts including directional splits, which may be obtained from district Traffic staff, will be required. In addition, the ITE Trip Generation Manual may be used as an estimate for peak-hour traffic counts. Planning can provide necessary growth rates for design-year analyses. o Left-turn lane guidelines for two-lane roads less than or equal to 40, 45, 50, 55, 60 mph o Left turn lane guidelines for four-lane undivided roadways o [these figures are similar to NCHPR Report 279/Harmelink’s graphs] • The use and design of auxiliary left-turn lanes are to be guided by a traffic study. In general, auxiliary left-turn lanes must be long enough to accommodate a safe deceleration distance and provide adequate storage for an expected peak-hour turning traffic queue. Refer to storage and deceleration lengths for additional information. Montana http://www.mdt.mt.gov/ot her/roaddesign/external/m ontana_road_design_man ual/13_intersection_at- grade.pdf http://www.mdt.mt.gov/p ublications/manuals.shtml Road Design Manual, Chapter 13: Intersections At-Grade December 2004 13.3.1.2 Guidelines for Left-Turn Lanes Exclusive left-turn lanes should be considered: 1. at all public intersections on all multilane urban and rural highways, regardless of traffic volumes; 2. at the free-flowing leg of any unsignalized intersection on a two-lane urban or rural highway which satisfies the criteria in Figures 13.3C, 13.3D, 13.3E or 13.3F [appears to be based on Harmelink/NCHRP Report 279]; 3. at any intersection where a capacity analysis determines a left-turn lane is necessary to meet the level-of-service criteria; 4. as a general rule on the major roadway, at any signalized intersection; 5. at high-volume driveway approaches which satisfy the criteria in Figures 13.3C, 13.3D, 13.3E or 13.3F; or 6. at any intersection where the accident experience, traffic operations, sight distance restrictions (e.g., intersection beyond a crest vertical curve), or engineering judgment indicates a significant conflict related to left-turning vehicles.

C-12 State Material on Left-Turn Guidelines or Warrants Nebraska http://www.dor.state.ne.us /roadway- design/pdfs/rwydesignma n.pdf Roadway Design Manual July 2006 Left turn treatments may be necessary on two-lane highways where traffic volumes are high and safety considerations are sufficient to warrant them. Left turn lanes should be provided on divided arterials at intersections and at other median breaks where left turn volumes and/or vehicle speeds are high. The “Intersection Channelization Design Guide” (Transportation Research Board, “Intersection Channelization Design Guide,” National Cooperative Highway Research Program Report 279, Washington, DC, 1994.) provides warrants and guidelines for auxiliary lane design. Nevada http://www.nevadadot.co m/reports_pubs/ Reviewed main website page but did not identify link that would appear to have the left-turn warrants/guidelines material

C-13 State Material on Left-Turn Guidelines or Warrants Nevada http://www.nevadadot.co m/business/forms/pdfs/Tr afEng_AccesMgtSysStan dards.pdf Access Management System and Standards July 1999 4.8 Left Turn Lane Requirements, Two Lane Unsignalized Roads Table 4.8 [table is a reproduction of the Green Book values for 40, 50 and 60 mph—Nevada’s table also includes data for 70 mph] lists the projected 20 year design-hour volumes and the operating speeds of traffic which necessitate the installation of left turn lanes. The traffic volumes to be considered in making this determination are the opposing (oncoming) traffic volumes, the advancing traffic volumes and the percent of advancing traffic which is turning left. Turn lanes may be required at lower volumes, by a traffic impact study or by the Department, to protect the traveling public. 4.9 Left Turn Lane Requirements, Four Lane, Undivided, Unsignalized Roads Table 4.9 [see below] lists the projected 20 year design-hour volumes of traffic which necessitate the installation of left turn lanes on multilane, undivided, unsignalized roads. The traffic volumes which are to be considered in making this determination are the opposing (oncoming) traffic volumes, the advancing traffic volumes, and the percent of advancing traffic which is turning left. Turn lanes may be required at lower volumes, by a traffic study or by the Department, to protect the traveling public. Table 4.9 Left-Turn Lane Requirements for Multilane Roads (Unsignalized) Opposing Volume (ddhv) Advancing Volume (ddhv) for left turn percentages of 5% 10% 20% 30% 800 600 400 200 100 140 220 350 530 650 110 160 250 380 480 80 120 190 290 350 70 100 160 250 310 4.10 Left Turn Lane Requirements, Four Lane, Divided, Unsignalized Roads Table 4.10 [see below] lists the projected 20 year design-hour volumes of traffic which necessitate the installation of left turn lanes on divided, unsignalized, multilane roads. The traffic volumes which are to be considered in making this determination are the opposing (oncoming) traffic volumes, the advancing traffic volumes, and the percent of advancing traffic which is turning left. Turn lanes may be required at lower volumes, by a traffic study or by the Department, to protect the traveling public. Table 4.10 Left-Turn Lane Requirements for Multilane Divided Roads (Unsignalized) Opposing Volume (ddhv) Advancing Volume (ddhv) for left turn percentages of 5% 10% 20% 30% 800 600 400 200 100 210 340 520 800 1000 150 240 380 580 720 110 180 290 440 550 100 150 250 390 480

C-14 State Material on Left-Turn Guidelines or Warrants New Hampshire http://www.nh.gov/dot/bu reaus/highwaydesign/nhhi ghwaydesignmanuals.htm Highway Design Manual Updated URL: http://www.nh.gov/dot/or g/projectdevelopment/hig hwaydesign/designmanual /index.htm Did not find discussion on left-turn lane warrants/guidelines New Jersey http://www.state.nj.us/tra nsportation/eng/document s/RDME/sect6E2001.sht m Roadway Design Manual, Section 6: At-Grade Intersections December 27, 2002 Updated URL: http://www.state.nj.us/tra nsportation/eng/document s/RDM/sec6.shtm Roadway Design Manual, Section 6: At-Grade Intersections March 2009 6.06 Median Left-Turn Lane 6.06.1 General A median lane is provided at an intersection as a deceleration and storage lane for vehicles turning left to leave the highway. Median lanes may be operated with traffic signal control, with stop signs, or without either, as traffic conditions warrant. Figure 6-T [not included] shows a typical median left-turn lane.

C-15 State Material on Left-Turn Guidelines or Warrants New Mexico http://nmshtd.state.nm.us/ main.asp?secid=11703 State Access Management Manual September 2001 Table 17.B-1. Criteria for Deceleration Lanes on URBAN TWO-LANE HIGHWAYS Turning Volume1 (veh/hr) LEFT-TURN DECELERATION LANE Minimum Directional Volume in the Through Lane (veh/hr/ln)2 ≤ 30 mph 35-45 mph 45-55 mph < 5 Not Required Not Required Not Required 5 510 450 330 10 390 330 210 15 320 250 150 20 270 200 120 25 230 160 100 30 200 130 Required 35 170 110 Required 40 150 Required Required 45 130 Required Required ≥ 46 Required Required Required Left-turn Deceleration Lanes are Required on Urban Two-Lane Highways for the following Left-turn Volumes: • ≤ 30 mph : 46 veh/hr or more • 35 to 40 mph : 36 veh/hr or more • 45 to 55 mph : 26 veh/hr or more Notes: 1. Use linear interpolation for turning volumes between 5 and 45 veh/hr. 2. The directional volume in the through lane includes through vehicles and turning vehicles. Table 17.B-2 Criteria for Deceleration Lanes on URBAN MULTI-LANE HIGHWAYS Turning Volume1 (veh/hr) LEFT-TURN DECELERATION LANE Minimum Directional Volume in the Adjacent Through Lane (veh/hr/ln)2 ≤ 30 mph 35-45 mph 45-55 mph < 5 Not Required Not Required Not Required 5 Not Required 490 420 10 420 370 300 15 360 290 220 20 310 230 160 25 270 190 130 30 240 160 110 35 210 130 100 40 180 120 Required 45 160 110 Required ≥ 46 140 Required Required Left-turn Deceleration Lanes are Required on Urban Multi- Lane Highways for the following Left-turn Volumes: • ≤ 30 mph : 56 veh/hr or more • 35 to 40 mph : 46 veh/hr or more • 45 to 55 mph : 36 veh/hr or more Notes: 1. Use linear interpolation for turning volumes between 5 and 55 veh/hr. 2. The directional volume in the adjacent through lane includes through vehicles and turning vehicles.

C-16 State Material on Left-Turn Guidelines or Warrants New Mexico (continued) Table 17.B-3 Criteria for Left-turn Deceleration Lanes on RURAL TWO-LANE HIGHWAYS Left- Turn Volume1 (veh/hr) LEFT-TURN DECELERATION LANE Minimum Directional Volume in Through Lane (veh/hr/ln)2 ≤ 30 mph 35-40 mph 45-55 mph > 55 mph < 5 Not Required Not Required Not Required Not Required 5 400 20 120 60 10 240 140 80 40 15 160 100 60 Required 20 120 80 Required Required 25 100 Required Required Required ≥ 26 Required Required Required Required Left-turn Deceleration Lanes are Required on Rural Two-Lane Highways for the following Left-turn Volumes: • ≤ 30 mph : 26 veh/hr or more • 35 to 40 mph : 21 veh/hr or more • 45 to 55 mph : 16 veh/hr or more • > 55 mph : 11 veh/hr or more Notes: 1. Use Linear interpolation of left-turn volumes between 5 and 25 veh/hr. 2. The directional volume in the through lane includes through vehicles and turning vehicles. Table 17.B-4 Criteria for Left-turn Deceleration Lanes on RURAL MULTI-LANE HIGHWAYS Left-Turn Volume1 (veh/hr) LEFT-TURN DECELERATION LANE Minimum Directional Volume in Through Lane (veh/hr/ln)2 ≤ 30 mph 35-40 mph 45-55 mph > 55 mph < 5 Not Required Not Required Not Required Not Required 5 400 20 120 60 10 240 140 80 40 15 160 100 60 Required 20 120 80 Required Required 25 100 Required Required Required 30 130 Required Required Required 35 110 Required Required Required ≥ 36 Required Required Required Required Left-turn Deceleration Lanes are Required on Rural Multi-Lane Highways for the following Left-turn Volumes: • ≤ 30 mph : 36 veh/hr or more • 35 to 40 mph : 26 veh/hr or more • 45 to 55 mph : 21 veh/hr or more • > 55 mph : 16 veh/hr or more Notes: 1. Use Linear interpolation of left-turn volumes between 5 and 35 veh/hr. 2. The directional volume in the through lane includes through vehicles and turning vehicles.

C-17 State Material on Left-Turn Guidelines or Warrants New York https://www.nysdot.gov/d ivisions/engineering/desig n/dqab/hdm/hdm- repository/chapt_05.pdf Highway Design Manual, Chapter 5—Basic Design August 23, 2006 The decision to construct left-turn lanes should consider: • The volume of left-turning traffic and the volume of opposing traffic. In some cases, capacity analysis may clearly indicate a need for left-turn lanes. Exhibit 9-75 in Chapter 9 of AASHTO’s A Policy on Geometric Design of Highways and Streets, 2004, includes traffic volume criteria to be considered in determining the need for left turn lanes along two-lane highways. • The accident history. An accident pattern of rear-end accidents involving queued left turners or vehicles turning left in front of opposing traffic is often mitigated by exclusive left-turn lanes. NYSDOT accident reduction factors show an average reduction of around 30% when a left-turn lane is installed and is an appropriate alternative to mitigate a left-turn accident problem. • The accident potential and the anticipated operating speeds (i.e., the possible severity of an accident). • Sight distance on the mainline affecting the ability to see a vehicle waiting to turn. • The construction costs. • The right of way impacts. North Carolina http://www.ncdot.org/doh /preconstruct/altern/value/ manuals/RDM2001/part1/ chapter9/pt1ch9.pdf Roadway Design Manual, Chapter 9: At Grade Intersections Found discussion on right-turn lane warrants but not left-turn lane warrants North Dakota Design Manual http://www.dot.nd.gov/ma nuals/design/designmanua l/designmanual.htm Reviewed main website page but did not identify link that would appear to have the left-turn warrants/guidelines material

C-18 State Material on Left-Turn Guidelines or Warrants Ohio http://www.dot.state.oh.us /Divisions/ProdMgt/Road way/roadwaystandards/Lo cation%20and%20Design %20Manual/400_jul06.pd f Location and Design Manual Section 400 Intersection Design July 2006 Updated URL: Location and Design Manual, Volume 1 Roadway Design (October 2010) http://www.dot.state.oh.us /Divisions/ProdMgt/Road way/roadwaystandards/Pa ges/locationanddesignman uals.aspx 401.6 Approach Lanes 401.6.1 Left Turn Lanes Probably the single item having the most influence on intersection operation is the treatment of left turn vehicles. Left turn lanes are generally desirable at most intersections. However, cost and space requirements do not permit their inclusion in all situations. Intersection capacity analysis procedures of the current edition of the Highway Capacity Manual should be used to determine the number and use of left turn lanes. For unsignalized intersections, left turn lanes may also be needed if they meet warrants as provided in Figures 401-5a, b, and c. The warrants apply only to the free- flow approach of the unsignalized intersection. [Some of the curves in Figures 401-5a and b are based on Green Book values. The graphs include several additional percent left turn values than included in the Green Book. Figure 401-5c is for four-lane highways with curves for divided and undivided roadways. The source for curves is not apparent.] Oklahoma Roadway Design Standards and Specifications http://www.okladot.state. ok.us/roadway/standards. htm Did not find any discussion on left-turn warrants Oregon ftp://ftp.odot.state.or.us/te chserv/roadway/web_dra wings/HDM/Rev_E_2003 Chp09.pdf http://egov.oregon.gov/O DOT/HWY/ENGSERVIC ES/hwy_manuals.shtml#2 003_English_Manual Highway Design Manual, 2003 English Manual, Chapter 9 (Intersection and Interchange Design) Left Turn Lanes Providing a left turn lane at an intersection will significantly improve the safety of the intersections. Eliminating conflicts between left turning vehicles decelerating or stopping and through traffic is an important safety consideration. A left turn lane must be provided at all non-traversable median openings. Left turn lanes may be installed at intersections meeting the installation criteria. The left turn lane installation criteria are different for signalized and unsignalized intersections. Refer to Section 9.3 Signalized Intersections, and Section 9.4 Unsignalized Intersections, for the appropriate siting criteria. 9.4 Unsignalized Intersections, Left Turn Lanes Left turn lanes at unsignalized intersections must meet the siting criteria to justify installation. Regardless of the funding source, the Region Traffic Engineer must approve all unsignalized channelized left turn lanes. The design should work with the Traffic Management Section in locations where left turn lanes are being considered. Left turn siting criteria has been established and is located in Appendix F along with a left and right turn lane siting example.

C-19 State Material on Left-Turn Guidelines or Warrants Oregon (continued) Appendix F — Left and Right Turn Lane Siting Criteria, Left Turn Lane Criteria Purpose A left turn lane improves safety and increases the capacity of the roadway by reducing the speed differential between the through and the left turn vehicles. Furthermore, the left turn lane provides the turning vehicle with a potential waiting area until acceptable gaps in the opposing traffic allow them to complete the turn. Installation of a left turn lane must be consistent with the access management strategy for the roadway. Left Turn Lane Evaluation Process 1. A left turn lane should be installed if criteria 1 (Volume), or 2 (Crash), or 3 (Special Case) are met, unless a subsequent evaluation eliminates it as an option, and 2. The Region Traffic Engineer must approve all proposed left turn lanes on state highways, regardless of funding source, and 3. The State Traffic Engineer shall review and approve all proposed left turn lanes at signalized intersection locations on State Highway System to ensure proper signal operation, prior to design and construction, and 4. Complies with Access Management Spacing Standards, and 5. Conforms to applicable local, regional, and state plans. I Criterion 1: Vehicular volume The vehicular volume criterion is intended for application where the volume of intersecting traffic is the principal reason for considering installation of a left turn lane. The volume criteria is determined by the Texas Transportation Institute (TTI) curves in Figure F-1 [The graph is similar to work done by Hawley and Stover, which is based on delay and conflict avoidance]. The criterion is not met from zero to ten left turn vehicles per hour, but indicates that careful consideration be given to installing a left turn lane due to the increase potential for accidents in the through lanes. While the turn volumes are low, the adverse safety and operations impact may require installation of a left turn. The final determination will be based on a field study.

C-20 State Material on Left-Turn Guidelines or Warrants Oregon (continued) II Criterion II: Crash experience The crash experience criterion is satisfied when: 1. Adequate trial of other remedies with satisfactory observance and enforcement has failed to reduce the accident frequency; and 2. A history of crashes of the type susceptible to correction by a left turn lane; and 3. The safety benefits outweigh the associated improvement costs; and 4. The installation of the left turn lane does not adversely impact the operations of the roadway. III Criterion 3: Special Cases 1. Railroad crossings — If a railroad is parallel to the roadway and adversely affects left turns, a worst case scenario should be used in determining the storage requirements for the left turn lane design. Other surrounding conditions, such as a drawbridge, could adversely affect left turns and must be treated in a similar manner. The left turn lane storage length depends on the amount of time the roadway is closed, the expected number of vehicle arrivals, and the location of the crossing or other obstruction. The analysis should consider all the variables influencing the design of the left turn lane, and may allow a design for conditions other than the worst case storage requirements, providing safety is not compromised. 2. Passing lane — Special consideration must be given to installing a left turn lane for those locations where left turns may occur and other mitigation options are not acceptable. 3. Geometric/safety concerns — Consider sight distance, alignment, operating speeds, nearby access movements, and other safety related concerns. 4. Non-traversable median — As required in the Median Policy, a left turn lane must be installed for any break in a non-traversable median. 5. Signalized intersection — Consideration shall be given to installing left turn lanes at signalized intersections. The State Traffic Engineer shall review and approve all proposed left turn lanes at signalized intersection locations on the state highway system. IV Evaluation Guidelines 1. The evaluations should indicate the installation of a left turn lane will improve the overall safety and/or operation of the intersection and the roadway. If these requirements are not met, the left turn lane should not be installed or, if already in place, not allowed to remain in operation. 2. Alternatives Considered — list all alternatives that were considered, including alternative locations. Briefly discuss alternatives to the left turn lane considered to diminish congestion/delays resulting in criteria being met. 3. Access management — address access management issues such as the long term access management strategy for the state roadway, spacing standards, other accesses that may be located nearby, breaks in barrier/curb, etc. 4. Land Use Concerns — Include how the proposed left turn lane address land use concerns and transportation plans. 5. Plan — Include a plan or diagram of proposed location of left turn lane 6. Operational requirements — consider storage length requirements, deceleration distance, desired alignment distance, etc. For signalized intersections, installing a left turn lane must be consistent with the requirements in the Traffic Signal Guidelines.

C-21 State Material on Left-Turn Guidelines or Warrants Oregon (continued) Volume Criterion Example Figure B-7 shows an unsignalized intersection with a shared through-right lane and a shared through-left lane on the Highway. The peak-hour volumes and lane configurations are included in the figure. The 85th percentile speed is 45 mph and the intersection is located in a city with a population of 60,000. Southbound: The southbound advancing volume is 555 (90 + 250 + 200 + 15) and the northbound opposing volume is 515 vehicles (the opposing left turns are not counted as opposing volumes). The volume for the y-axis on Figure B-6 is determined using the equation: y-axis volume = (Advancing Volume/Number of Advancing Lanes) + (Opposing Volume/ Number of Opposing Lanes) y-axis = (555/2 + 515/2) = 535 To determine if the southbound left turn volume criteria is met, use the 45 mph curve in Figure B-6, 535 for the y-axis, and 15 left turns for the x-axis. The volume criterion is not met in the southbound direction. Northbound: The northbound advancing volume is 555 (40 + 200 + 300 + 15) and the southbound opposing volume is 540 vehicles (the opposing left turns are not counted as opposing volumes). The volume for the y-axis on Figure B-6 is (555/2+ 540/2) = 548. To determine if the northbound left turn volume criteria is met, use the 45 mph curve in Figure B-6, 548 for the y-axis, and 40 left turns for the x-axis. The volume criterion is met in the northbound direction. Figure B-7. Oregon, Sample Intersection for Volume Criterion Example.

C-22 State Material on Left-Turn Guidelines or Warrants Pennsylvania ftp://ftp.dot.state.pa.us/pu blic/Bureaus/design/Pub1 3M/Chapters/Chap03.pdf ftp://ftp.dot.state.pa.us/pu blic/Bureaus/design/Pub1 3M/Chapters/Chap01.pdf Design Manual, Part 2 Highway Design, Chapter 3 Intersections, Chapter 1 General Design June 2007 H. Direct and Indirect Left Turns and U-Turns. The various design methods and arrangements to accommodate left-turn and U-turn movements are predicated on the design control dimensions (width of median and width of crossroad or street) and the size of vehicle used for design control. The necessity to turn left or to make a U-turn in the urban or heavily developed residential or commercial sectors represents serious problems with respect to safety and efficient operations. Although warrants for the use of speed-change lanes cannot be stated definitely, the general conclusions and considerations for their use are contained in Design Manual, Part 2, Chapter 1, Section 1.6. 1.6 ACCELERATION AND DECELERATION (SPEED-CHANGE) LANES The term speed-change lane, acceleration lane or deceleration lane, as used herein, applies broadly to the added pavement joining the traveled way of the highway with that of the turning roadway and does not necessarily imply a definite lane of uniform width. The warrants for the use of speed-change lanes cannot be stated definitely. However, based on observations and past experience, the following general conclusions have been made: 1. Speed-change lanes are warranted on high-speed and on high-volume highways where a change in speed is necessary for vehicles entering or leaving the through- traffic lanes. 2. All drivers do not use speed-change lanes in the same manner. 3. Use of speed-change lanes varies with volume, the majority of drivers using them at high volumes. 4. The directional type of speed-change lane consisting of a long taper fits the behavior of most drivers and does not require maneuvering on a reverse-curve path. 5. Deceleration lanes on the approaches to intersections that also function as storage lanes for turning traffic are particularly advantageous, and experience with them generally has been favorable. For additional information on speed-change lanes as applicable to intersections and interchanges, refer to the AASHTO Green Book, Chapter 9 and Chapter 10 and Design Manual, Part 2, Chapter 3 and Chapter 4. Rhode Island https://www.pmp.dot.ri.go v/PMP/DesktopDefault.as px?aM=udoc&oM=pages &c1P=info&c1I=833&ap pindex=0&appid=0&podi d=- 1&mth=1#pageAnchor6 Highway Design Manual 2008 Left-turn lanes not mentioned other than discussion on lane widths South Carolina Reviewed main website page but did not identify link that would appear to have the left-turn warrants/guidelines material

C-23 State Material on Left-Turn Guidelines or Warrants South Dakota http://www.sddot.com/pe/ roaddesign/docs/rdmanual /rdmch12.pdf Road Design Manual, Chapter 12: Intersections 2007 Left Turn Lane Criteria — Unsignalized Intersections Left turn lanes should be provided where through and turning volumes create an operational or a potential accident problem. Left Turn Lane Evaluation Process • A left turn lane should be installed if Criteria 1 (Volume), or 2 (Crash), or 3 (Special Cases) are met, unless a subsequent evaluation eliminates it as an option; and • The left turn lane complies with access management spacing standards; and • The left turn lane conforms to applicable local, regional and/or state design guidelines. [See the manual for specific criteria. South Dakota’s graph references the Oregon DOT procedures manual.] Tennessee http://www.tdot.state.tn.us /Chief_Engineer/assistant _engineer_design/design/ DGpdf/ENGLISH%20GU IDELINES.pdf TDOT Roadway Design Guidelines March 2006 Updated URL: Roadway Design Guidelines (August 2008) http://www.tdot.state.tn.us /Chief_Engineer/assistant _engineer_design/design/ DGpdf/DESIGN%20GUI DELINE.pdf To determine a warrant for and a required storage length, use the attached charts (Figures 2-13 through 2-16f) by M. D. Harmelink. (See also A Policy of Geometric Design of Highways and Streets 2004, Exhibit 9-75, page 685. This table is a condensed version of the Harmelink charts for two-lane highways.) The first chart applies to four-lane highways, all speeds [Figure 2-13]. The remaining charts are a function of speed and the percentage of lefts in the approaching traffic, and are applicable only to two-lane highways [Figures 2-14 to 2-16f]. Texas http://onlinemanuals.txdot .gov/txdotmanuals/rdw/rd w.pdf Roadway Design Manual October 2006 Left-turn lanes on two-lane highways at intersecting crossroads generally are not economically justified. For certain moderate or high volume two-lane highways with heavy left-turn movements, however, left-turn lanes may be justified in view of reduced road user accident costs. Figure 3-11 provides recommendations for when left-turn lanes should be considered based on traffic volumes. [Includes table with values similar to Green Book values.]

C-24 State Material on Left-Turn Guidelines or Warrants Utah http://www.udot.utah.gov/ main/f?p=100:pg:414701 0608516685::::V,T:,1498 Roadway Design Manual of Instruction May 2007 Highway Capacity Provide a level of service C or higher for a 20-year design in a rural area and a level of service D or higher for a 20-year design in an urban area. Decisions from going through the environmental process and implementing CSS may impact the decision of which level of service to provide. Design all elements of the roadway (including intersections) to the selected level of service. The need for channelization, left-turn lanes, etc., is directly related to the acceptable level of service. The Highway Capacity Manual presents a more thorough discussion of the level-of-service concept. It also supplies the analytical base for design calculations and decisions including capacity analysis. 7.18 Lane Types Auxiliary Lanes Under conditions of relatively high traffic volumes, traffic congestion problems can be significantly alleviated with auxiliary lanes to handle turning movements. In rural areas, consider left-turn lanes where there are 25 or more left-turn movements from the main highway in the peak hour. Vermont http://www.aot.state.vt.us/ progdev/standards/statabt a.htm Vermont State Design Standards October 22, 1997 Did not find information on left-turn lanes Virginia http://www.extranet.vdot. state.va.us/locdes/Electro nic%20Pubs/2005%20RD M/appendc.pdf Road Design Manual, Appendix C: Section C-1—Design Features, Virginia Department of Transportation, Location and Design Division July 2008 LEFT-TURN LANES As a general policy, left-turn lanes are to be provided for traffic in both directions in the design of all median crossovers on non-access controlled four-lane or greater divided highways using controls as shown in Figure C-1-1 [not included] and adjusted upward as determined by Figure C-1-1.1 or by capacity analysis for left-turn storage. Left-turn lanes should also be established on two-lane undivided highways where needed for storage of left-turn vehicles and/or prevention of thru-traffic delay as shown in Figure C-1-1 and adjusted upward as determined by Table C-1-2 [not included] and Figure C-1-1.2 [not included] through C-1-1.19 [not included] or by capacity analysis for left-turn storage. See Table C-1-2.1 [not included] for TRUCK ADJUSTMENTS. In general, when left-turn volumes are higher than 100 veh/hr, an exclusive left-turn lane shall be considered. Figures C-1-1.2 through C-1-1.19 provide warrants for left-turn storage lanes on two- lane highways based on 5 to 30 percent left-turn volumes and operating speeds of 40, 50, and 60 MPH. Table C-1-2.1provides the additional storage length required for 10 to 50 percent truck volumes. These figures were derived from Highway Research Report No. 211. This study was undertaken to provide consistent volume warrants for left-turn storage lanes at unsignalized intersections. Intersections with poor visibility and/or a bad accident record may require the designer to use engineering judgment when volume conditions alone do not warrant a storage lane. [COMMENTS: Warrants for left-turn storage lanes on two-lane highways table (Table C-1-2) is a reproduction of the Green Book table. HRR No. 211 = Harmelink’s paper]

C-25 State Material on Left-Turn Guidelines or Warrants Washington http://www.wsdot.wa.gov /Publications/Manuals/M2 2-01.htm Design Manual January 2006 (2) Left-Turn Lanes and Turn Radii Left-turn lanes provide storage, separate from the through lanes, for left-turning vehicles waiting for a signal to change or for a gap in opposing traffic. (See 910.07(4) for a discussion on speed change lanes.) Design left-turn channelization to provide sufficient operational flexibility to function under peak loads and adverse conditions. (a) One-Way Left-Turn Lanes are separate storage lanes for vehicles turning left from one roadway onto another. When recommended, one-way left-turn lanes may be an economical way to lessen delays and accident potential involving left-turning vehicles. In addition, they can allow deceleration clear of the through traffic lanes. When evaluating left-turn lanes, include impacts to all intersection movements and users. At signalized intersections, use a traffic signal analysis to determine whether a left-turn lane is needed and what the storage requirements are (see Chapter 850). At unsignalized intersections, use the following as a guide to determine whether or not to provide one-way left-turn lanes: • A traffic analysis indicates that a left-turn lane will reduce congestion. On two- lane highways, use Figure 910-12a, based on total traffic volume (DHV) for both directions and percent left-turn traffic, to determine whether further investigation is needed. On four-lane highways, use Figure 910-12b to determine whether a left- turn lane is recommended. • An accident study indicates that a left-turn lane will reduce accidents. • Restrictive geometrics require left-turning vehicles to slow greatly below the speed of the through traffic. • There is less than decision sight distance at the approach to the intersection. An HCM analysis may also be used to determine whether left-turn lanes are necessary to maintain the desired level of service. Determine the storage length required on two-lane highways by using Figures 910-13a through 13c. On four-lane highways, use Figure 910-12b. These lengths do not consider trucks. Use Figure 910-7 for storage length when trucks are present. [Figure 910-12a for two-lane unsignalized intersections appears to be a drawing using Green Book numbers. Figure 910-12b for four-lane unsignalized intersections may be from NCHRP Report 279/Harmelink.] West Virginia Design Directives Manual (October 2006) http://www.transportation. wv.gov/highways/enginee ring/DD/2006%20DD%2 0Manual%20MASTER.p df Left-turn lanes not mentioned Wisconsin Design Manual Left-turn lanes not mentioned Wyoming Road Design Manual Left-turn lanes not mentioned

D-1 APPENDIX D INTERVIEW QUESTIONS PLANNING 1. Do you provide left-turn treatments at unsignalized intersections? If so, please give some examples (e.g., bypass lanes, TWLTL, etc.). 2. Please provide details on your process for determining where left-turn treatments should be installed at unsignalized locations (e.g., who is involved, what affects the decision, etc.). 3. Does this process vary depending upon whether the left-turn treatment is needed for (a) a new development, (b) as part of a reconstruction project, or (c) as a spot improvement (e.g., based on crash experience)? If so, how? 4. When is inclusion of a left-turn accommodation considered on a project (e.g., during project scoping, during design, or after completion of a traffic study)? 5. What criteria do you use to determine whether a left-turn lane should be provided at an unsignalized location? (Please check all that apply.) ___ Harmelink’s guidelines ___ state DOT guidelines ___ local jurisdiction guidelines ___ none ___ other __________________ 6. Do the criteria vary if the site is in an urban or rural area? Or by speed limit? Or if the road is divided or undivided? If so, how? 7. If other left-turn accommodations (such as a bypass lane) are an option, what criteria do you use to determine where they should be provided? 8. If your local agency uses its own installation guidelines, how do they vary from state DOT guidelines? Would you be willing to provide us a copy? 9. Have your recent left-turn lane installations at unsignalized locations tended to be retrofit/restriping projects* or full-width build-out projects**? Why? 10. Have your recent left-turn lane installations at unsignalized locations tended to be for new developments, general street/highway improvements, or specific treatments to improve safety/reduce crashes? Why? 11. What methods and/or measures, if any, do you use to evaluate the effectiveness of unsignalized left-turn lanes or other accommodations (e.g., before-after study)? DESIGN 12. To what extent do you encourage positive offset of your left-turn lanes? 13. How do you determine the appropriate values for design elements of a left-turn lane (e.g., queue length, storage length, taper length, lane width, offset, sight distance, etc.)? 14. How do you determine the appropriate values for design elements of other left-turn accommodations?

D-2 15. What is the cost of a typical left-turn lane installation for a retrofit? For a full-width build-out? For other left-turn accommodations? 16. Is there a cross section on which you would not consider installing a left-turn lane at an unsignalized location? LEGAL/POLICY/FINANCE 17. Who pays for the installation of a left-turn treatment at (a) a driveway to a new development and (b) an intersection that is projected to be impacted by a new development (i.e., not at a driveway for the development)? 18. If construction of a left-turn lane would require obtaining property from a third party, how would that property be acquired and by whom? 19. For proposed developments, are the costs for left-turn lanes treated any differently than other mitigation needed to accommodate the additional site traffic? 20. Have there been any recent (i.e., within the last 5 years) changes in the decision-making process related to the installation of left-turn lanes or other accommodations at unsignalized intersections? Why? What were the changes? 21. Are you aware of the Supreme Court decisions relating to essential nexus or rough proportionality: Nollan v. California Coastal Commission and Dolan v. City of Tigard? 22. If yes, are you aware of any ramifications of these court cases as they relate to decisions on left-turn accommodations (including when to install and who pays)? POTENTIAL FUTURE APPLICATIONS 23. What lessons have you learned that will guide future installations of left-turn accommodations at unsignalized intersections? 24. Are you considering making any changes to policies on this topic in your jurisdiction? 25. Are there any regulatory/policy changes needed or anticipated at other levels to deal with issues related to the installation of left-turn accommodations at unsignalized intersections? * Could include restriping the roadway, changing parking regulations, or reducing or removing shoulders. ** Could include minor widening within the right-of-way or acquiring additional right-of-way.

E-1 APPENDIX E INTERVIEWS To help investigate the implementation of left-turn accommodations at unsignalized intersections, interviews were conducted of representatives from the following agencies/organizations: • State DOTs, • County governments, • City governments, and • Consultants. The 25 questions in the interview were structured into the following four topics: • Planning, • Design, • Legal/policy/finance, and • Potential future applications. The list of questions is included in Appendix D. This appendix presents the interview findings for each of the four topics listed above. Within the discussion of each topic, the results are discussed for the agencies/organizations included in the interviews (i.e., state DOT, county governments, city governments, and consultants). The findings are based on interviews with 11 state DOT representatives, 5 counties, 4 cities, and 2 consultants. The 11 state DOT interview participants represented nine state DOTs; two of the state DOTs reflected in the survey had two interview participants—one state DOT had two different districts represented, and one state DOT had two different divisions within the central office represented. PLANNING (QUESTIONS 1 to 11) State DOT Practices Process All state DOT interview participants indicated that left-turn treatments are provided at unsignalized intersections. The treatments used by all state DOTs are left-turn lanes and two- way, left-turn lanes. Several participants noted that TWLTLs are typically used in areas where there may be poor access control. Four of the state DOTs indicated they consider bypass or shoulder widening but noted these techniques are typically used when there are no other viable options. Three of the state DOTs noted that they generally would not use bypass or shoulder widening. From the responses, there does not appear to be a clear distinction between the terms “left-turn bypass lanes” and “shoulder widening” (may be called “shoulder lanes”). Bypass lanes

E-2 may be striped, but this does not appear to be universal. One state DOT noted that roundabouts are another treatment considered for dealing with left-turn movements. The process for determining where left-turn treatments should be installed at unsignalized locations, who is involved, and what affects the decision may vary depending upon the circumstances involved and whether the question relates to: • A developer seeking access onto the roadway system for a new development or major redevelopment, • The state DOT reevaluating the roadway condition as part of an improvement project, • A county- or city-initiated project that is being coordinated with the state DOT, or • A problem location identified by citizen complaints or its crash rate. One state DOT participant noted that the process must vary due to the limited timeframe the DOT staff has for reviewing driveway permit applications submitted by developers for access onto the state roadway system. A developer is often required to submit a traffic impact study for a major proposed development. This study is intended to assess the roadway network’s capability to accommodate the site- generated traffic. A major conclusion of the study is the mitigation to the roadways and intersections that is needed to handle the additional traffic. Left-turn treatments are often part of the identified mitigation. The DOT permitting staff or others involved in the study review play a critical role in the process to decide what treatment is needed. The design staff often leads the actual design effort that is done in accordance with the DOT’s design manual. For an improvement project, it typically would be the traffic engineering staff supporting the designers in helping to identify needed improvements based on existing or projected future conditions. Generally, the design staff plays a critical role in the process to decide what improvements should be made, including what left-turn treatments should be incorporated. When a project is initiated by the local county or city, extensive coordination of the participating agencies is involved. The project could be related to the roadway system or economic development. The DOT staff involved would vary depending upon the nature and extent of the project. When a problem location is identified, based on its poor safety or operating conditions, a team may be formed to investigate the causes and explore improvement options. The team considers the crash patterns, traffic volumes and movements, and physical conditions to help identify options for improving the traffic conditions. All state DOTS responded to the question regarding when left-turn accommodations are considered in a highway reconstruction or improvement project. Seven state DOTs first consider left-turn accommodations during the planning or scoping of a project. Two state DOTs consider left-turn accommodations during the design phase of a project.

E-3 Criteria Of the nine state DOTs represented in the interviews, three specified they use criteria based on the guidelines developed by Harmelink to determine whether a left-turn lane should be provided at an unsignalized location. A majority of the state DOTs referred to their own guidelines as the criteria used in the decision-making process. However, it appears that a number of these state DOT guidelines are derived from the research performed by Harmelink. Based on responses from state DOTs that had more than one participant, the criteria may vary by district; one district replied that the Harmelink guidance was used, whereas the other district noted that state DOT guidelines were used. The criteria used by a state DOT may also vary based on the need being addressed. One state DOT had participants from both the traffic engineering unit and the driveway permitting unit. The traffic engineering unit would use, in the decision- making process related to left-turn accommodations, information on crash rates and patterns as well as on traffic volumes and conditions. The permitting unit would use state DOT guidelines regarding left-turn lane warrants. The following are the responses to the question regarding whether the criteria for left-turn treatments vary if the site is in an urban or rural area, by speed limit, and/or whether it is located on a divided or undivided road: • All of the above are taken into account. • Speed limit does have an impact as well as whether the road is divided or not. • Yes, the roadway design manual addresses all these varying conditions. • The criteria for left-turn treatment may vary by the location setting. • The criteria do not vary based on urban or rural area, speed limit, or divided/undivided road. However, typically there is a greater need to investigate these treatments in more rural or urbanizing areas. • Criteria are essentially the same, but the design will vary. • Since the treatment will vary on a case-by-case basis, all of the above need to be considered. • Yes, they vary based on functional class and type of development. Also, speed for left- turn lanes on two-lane roads is considered. • Urban/rural location is not a factor. Speed is not an issue with current criteria. The divided/undivided cross section needs to be considered. • Urban/rural location is generally not a factor; however, urban areas are more difficult to deal with. High speed gets factored in qualitatively but is reflected in the deceleration length calculation. Divided/undivided road would have some effect. The following are the responses to the question regarding what criteria are used for identifying when other left-turn accommodations (such as a bypass lane) should be provided: • The respondent was not familiar with bypass lanes. • The respondent’s DOT does not really use bypass lanes. • Requirements are set forth in the roadway design manual. • The effectiveness of countermeasures, right-of-way requirements, cost, and schedule are considered. • Respondent’s DOT does not usually use other options.

E-4 • Respondent’s DOT does not use bypass lanes. The DOT occasionally builds TWLTLs for low-speed conditions. • Respondent’s DOT would generally not use bypass lanes but may use TWLTLs. • A bypass lane is not an option. The DOT has allowed a TWLTL in lieu of a left-turn lane where access is poorly controlled. The DOT has installed a median to block left-turn access. • The roadway design manual has provisions for when bypass lanes should be installed and also has a discussion of TWLTLs. • Bypass lanes are generally not used. TWLTLs are considered on a case-by-case basis. Most state DOT participants did not know whether there are local agencies that use their own installation guidelines. One respondent indicated that the majority of counties and local governments use the state DOT guidance. One respondent indicated that one city does not use state DOT criteria; instead it uses criteria it established based mainly on a roadway’s functional class. Applications According to a majority of the state DOTs represented in the interviews, recent left-turn lane installations at unsignalized locations have tended to include both retrofit/restriping projects and full-width build-out projects. The retrofit projects could involve restriping the roadway, changing parking regulations, or reducing or removing shoulders. The full-width projects could include minor widening within the right-of-way or acquiring additional right-of-way. One state DOT responded that its recent left-turn lane installations were retrofit projects. Three state DOTs indicated their recent left-turn lane installations were full-width projects. These three state DOTs offered more information regarding the full-width projects: • Most are the addition of left-turn lanes at mostly rural county road intersections as part of the DOT’s safety program. These were accomplished within the existing state right-of- way and involved minor widening of the pavement to accommodate the extra pavement width required to provide these turn lanes. • The recent left-turn lane projects are on two-lane conventional roadways that require widening to accommodate the greater pavement width. • In most cases full-width projects are done since retrofit has already been done where possible. There was a range of responses to the question regarding whether recent left-turn lane installations at unsignalized locations tended to be for new developments, general street/highway improvements, or specific treatments to improve safety/reduce crashes. The responses from state DOTs that had more than one participant varied. In one case, there were two districts responding, and in a second case there were two units from within the central office that were represented. For the state that had two districts responding, one district replied that left-turn lane installations have tended to be related to safety improvements, whereas the other district noted the installations appeared to be related to new developments. For the state that had two units responding, one unit replied that left-turn lane installations have tended to be

E-5 related to new developments, highway improvements, and safety improvements, whereas the other unit noted the installations appeared to be related to new developments. In response to the question regarding what methods and/or measures, if any, are used to evaluate the effectiveness of unsignalized left-turn lanes or other accommodations, the state DOTs noted that before and after studies or other evaluations are not generally performed. Safety improvements are the exception where evaluations are performed to identify the effectiveness of the improvement that was implemented. County Practices Process All five county interview participants indicated that left-turn treatments are provided at unsignalized intersections. All counties surveyed use left-turn lanes. Three of the counties indicated they use TWLTLs. One noted that TWLTLs exist as a retrofit option when access management is not practical. Two counties indicated they consider bypass or shoulder widening but noted these techniques are typically used when options are limited. The following are the responses to the question regarding the process used for determining where left-turn treatments should be installed at unsignalized locations, who is involved, and what affects the decision: • Access management guidelines are used for evaluating the needs for new developments as well as for identifying improvements to include in roadway improvement projects. The director of transportation who is also the county highway superintendent or a designee would make the decision. • Concurrency rules similar to those in Florida are part of the process. A number of different units would need to be involved, including design, traffic engineering, and property development. If the process is applied to a new development, coordination with the developer and associated traffic engineer is required. • The process has two main components—one related to a special fund for making improvements and a second relating to new developments desiring access. The default option is to add a left-turn lane in a county improvement project. Current policy requires a developer to build a left-turn lane at a site driveway. • The two main elements of the process relate to development review and traffic studies. When the need for a left-turn lane is identified based on a development review, the developer is usually conditioned to provide a dedicated left-turn lane. This is typically determined as part of a traffic impact study conducted by a traffic engineer for the developer. The need for a left-turn treatment may also be identified based on safety studies, traffic operational studies, and citizen complaints. These studies are a collaborative effort involving the traffic and capital divisions within public works. • The access proposal for a new development is reviewed by county staff for safety and consistency with access management guidelines.

E-6 The following are the responses to the question regarding whether the process used for determining where left-turn treatments should be installed at unsignalized locations varies for a new development, reconstruction project, and/or spot improvement: • Generally no. The same guidelines are used, and the same methodology is applied. For spot improvements, however, crashes need to be addressed. • It does. New developments result in right-of-way questions. How far do the decision makers want to go in pushing for a left-turn lane? There needs to be more of a give and take. For a reconstruction project, however, it would likely be a county project, and the county would acquire the right-of-way needed for the desired level of improvement. • Except for low-volume developments, a left-turn lane would be installed. Generally, the county would provide left-turn lanes or install a median to eliminate the left-turn movement. TWLTLs are not installed, and there are efforts to remove them and replace them with medians. • If a road is being reconstructed, the county will add medians or left-turn lanes consistent with what would be required from a private developer. All county participants responded to the question on when left-turn accommodations are considered in a highway reconstruction or improvement project. Four of the counties consider left-turn accommodations during the planning or scoping of a project. One county considers left- turn accommodations during the preliminary design phase of a project. Criteria The following are the responses to the question regarding the criteria used to determine whether a left-turn lane should be provided at an unsignalized location: • The county uses a combination of Harmelink, state DOT guidelines, and local jurisdiction guidelines. • Criteria are based on design considerations. • The county uses Harmelink guidance and crash data. • The county uses local jurisdiction guidelines. • The county uses Harmelink guidance and local access management guidelines based on Harmelink. The following are the responses to the question regarding whether the criteria for left-turn treatments vary if the site is in an urban or rural area, by speed limit, and/or whether it is located on a divided or undivided road: • Speed limit is considered. The county is more likely to require a left-turn lane on a higher-speed road. Since the county does not have many divided roadways, this is not an issue. • The criteria do not vary by speed limit, at least for minor approaches, and there is no difference for urban/rural areas. Since the county does not have many divided roadways, this is not an issue. • The default decision is to install a left-turn lane. Higher speed is better justification for a left-turn lane. If the roadway is divided, there would not be a median opening unless the developer is able to justify it. • Yes, the criteria are based on the Harmelink guidance

E-7 • No. However, if a proposed access or street is in a built-up area where roadway expansion is not practical or consistent with the corridor plan, then turn lanes may not be required. The following are the responses to the question regarding what criteria are used for identifying when other left-turn accommodations (such as a bypass lane) should be provided: • The county uses a combination of Harmelink and the access management guidelines with consideration of available gaps in the traffic stream. • The county has discussed bypass lanes but has not used them. • A bypass lane would be used only for maintenance. • A bypass lane is not considered an option. • Bypass lanes are only considered at T-intersections where turning volumes are not expected to cause queuing of multiple vehicles at any given time. Two of the counties use state DOT installation guidelines for left-turn treatments. One county has established its own set of guidelines. Applications According to a majority of the counties represented in the interviews, recent left-turn lane installations at unsignalized locations have tended to include both retrofit/restriping projects and full-width build-out projects. The retrofit projects could involve restriping the roadway, changing parking regulations, or reducing or removing shoulders. The full-width projects could include minor widening within the right-of-way or acquiring additional right-of-way. Two counties indicated their recent left-turn lane installations were full-width projects. All five county participants replied to the question regarding whether recent left-turn lane installations at unsignalized locations tended to be for new developments, general street/highway improvements, or specific treatments to improve safety/reduce crashes. Four counties replied that left-turn lane installations have tended to be related to a combination of new developments, highway improvements, and safety improvements. One county responded that the majority of installations appeared to be related to new developments. In response to the question regarding what methods and/or measures, if any, are used to evaluate the effectiveness of unsignalized left-turn lanes or other accommodations, the counties noted that before and after studies or other evaluations are not generally performed. One county indicated they do evaluations of improvements done as part of safety projects. One county noted they sometimes do operational analysis or field assessment after an improvement is completed. City Practices Process All four city interview participants indicated that left-turn treatments are provided at unsignalized intersections. All cities use left-turn lanes and TWLTLs. None of the cities indicated they use left-turn bypass lanes or shoulder widening.

E-8 The following are the responses to the question regarding the process used for determining where left-turn treatments should be installed at unsignalized locations, who is involved, and what affects the decision: • A lot of the decisions are based on traffic studies for new developments. • The traffic engineering staff is involved and make a decision based on left-turn volume, crashes, delay, and funding availability. • Due to limited funding, there are not a lot of projects. When funds are available for major projects, roads are typically built to a three-lane or five-lane cross section. • Typically, on divided roadways, left-turn treatments at unsignalized intersections are instituted in the design process by the roadway designer. On undivided roadways, left- turn treatments are typically added by the traffic engineer in response to increased left- turn traffic volumes and/or a high number of accidents. The following are the responses to the question regarding whether the process used for determining where left-turn treatments should be installed at unsignalized locations varies for a new development, reconstruction project, and/or spot improvement: • Yes. The process is essentially the same for new developments or reconstruction projects and is based primarily on operational analysis. For locations with higher crash rates, safety is the predominant focus. • No. • The decision to install a left-turn lane or TWLTL is based on volumes and crashes. Therefore, the process would be the same, although the funding would be an issue. • Yes, left-turn pockets are required for all new developments generating more than 250 vehicle trips per day on an existing divided roadway. For reconstruction projects, left-turn pockets are provided at all unsignalized intersections on divided roadways; on undivided roadways, dedicated left-turn lanes or TWLTLs are rarely provided at unsignalized intersections. For spot improvements on undivided roadways, left-turn treatments are typically added by the traffic engineer in response to increased left-turn traffic volumes and/or a high number of accidents. All city participants responded to the question regarding when left-turn accommodations are considered in a highway reconstruction or improvement project. Two participants stated that it was after a traffic study, one participant indicated during project scoping, and one participant indicated during the design process. Criteria The following are the responses to the question regarding the criteria used to determine whether a left-turn lane should be provided at an unsignalized location: • The city uses a combination of state DOT guidelines and local jurisdiction guidelines. • The city uses a combination of local jurisdiction guidelines, traffic impact studies, and staff site observations. • The city uses local guidelines that reflect the street’s average daily traffic and operating speed, the driveway’s volume, and the driveway’s left-turn ingress volume as a percentage of the street’s peak-period traffic volume. • The city uses local jurisdiction guidelines.

E-9 The following are the responses to the question regarding whether the criteria for left-turn treatments vary if the site is in an urban or rural area, by speed limit, and/or whether it is located on a divided or undivided road: • Each situation is considered individually. • No. • Operating speed is considered. • Typically, the criteria do not vary if the site is an urban versus rural area or by speed limit. However, they do vary based on whether the road is divided or undivided. The following are the responses to the question regarding what criteria are used for identifying when other left-turn accommodations (such as a bypass lane) should be provided: • Bypass lanes are not used. • Bypass lanes are used by the state DOT in rural areas. They are typically not used in a city. • Other left-turn treatments are not used. One of the cities uses state DOT installation guidelines for left-turn treatments since most of its routes are state routes. One city has no formal written guidelines. Applications Two of the cities indicated their recent left-turn lane installations at unsignalized locations were predominantly full-width projects. The full-width projects could include minor widening within the right-of-way or acquiring additional right-of-way. One city indicated that recent left-turn lane installations have tended to include both retrofit/restriping projects and full-width build-out projects. The retrofit projects could involve restriping the roadway, changing parking regulations, or reducing or removing shoulders. One city indicated that their recent installations at unsignalized intersections tended to be restriping projects because of costs. All four city participants replied to the question regarding whether recent left-turn lane installations at unsignalized locations tended to be for new developments, general street/highway improvements, or specific treatments to improve safety/reduce crashes. One city replied that left- turn lane installations have tended to be related to a combination of new developments, highway improvements, and safety improvements. One city responded that the majority of installations appeared to be related to new developments. One city indicated that the installations have been predominantly general improvement projects or safety projects. One city indicated their recent left-turn installations have been in response to increased left-turn traffic volumes and/or a high number of accidents. In response to the question regarding what methods and/or measures, if any, are used to evaluate the effectiveness of unsignalized left-turn lanes or other accommodations, two of the cities noted they do before and after studies; one does studies related to crashes and delay, and one does benefit/cost analysis.

E-10 Consultant Practices A consultant must follow the process and apply the criteria used by the jurisdiction for which (i.e., highway improvement) or in which (i.e., new development) a project is being done. Therefore, the responses from the two consultants to these questions of the survey are not included. DESIGN (QUESTIONS 12 TO 16) State DOT Practices All state DOT representatives indicated that they have guidelines that they use for determining the appropriate values for design elements of a left-turn lane (e.g., queue length, storage length, taper length, lane width, offset, sight distance, etc.). These guidelines could be contained within a range of documents including a road design manual, highway design manual, access management guide, project development design manual, corridor management policy, manual on uniform traffic control devices, or other design standards. Some of the state DOT representatives indicated they used state criteria for certain design elements and AASHTO guidance for others. State DOT representatives generally cited the same guidelines for design elements of other left- turn accommodations as they did for left-turn lanes. The representatives from two of the state DOTs—Kansas and Minnesota—cited guidelines that were developed for the design of left-turn bypass lanes. (Although left-turn lanes are allowed by the Kansas DOT, they are not commonly used.) In its Corridor Management Policy (Section C: Access Criteria), the Kansas DOT has guidelines based on posted highway speed for left-turn bypass lanes at T-intersections. The Minnesota DOT, in its Road Design Manual (Section 5-4.0: Rural Intersections), has two guidelines for the dimensions of left-turn bypass lanes at T-intersections and 4-leg intersections in rural areas. The Minnesota DOT in the Road Design Manual indicates that left-turn lanes are “the most effective and safe way to separate the left-turning from the through traffic streams”; left-turn bypass lanes should be considered only when there is no other alternative left-turn treatment. For the question regarding whether there is a cross section on which installing a left-turn lane at an unsignalized location would not be considered, several interview participants indicated there was no cross section on which they would not consider the installation of a left-turn lane. Other responses included: • The Road Design Manual needed to be checked for all design-related questions. • Installing the left-turn lane was more a function of site conditions than the cross section. • There was no cross section on which installing a left-turn lane would not be considered unless the location was proven to be economically or technically infeasible. • The DOT is reluctant to impact existing private development buildings. • Installing the left-turn lane is more of a design question. A left-turn lane as narrow as 10 ft in an urban, low-speed area may be acceptable.

E-11 • Yes, but it depends on truck volumes. The DOT may accept a 10-ft left-turn lane on low- speed roadways. • Yes, the DOT would not accept less than a 10-ft left-turn lane on a low-speed roadway and a 12-ft left-turn lane on a high-speed roadway. • A 12-ft-wide left-turn lane is preferred. No less than 11 ft on a rural facility is acceptable. There was a situation in an urban area where the left-turn lane was 11 ft wide and the through lane was 10 ft wide. About half of the interview participants did not know the cost of constructing left-turn treatments. They explained that where left-turn treatments are part of more comprehensive improvements, the cost of left-turn treatments is not computed separately. They further explained they would not be aware of the cost if a developer paid for the installation. The cost of constructing left-turn lanes on the major approaches at a four-leg intersection could range from $250,000 to $400,000 in urban areas with a shorter deceleration length. The cost in rural areas can be greater, ranging from $600,000 to $800,000, due to the higher speeds and greater length of the design elements. One respondent indicated the full-width build-out can be about $1,000,000 per intersection, including construction and right-of-way cost. Based on the responses from the interview participants, there appears to be limited usage of positive offset left-turn lanes. Reasons for the limited usage include cost and a lack of familiarity. Only one respondent indicated that the state DOT always tries to achieve this design. County Practices All county interview participants indicated that they have guidelines that they use for determining the appropriate values for design elements of a left-turn lane (e.g., queue length, storage length, taper length, lane width, offset, sight distance, etc.). These guidelines could be contained within a range of formal and informal documents; two use their state DOT design manuals, one uses internal guidelines that have been established, one uses the Harmelink methodology and available right-of-way, and one uses a combination of internal guidelines and other references. Three of the five county interview participants cited the same guidelines for design elements of other left-turn accommodations as they did for left-turn lanes—a state DOT design manual, internal guidelines, and a combination of internal guidelines and other references. Of the other two participants, one indicated this is not relevant, and one did not know how to reply. The following are the responses to the question regarding whether there is a cross section on which installing a left-turn lane at an unsignalized location would not be considered: • A left-turn lane would not be considered if there is inadequate width. • A left-tune lane would not be considered if there is insufficient sight distance related to horizontal or vertical curves. • The turn lane must be compatible with other geometry at the intersection. For example, an opposing left-turn lane should be present if the intersection has four legs.

E-12 Two of the interview participants provided an estimated cost of constructing left-turn treatments. The first indicated that his rule of thumb is between $75,000 and $100,000 for a left-turn lane at a T-intersection, $175,000 for two approaches at a four-leg intersection, and $300,000 for all four approaches. The second participant indicated that an estimated cost of several hundred thousand dollars is typical for left-turn lanes and that right-turn/bypass lanes may run about half the cost. Other participants explained that costs vary widely and that, where left-turn treatments are part of more comprehensive improvements, the cost of left-turn treatments is not computed separately. They further explained they would not be aware of the cost if a developer paid for the installation. Based on the responses from the interview participants, there appears to be limited usage of positive offset left-turn lanes. Emphasis is placed on aligning the opposing left-turn lanes. City Practices Three of the four interview participants replied to this question regarding the guidelines they use for determining the appropriate values for design elements of a left-turn lane (e.g., queue length, storage length, taper length, lane width, offset, sight distance, etc.). One indicated AASHTO guidance, one identified an internal document that is used along with field observations, and one indicated that these design elements are specified in the city’s pavement design manual. None of the respondents offered any insights into the guidelines used for design elements of other left- turn accommodations. The following are the responses to the question regarding whether there is a cross section on which installing a left-turn lane at an unsignalized location would not be considered: • The conditions at each intersection need to be considered, including sight distance, speed limit, other access points, etc. • An 11-ft width would be preferable. • No. One of the consultants had the following response to the question regarding whether there is a cross section on which installing a left-turn lane at an unsignalized location would not be considered: in a central business district where the distance between curbs is 30 ft, the sidewalk width could not be reduced, buildings would not allow the pavement to be widened, and additional right-of-way could not be acquired. One of the interview participants provided an estimated cost of constructing left-turn treatments of $300,000 to $1,000,000 for full intersection widening. Another participant indicated that the estimated cost for widening a two-block segment to include TWLTLs is $1,000,000. Another city indicated costs of $2,000 (retrofit [restriping]), $25,000 (full-width build-out [no existing median opening]), and $17,000 (full-width build-out [with existing median opening]). Regarding the usage of positive offset left-turn lanes, one participant indicated the usage of positive offset is strongly encouraged, one replied that it is done, and two indicated it is not used.

E-13 Consultant Practices The two consultant interview participants replied to this question regarding the guidelines they use for determining the appropriate values for design elements of a left-turn lane (e.g., queue length, storage length, taper length, lane width, offset, sight distance, etc.). One indicated the dimensions used without citing a reference. The second indicated that the values of the design elements are based on first considering sight distance and queue length. Regarding the guidelines used for design elements of other left-turn accommodations, one participant indicated the application of state standards based on speed. The second noted that the values of these design elements depend on the constraints involved that limit widening opportunities. One of the consultant participants indicated that the usage of positive offset left-turn lanes is generally limited due to right-of-way constraints. LEGAL/POLICY/FINANCE (QUESTIONS 17 to 22) State DOT Practices Who Pays for Left-Turn Treatment Of the nine state DOTs represented in the interviews, six indicated unequivocally that the developer of a new development would be responsible for paying for the installation of a left- turn treatment at a site driveway. Two indicated that a developer would “normally” or “typically” be responsible. However, the Minnesota DOT cannot require a developer to pay for the installation of a left-turn lane at a site driveway. The following provisions in its Access Management Manual (Chapter 4, page 16) would apply: For private driveway permits: Conditions of approval must fall within the reasonable exercise of the state’s police power. Conditions related to restricted movements and turn lanes will include only the design of the driveway entrance itself to restrict movements (e.g., pork chop design) and construction of warranted right-turn lanes provided there is adequate existing right-of-way to accommodate the right-turn lane. The applicant will be responsible for the costs associated with constructing these entrance design features. Other design features related to medians or turn lanes, warranted under the guidelines, will not be treated as conditions of the driveway permit, but will be recommended for consideration by the Local Governmental Unit as locally- initiated improvements to the trunk highway system. The Local Governmental Unit will need to determine whether it can assign any of the costs of these improvements, including the acquisition or dedication of additional right-of-way, to private property owners through assessments or related zoning and subdivision approvals. For an off-site improvement, such as an intersection that is projected to be impacted by a new development (but not a site driveway), there was a range of responses of who would be

E-14 responsible for paying for the installation of a left-turn treatment. Almost all of the state DOTs indicated that the developer would be expected to have some level of participation. Several indicated that the developer would be responsible for the mitigation. Some indicated that the developer would be responsible to pay a “fair share” or “proportional” cost. Several indicated that the local government agency (possibly city or county) would be expected to participate in the funding. One indicated that the developer would pay for the full cost of the treatment if the need for the off-site improvement on opening day was driven by the development. There was a wide range of responses to the question regarding who would obtain the property from a third party for the construction of a left-turn treatment. Several indicated the developer would need to take the lead in the coordination for obtaining the property. Two state DOT representatives indicated that it may be done in a number of ways that would vary depending upon the nature of the development and the improvement. Although most indicated that the DOT would leave it to other entities to coordinate this, one DOT responded that the agency could get involved. Another DOT indicated it would be up to the city in which the project is located. One DOT indicated it may take the lead in a rural area but would look for a city to take the lead in an urban area. All state DOT representatives that were interviewed indicated that the costs for left-turn accommodations are not treated differently than other mitigation needed to handle projected site- generated traffic. Policy Interview participants from three of the nine state DOTs indicated there had been changes in the past 5 years related to the decision-making process for the installation of left-turn lanes or other accommodations at unsignalized intersections. One indicated that the DOT has shifted away from using the Harmelink graphs, for establishing the need for a left-turn lane, to more general guidelines that include safety. Legal Six of the eleven state DOT interview participants were not aware of the two Supreme Court decisions relating to essential nexus or rough proportionality—Nollan v. California Coastal Commission and Dolan v. City of Tigard. Five were aware of the Supreme Court decisions. Two of the state DOTs reflected in the survey had two interview participants—one state DOT had two different districts represented, and one state DOT had two different divisions within the central office represented. For both states with two representatives, one respondent was aware of the Supreme Court decisions, and one was not. All of the interview participants who expressed an awareness of the Supreme Court decisions indicated they were aware of no ramifications of these court cases to state DOT decisions related to left-turn accommodations. Interview participants indicated that their policies are consistent with these decisions and already reflect rough proportionality, rational nexus, etc.

E-15 County Practices Who Pays for Left-Turn Treatment The interview participants from all five of the counties included in the interviews indicated the developer of a new development would be responsible for paying for the installation of a left- turn treatment at a site driveway. For an off-site improvement, such as an intersection that is projected to be impacted by a new development (but not a site driveway), the following are the responses to the question regarding who would be responsible for paying for the installation of a left-turn treatment: • The developer would be responsible. • It would depend on the specific concurrency rules. • It would be a political question. • It would depend on if the impact was direct or indirect. • It would vary: it could be the developer or a shared arrangement. The following are the responses to the question regarding who would obtain the property from a third party for the construction of a left-turn treatment: • The developer would need to take the lead in the coordination for obtaining the property for a large development. • The county would acquire the property at the cost of the developer. • The county may acquire the property if funds are available. • The developer would acquire the property if the property is required for a left-turn lane needed for a development permit; otherwise, the county would do so. • The county would acquire the property if it was a county project. The developer would acquire the property if the need was generated by a proposed development. Four of the five county interview participants indicated that the costs for left-turn accommodations are not treated differently than other mitigation needed to handle projected site- generated traffic. One noted that there could be a difference related to whether the improvement was being made along the property frontage (of the development site) versus where other properties are involved. Policy All five of the county interview participants indicated there were no major changes made in the past 5 years related to the decision-making process for the installation of left-turn lanes or other accommodations at unsignalized intersections. Legal Two of the five county interview participants were aware of the two Supreme Court decisions relating to essential nexus or rough proportionality—Nollan v. California Coastal Commission and Dolan v. City of Tigard. Three were not aware of the Supreme Court decisions.

E-16 Both of the interview participants who expressed an awareness of the Supreme Court decisions indicated they were aware of no ramifications of these court cases to county decisions related to left-turn accommodations. City Practices Who Pays for Left-Turn Treatment The interview participants from all four of the cities included in the interviews indicated the developer of a new development would be responsible for paying for the installation of a left- turn treatment at a site driveway. For an off-site improvement, such as an intersection that is projected to be impacted by a new development (but not a site driveway), the following are the responses to the question regarding who would be responsible for paying for the installation of a left-turn treatment: • The developer pays (indicated by two participants). • The developer pays if the treatment is required by ordinance. • The city pays, with some developer contribution. The following are the responses to the question regarding who would obtain the property from a third party for the construction of a left-turn treatment: • The developer is responsible. • The city is responsible, either by way of dedication or purchase. • It would depend. The developer or city may be responsible, depending upon whether the work would be considered a developer job or a city job. • The developer must purchase the property and dedicate it to the city. All four of the city interview participants indicated that the costs for left-turn accommodations are not treated differently than other mitigation needed to handle projected site-generated traffic. Policy Three of the four city interview participants indicated there were no major changes made in the past 5 years related to the decision-making process for the installation of left-turn lanes or other accommodations at unsignalized intersections. One noted that there are more innovative left-turn treatments as a result of successful applications elsewhere. Legal Three of the four city interview participants were not aware of the two Supreme Court decisions relating to essential nexus or rough proportionality—Nollan v. California Coastal Commission and Dolan v. City of Tigard. The third indicated a marginal familiarity based on a developer referencing one of the Supreme Court decisions to be released from mitigation responsibility.

E-17 Consultant Practices Who Pays for Left-Turn Treatment The two interview participants from consulting firms indicated the developer of a new development would be responsible for paying for the installation of a left-turn treatment at a site driveway. For an off-site improvement, such as an intersection that is projected to be impacted by a new development (but not a site driveway), the following are the responses to the question regarding who would be responsible for paying for the installation of a left-turn treatment: • The developer or government agency is responsible, on a case-by-case basis. • The developer is responsible, but sometimes a fair share is paid based on projected impacts. The following are the responses to the question regarding who would obtain the property from a third party for the construction of a left-turn treatment: • It is determined on a case-by-case basis and the willingness of the owner to sell to a developer. Owner willingness is typically a determining factor since only governmental agencies can use eminent domain. • Usually the developer would try to obtain the property, but if the developer is unsuccessful, the government agency with jurisdiction could get involved. Both of the consultants indicated that the costs for left-turn accommodations are not treated differently than other mitigation needed to handle projected site-generated traffic. Policy Both consultants participating in the survey indicated they were aware of no major changes made in the past 5 years related to the decision-making process for the installation of left-turn lanes or other accommodations at unsignalized intersections. Legal One of the two consultant interview participants was not aware of the two Supreme Court decisions relating to essential nexus or rough proportionality—Nollan v. California Coastal Commission and Dolan v. City of Tigard. The second indicated a limited familiarity. POTENTIAL FUTURE APPLICATIONS (QUESTIONS 23 TO 25) State DOT Practices The following are the responses related to lessons learned at the state DOT level that will guide future installations of left-turn accommodations at unsignalized intersections: • It is easier to install a left-turn lane as part of a project than it would be in the future as a retrofit.

E-18 • The application of a left-turn lane is preferred to the usage of a bypass lane. • Decisions should be made based on the corridor and not only on the specific location under study. • Left-turn warrants need to be updated. • Left-turn lanes should be installed even where left-turn volumes may be low. • Geometric constraints often dictate the dimensions of left-turn lanes. • DOTs would like to implement more positive left-turn offset. Generally, state DOT representatives indicated they were either unaware of or not considering changes to their policies related to left-turn accommodations at unsignalized intersections. Three interview participants indicated they either were initiating the process or in the long-term process to update their policies. State DOT representatives generally indicated they were either unaware of or not considering policy/regulatory changes at other levels of government related to left-turn accommodations at unsignalized intersections. One respondent noted the risk of going this route since it would provide legislators the opportunity to change what is in effect already. County Practices The following are the responses related to lessons learned at the county level that will guide future installations of left-turn accommodations at unsignalized intersections: • Counties need to address issues of safety, capacity, and feasibility, including right-of-way and environmental constraints. • There are pitfalls to allowing bypass lanes at four-way intersections and problems related to direct driveway access downstream of a bypass lane. • Counties need to get rid of TWLTLs and put in medians to manage access. • Counties need to be uniform in application of standards for a development. County representatives generally indicated they were not considering changes to their policies related to left-turn accommodations at unsignalized intersections. One indicated the county plans to review the existing guidelines as part of an upcoming transportation plan revision. County representatives generally indicated they were not aware of policy/regulatory changes at other levels of government related to left-turn accommodations at unsignalized intersections. One respondent noted the ability to charge “impact fees” to developers could change how such projects are handled. City Practices Three of the four city representatives surveyed did not provide a response to this question. The fourth responded that, based on implementing a few designs, he is more comfortable with identifying in what situations these treatments might be most effective.

E-19 The city representatives indicated they were not considering changes to their policies related to left-turn accommodations at unsignalized intersections. They also replied that they were not aware of policy/regulatory changes needed at other levels of government. Consultant Practices The two consultants that participated in the interviews did not provide any information related to these survey questions.

F-1 APPENDIX F LEGAL REVIEW THE IMPACT OF ESSENTIAL NEXUS AND ROUGH PROPORTIONALITY ON DEVELOPMENT CONDITIONS: A LEGAL REVIEW—“ONE THING IS CLEAR: THERE IS STILL A FUNDAMENTAL LACK OF CLARITY” This chapter addresses the following question: When a government seeks to fulfill a broad public objective such as safety and, in this project, left-turn accommodation, who should bear the costs—the developer who would be adding traffic to the roadway network or the general public? One of the critical issues that needs to be addressed in order to respond to this question is determining the application of the U.S. Supreme Court and other jurisdictional holdings on the takings clause as it may relate to development of property, in particular to determine under what circumstances the requirements of essential nexus and rough proportionality must be applied. The analytical framework presented in this chapter to address these issues attempts to make sense out of court holdings that are often in conflict, erratic, and confusing. In Part I, this chapter begins by looking at the big picture of land use/takings as articulated in the seminal cases of Nollan v. California Coastal Commission1 and Dolan v. City of Tigard.2 This part provides an overview of the development of the legal theories of essential nexus and rough proportionality. Part II attempts to discern the scope of what the Supreme Court meant in laying out guidelines for examining development exactions, and then focuses on small, individual pieces of the puzzle by exploring how essential nexus and rough proportionality have been applied to development exaction in various cases on the federal and state level. PART I. EXPLANATIONS OF THE LEGAL THEORIES OF ESSENTIAL NEXUS AND ROUGH PROPORTIONALITY A rudimentary review of the general principles of the takings doctrine as applied to regulatory takings3 1 Nollan v. Cal. Coastal Comm’n, 483 U.S. 825 (1987). is necessary to evaluate potential applications to left-turn accommodations and to describe the contradictory judicial milieu in which various jurisdictions find themselves. The Fifth Amendment to the United States Constitution provides that no land may be taken for public use without just compensation nor shall private property be taken for public use without just 2 Dolan v. City of Tigard, 512 U.S. 374 (1994). 3 Although outside the scope of this research, note should be made of a relatively recent U.S. Supreme Court case that negated a previous case, Agins v. City of Tiburon, 447 U.S. 255 (1980), on regulatory takings issues. In Lingle v. Chevron U.S.A. Inc., 544 U.S. 528 (2005), the Supreme Court noted the imprecision of Agins and spelled out four types of regulatory takings: “Twenty-five years ago, the Court posited that a regulation of private property ‘effects a taking if [it] does not substantially advance [a] legitimate state interest’ [Agins, 447 U.S. at 260]. The lower courts in this case took that statement to its logical conclusion, and in so doing, revealed its imprecision. Today we correct course. We hold that the ‘substantially advances’ formula is not a valid takings test, and indeed conclude that it has no proper place in our takings jurisprudence. In so doing, we reaffirm that a plaintiff seeking to challenge a government regulation as an uncompensated taking of private property may proceed under one of the other theories…by alleging a ‘physical’ taking, a Lucas-type ‘total regulatory taking,’ a Penn Central taking, or a land-use exaction violating the standards set forth in Nollan and Dolan,” Lingle, 544 U.S. at 540.

F-2 compensation.4 However, there has been a dramatic expansion of the definition of taking as related to the Fifth Amendment. The original focus was on per se takings. Federal governmental action unequivocally violated the takings clause if such action resulted in the permanent physical occupation of property or if the action denied the owner all economically beneficial use of her or his property.5 These actions triggered the “just compensation” requirement no matter whether the invasion was minor or whether the public purpose was greatly served.6 Thus, the Takings Clause protected private property from physical appropriation by the federal government.7 The first expansion of the Takings Clause came in 1922 when the Supreme Court gave “birth to our regulatory takings jurisprudence.”8 In Pennsylvania Coal Co. v. Mahon,9 Justice Holmes stated that “if regulation goes too far, it will be recognized as a taking.”10 Overly burdensome regulations, then, will be seen as a taking. What constitutes “overly burdensome,” however, was not defined by Holmes. Instead, he noted that it would be a question of degree.11 This lack of clarity and formulaic framework continued until the Supreme Court finally began to develop specific tests to determine when, in order to improve the public condition, the regulation went too far. Unfortunately, these attempts only resulted in more confusion. Today, the courts may use one of several tests: a two-pronged inquiry, the Nollan/Dolan rule; a three-part test, the Penn Central rule;12 and a per se rule, referred to as the Lucas rule.13 For purposes of this study, the focus is on the two-pronged test that was espoused in Nollan v. California Coastal Commission14 and Dolan v. City of Tigard.15 4 U.S. Const. Amend. V. 5 See Lucas v. S.C. Coastal Council, 505 U.S. 1003, 1015 (1992). 6 Id. 7 See, e.g., Robert Meltz et al., The Takings Issue: Constitutional Limits on Land Use Control and Environmental Regulation 129–30 (Island Press, 1999); J. Peter Byrne, Ten Arguments for the Abolition of the Regulatory Takings Doctrine, 22 Ecology L.Q. 89, 91–96 (1995); Joseph Sax, Takings and the Police Power, 74 Yale L.J. 36, 58–60 (1964); William M. Treanor, The Original Understanding of the Takings Clause and the Political Process, 95 Colum. L. Rev. 782 (1995); William M. Treanor, The Origins and Original Significance of the Just Compensation Clause of the Fifth Amendment, 94 Yale L.J. 694 (1985). 8 See, e.g., Tahoe-Sierra Preservation Council, Inc. v. Tahoe Reg’l Planning Agency, 535 U.S. 302, 325 (2002) (describing Pennsylvania Coal Co. v. Mahon, 260 U.S. 393 (1922) as the case “that gave birth to our regulatory takings jurisprudence.” However, some scholars assert that the Court’s regulatory takings jurisprudence began in the 19th century with the Court’s decision in Yates v. Milwaukee, 77 U.S. (10 Wall.) 497 (1870). See, e.g., Kris W. Kobach, The Origins of Regulatory Takings: Setting the Record Straight, 1996 Utah L. Rev. 1211, 1267–72 (1996). 9 Pennsylvania Coal Co. v. Mahon, 260 U.S. 393 (1922). 10 Id. at 415. 11 Id. at 416 (stating that “[w]e are in danger of forgetting that a strong public desire to improve the public condition is not enough to warrant achieving the desire by a shorter cut than the constitutional way of paying for the change. As we already have said, this is a question of degree—and therefore cannot be disposed of by general propositions.” 12 See Penn Cent. Transp. Co. v. City of New York, 438 U.S. 104, 124 (1978). “The factors include: the economic impact of the regulation on the claimant; the extent to which the regulation has interfered with distinct investment- backed expectations; and the character of the governmental regulation.” Id. 13 Lucas v. S.C. Coastal Council, 505 U.S. 1003, 1026–27 (1992). This case held that a regulatory taking occurs where a regulation completely devalued the land, that is, whether the regulations “deprived the land of all economically beneficial use. Id at 1015. See also supra note 3. 14 Nollan, supra note 1. 15 Dolan, supra note 2.

F-3 Nollan v. California Coastal Commission In the first case, the Nollans owned a beachfront lot that was located between two public areas, a beach and a park. Between the beach and the rest of the property was an 8-ft-high sea wall. Originally, the Nollans had leased the property with an option to buy; the option was conditioned on their promise to replace an existing house on the lot. The Nollans were required to and did seek the approval of the Coastal Commission to build the new, three-bedroom house on the site of the previous, substantially smaller house. The Commission recommended that the permit be granted on the condition that the Nollans provide a public easement for access across a portion of their property to protect the public’s “visual access” to the beach as well as prevent a psychological perception that there was no public access to the beach16 since the new house would be significantly larger. The Nollans argued that this easement constituted a taking.17 The U.S. Supreme Court, with Justice Scalia writing for the majority, stated that the Commission could have denied the permit outright if it would not have “interfered so drastically with the Nollans’ use of their property as to constitute a taking” and further a legitimate state interest.18 The Court reasoned that even though the government required the easement as a condition of the permit—as such, a land use regulation—and even though such land use regulation “does not affect a taking if it ‘substantially advance[s] a legitimate state interest’ and does not ‘deny an owner economically viable use of his land,’ ”19 the easement constituted a regulatory taking.20 In other words, protecting beach access by requiring a permit to build on the beach was a valid exercise of regulation.21 However, the Commission could not have simply acquired an easement from the Nollans without committing a taking. The Court looked at the connection between the conditions imposed and the interest being protected and held that the easement was not sufficiently related to “visual access”: “It is quite impossible to understand how a requirement that people already on the public beaches be able to walk across the Nollans’ property reduces any obstacles to viewing the beach created by the new house. It is also impossible to understand how it lowers any ‘psychological barrier’ to using the public beaches, or how it helps to remedy any additional congestion on them caused by construction of the Nollans’ new house.”22 16 Nollan, 483 U.S. at 827–28. Therefore, the easement condition was invalid since it sought to give the public direct access to the beach rather than focusing on the original condition, visual access: “The evident constitutional propriety disappears, however, if the condition substituted for the prohibition utterly fails to further the end advanced as the justification for the prohibition. When that essential nexus is eliminated, the situation becomes the same as if California law forbade shouting fire in a crowded theater, but granted dispensations to those willing to contribute $100 17 Id. at 829. For further takings analysis, see, e.g., John A. Humbach, A Unifying Theory for the Just- Compensation Cases: Takings, Regulation, and Public Use, 34 Rutgers L. Rev. 243, 254–62 (1982); Richard A. Epstein, Takings: Private Property and the Power of Eminent Domain 331–33 (1985); Andrea L. Peterson, The Takings Clause: In Search of Underlying Principles Part I—A Critique of Current Takings Clause Doctrine, 77 Cal. L. Rev. 1301, 1301 (1989); and Andrea L. Peterson, The Takings Clause: In Search of Underlying Principles Part II—Takings as Intentional Deprivations of Property without Moral Justification, 78 Cal. L. Rev. 53, 55 (1990). 18 Nollan, 483 U.S. at 836. 19 Id. at 834. 20 Nollan, 483 U.S. at 841–42. 21 Id. at 834–35. 22 Nollan, 483 U.S. at 838–39.

F-4 to the state treasury.”23 Thus, there would be no taking as long as the development condition furthered the same purpose as the development ban: the exaction—the easement—did not advance the same purpose as the development prohibition—psychological blocking of the view and, in turn, the perception of blocking access to the beach—because the exaction’s purpose lacked an “essential nexus”—that is, an essential connection—to the harm the building would cause.24 Important to this holding is the underlying principle for the essential nexus test, that of the unconstitutional conditions doctrine, implicitly invoked in Nollan,25 which states that the government cannot condition receipt of a benefit on the applicant’s foregoing a constitutional right. Therefore, even if the government is not constitutionally required to grant a particular privilege or benefit, once it offers that benefit, it may not condition the offer upon the recipient’s surrender or waiver of a constitutional right.26 Nollan left unanswered questions about the degree of relationship the government must prove for an exaction to be judicially sustained. Still, in practical terms, state and local governmental bodies must establish an essential nexus when implementing exactions.27 This became explicit in the subsequent exaction case, Dolan v. City of Tigard. Dolan v. City of Tigard The Dolan case involved the practice of zoning in relation to property rights. In that case, the U.S. Supreme Court refined its analysis of an exaction by looking at the degree of exaction in relation to the burden caused by the development.28 The holding in the case established limits on the ability of governmental agencies to use land-use regulations to require property owners to make unrelated public improvements. As background, Dolan owned and operated a plumbing store in Tigard, Oregon. She applied for a land-use variance in order to expand the store and pave its parking lot. The city planning commission granted only conditional approval: Dolan was required to dedicate land along an adjacent creek and develop a pedestrian and bicycle pathway to relieve traffic congestion. Dolan asserted that the requirements were not related to the proposed development and thus constituted an uncompensated taking. The Supreme Court held that a government agency “may not require a person to give up a constitutional right—here the right to receive just compensation when property is taken for public use—in exchange for a discretionary benefit conferred by the government where the benefit sought has little or no relationship to the property.”29 The Court applied a two-prong test. The first prong, established in Nollan, must be to determine whether an “essential nexus” exists between the legitimate state interest to be advanced by the restriction on development and the condition exacted by the government.30 23 Id. at 837–39. The Court found that the purpose 24 Id. at 837 (quoting J.E.D. Assoc. v. Atkinson, 432 A.2d 12, 14–15 [N.H. 1981]). 25 See Nollan, 483 U.S. at 836–37; see also Kathleen M. Sullivan, Unconstitutional Conditions, 102 Harv. L. Rev. 1413, 1463 (1989). 26 See, e.g., Epstein, Unconstitutional Conditions, State Power, and the Limits of Consent, 102 Harv. L. Rev. 4, 6–7 (1988). 27 See 483 U.S. at 836–37. 28 See Dolan, 512 U.S. at 374. 29 Dolan, 512 U.S. at 385. 30 Nollan, 483 U.S. at 837.

F-5 of the permit conditions fell within the legitimate state interest, that of protecting against flooding and preventing congestion, both interests of which would have been served had the development permit application been denied.31 Then, the Court added a second prong to the analysis. A determination must be made of whether there is a “rough proportionality” between the legitimate interest the government asserts and the actual impact on the landowner’s proposed property use, that is, rough proportionality between the costs or harm the development would impose and the cost imposed by the exaction on the developer.32 Citing the “reasonable relationship” test adopted by a majority of state courts, the Court held that although it did not adopt the phrase itself because it can too easily be confused with the term “rational basis” used to describe the minimal level of scrutiny under the Equal Protection Clause of the 14th Amendment, “a term such as ‘rough proportionality’ best encapsulates what we hold to be the requirement of the Fifth Amendment. No precise mathematical calculation is required, but the city must make some sort of individualized determination that the required dedication is related both in nature and extent to the impact of the proposed development.”33 Unfortunately, the proportionality requirement of Dolan was left ill defined. Interestingly, in a footnote the majority stated that determination of rough proportionality lies with the municipality.34 In this case, the city satisfied the nexus test but provided insufficient findings to demonstrate rough proportionality. Thus, governmental entities must make individualized determinations to justify the regulations they impose upon developers to ascertain “rough proportionality.” PART II: REVIEW OF OTHER LAW AS APPLIED TO DEVELOPMENT CONDITIONS IN LIGHT OF NOLLAN/DOLAN The question in this study is: from a legal perspective, when a government seeks to fulfill a broad public objective such as safety and in this project, left-turn accommodation, who should bear the costs—the developer who would be adding traffic to the roadway network or the general public? If the developer should shoulder the cost, frequently used alternatives include the development exaction and the development agreement, the latter of which may include an exaction. What the case law reveals, however, is a history of adjudicating outcomes that are erratic, conflicting, and inconsistent as will be seen in the following discussion. Government use of development exactions is not a recent phenomenon. However, due to a variety of factors in the last half century including urban sprawl, local governments have increasingly relied upon exactions to finance new development projects. Cities and towns use development exactions to offset the public burden of new development, with developers paying their fair share of public costs generated.35 In development exactions, a property owner is forced to relinquish something of value, e.g., land or money, in exchange for a building permit. Thus, development exactions occur when a local governmental entity conditions the grant of a development permit on the developer agreeing to pay money, provide materials or services, or dedicate land.36 31 Id. The cash payment form of an exaction is typically referred to as a monetary 32 Dolan, 512 U.S. at 391. 33 Id. 34 Id. at 391 n. 8. 35 See Donald G. Hagman, Public Planning and Control of Urban and Land Development 904 (2d ed. 1980). 36 See, e.g., Michael H. Crew, Development Agreements after Nollan v. California Coastal Commission, 483 U.S. 825 (1987), 22 Urb. Law. 23, 23–24 (1990).

F-6 exaction. One type of monetary exaction is the impact fee, a one-time financial assessment imposed as a condition of development37 that offsets the municipality’s capital expenditures used to construct public off-site infrastructure directly connected with or required because of the new development, for example. The fee is usually determined by legislatively adopted rates.38 Using this alternative, municipalities hope to avoid triggering the Nollan/Dolan level of scrutiny. However, even though a municipality uses monetary exactions based on legislative enactment rather than adjudicative, discretionary determinations, these alternatives may still be subjected to the Nollan/Dolan tests. The various development agreements that occur as the result of negotiations between a developer and the local agency appear to be effective and typically are not subject to the Nollan/Dolan mandates.39 The purpose of these agreements is to expressly limit municipalities from applying new requirements, ordinances, or other land-use changes to ongoing developments. The basis for development agreements is found in contractual law; they are voluntary, and that suggests that the agreed-upon terms are binding: “The developers…bargain out of their own choice, and municipalities should be able to exact as much as they can….”40 Courts have recognized the benefits of development agreements. For example, in Queen Anne’s Conservation, Inc. v. County Comm’rs of Queen Anne’s County,41 the Maryland Court noted that the purpose of the development agreement “is to vest development rights in the landowner or developer in exchange for the dedication and funding of public facilities.”42 Further, the municipality is not “granting the landowner the right to develop nor imposing conditions on such development”43 but rather “is promising to protect the developer’s investment by not enforcing any subsequent land use regulation that may burden the project.”44 37 See Ronald H. Rosenberg, The Changing Culture of American Land Use Regulation: Paying for Growth with Impact Fees, 59 SMU L. Rev. 177, 205–06 (2006). The commitment by a developer that adequate infrastructure to serve the development project will be in place in exchange for a guarantee that a developer may proceed for a specified period of time is beneficial to both parties. These agreements, however, must be based on specific legislative authorization—enabling legislation—that allows municipalities to negotiate these agreements. Such legislation should, for example, establish “minimum procedural requirements for the 38 Id. at 205. 39 Daniel J. Curtin, Jr., and Jonathan D. Witten, Windfalls, Wipeouts, Givings, and Takings in Dramatic Redevelopment Projects: Bargaining for Better Zoning on Density, Views, and Public Access, 32 B.C. Envt’l Aff. L. Rev. 325, 340–41 (2005). 40 Catherine Lockhard, Note, Gaining Access to Private Property: The Zoning Process and Development Agreements, 79 Notre Dame L. Rev. 765, 786 (2004). 41 855 A.2d 325 (Md. 2004). 42 Id. at 327. See also Lockhard, supra note 40 for a discussion of the Queen Anne’s Conservation case. 43 Lockard (citing David L. Callies and Julie A. Tappendorf, Unconstitutional Land Development Conditions and the Development Agreement Solution: Bargaining for Public Facilities after Nollan and Dolan, 51 Case W. Res. L. Rev. 663, 695 [2001]). 44 Id.

F-7 consideration and adoption of development agreements, and spell out the legal effects of such agreements with regard to subsequent land use regulations.”45 California was the first state to promulgate development agreement legislation. 46 At least a dozen other states have put statutes on their books to authorize local governments to enter into development agreements.47 According to one commentator, as it now stands, the majority of jurisdictions “do not subject development agreement cases to regulatory takings analysis…. Until the Supreme Court decides a development agreement case, municipalities will likely continue to use development agreements to exact more than would be available with development conditions.”48 Thus, the question of “who pays” would be answered according to the negotiation between the municipality and the developer. Because there is actual bargaining, exactions and the terms thereof will be as stringent as the parties allow through their negotiation. As the use of various types of development exaction fees expanded, in other jurisdictions litigants looked to the judicial system to determine the constitutionality of these regulatory takings. In response, the U.S. Supreme Court departed from its long tradition of deference to state police power actions and limited local governments’ constitutional ability to impose development conditions. In the Nollan/Dolan decisions discussed in Part I, the Court created a heightened scrutiny of development exactions.49 The Court did not explicitly provide in those cases an answer to the question of whether monetary exactions are subject to the Nollan/Dolan heightened scrutiny. However, a few days after the Dolan decision, a case addressing the issue of whether monetary exaction must be subjected to the Nollan/Dolan requirements was granted certiorari. In this case, Ehrlich v. City of Culver City,50 a property owner and developer, Ehrlich, requested from the city permission to construct a condominium complex on land he already owned and on which he had operated recreational facilities. Concerned about the loss of the recreational facility, the city approved the plaintiff’s application conditioned upon payment of a $280,000 fee to be used for additional public recreational facilities.51 The landowner challenged this monetary exaction as an unconstitutional taking. Ultimately, the Court remanded the case to the U.S. Court of Appeals for the Sixth Circuit for further consideration. On remand, the Court of Appeals reaffirmed its ruling in favor of Culver City. This caused the Supreme Court of California to grant review.52 45 Michael B. Kent, Jr., Forming a Tie That Binds: Development Agreements in Georgia and the Need for Legislative Clarity, 30 Environs Envtl. L. & Pol’y J. 1, 31 (2006). California’s Supreme Court reversed. The Court noted that the issue was properly analyzed within the statutory framework of California’s Mitigation Fee Act. The Court then held that the city met its burden of showing an essential nexus between the permit 46 Cal. Gov’t Code 65,865 (West 1997). 47 See, e.g., Lockard supra note 40 (citing Ariz. Rev. Stat. Ann. 9-500.05 [West 1996 and Supp. 2002] [amended 1997]; Colo. Rev. Stat. Ann. 24-68-101 [West 2001]; Fla. Stat. Ann. 163.3220 [West 2000]; Haw. Rev. Stat. Ann. 46-123 [Michie 2001]; Idaho Code 67-6511A [Michie 2001]; La. Rev. Stat. Ann. 33:4780.22 [West 2002]; Nev. Rev. Stat. Ann. 278.0201 [Michie 2002]; N.J. Stat. Ann. 40:55D-45 [West 1991]; Or. Rev. Stat. 94.504 [2001]; S.C. Code Ann. 6-31-10 [Law. Co-op. Supp. 2002]; Va. Code Ann. 15.2-2303.1 [Michie 2003]; Wash. Rev. Code Ann. 36.70B.170 [West 2003]). 48 Id. (citing John J. Delaney, Development Agreements: The Road from Prohibition to “Let’s Make a Deal!”, 25 Urb. Law. 49, 55 [1993]). 49 See Dolan, 512 U.S. at 394–96; Nollan, 483 U.S. at 838–39. 50 Ehrlich v. City of Culver City, 512 U.S. 1231 (1994). 51 Id. at 434–35. 52 Ehrlich v. City of Culver City, 911 P.2d 429, 433 (Cal. 1996).

F-8 condition and the public impact of the proposed development by demonstrating the connection between the rezoning necessary to construct the condominium complex and the imposition of the fee that would be expended in support of recreational purposes, as a means of mitigating that loss: In our view, the intermediate standard of judicial scrutiny formulated by the high court in Nollan and Dolan is intended to address just such indicators [leveraging] in land use “bargains” between property owners and regulatory bodies—those in which the local government conditions permit approval for a given use on the owner’s surrender of benefits which purportedly offset the impact of the proposed development. It is in this paradigmatic permit context—where the individual property owner-developer seeks to negotiate approval of a planned development—that the combined Nollan and Dolan test quintessentially applies.53 After carefully reviewing the Nollan/Dolan decisions as well as state court decisions, however, the Court reasoned: Under this view of the constitutional role of the consolidated “essential nexus” and “rough proportionality” tests, it matters little whether the local land use permit authority demands the actual conveyance of property or the payment of a monetary exaction. In a context in which the constraints imposed by legislative and political processes are absent or substantially reduced, the risk of too elastic or diluted a takings standard—the vice of distributive injustice in the allocation of civic costs—is heightened in either case.54 The Court held that “the city failed to show the required rough proportionality between the magnitude of the fiscal exaction and the effects of the proposed development, and the court therefore remanded for further findings.”55 In other words, the record was insufficient to sustain the city’s assertion that Ehrlich should pay a mitigation fee of $280,000 as a condition for approval of his request. Ultimately, the Court looked at whether the exaction imposed arose from legislatively formulated development assessments that would, therefore, deserve deference and be subject to a lesser standard of scrutiny than Nollan/Dolan requirements because “the heightened risk of the ‘extortionate’ use of the police power to exact unconstitutional conditions is not present.”56 Nollan/Dolan requirements would apply, then, if the exaction was pursuant to a discretionary administrative condition. In particular, the applicability of Nollan/Dolan to adjudicative versus legislative governmental actions is troubling. One commentator notes that it is possible to separate legislation into two categories: “1) legislation that gives a municipality discretion in applying it built into the very language of the ordinance; and (2) legislation that allows for no discretion and is, by default, 53 Id. at 438. 54 Id. at 444. 55 Id. 56 Id.

F-9 more even-handed in nature.”57 The commentator looked at two cases in the same Oregon Court of Appeals, Rogers Machinery, Inc. v. Washington County58 and Dudek v. Umatilla County59 to “illustrate the dichotomy between the different types of legislation.” In the first case, Rogers Machinery, the court refused to apply the Nollan/Dolan requirement to a traffic impact fee (TIF) assessed against developments: Indeed, nearly all proposed development that conceivably would burden the street and arterial infrastructure within the county is subject to the mandatory fee. Calculation of the fee is, likewise, nondiscretionary. To be sure, because different uses vary in the burden they place on street and arterial infrastructure, the legislation uses a classification scheme to adjust the fee based on those differences. Such classification is common for any number of generally applicable legislative tax, fee, and other assessments; legislative classifications do not render the scheme adjudicatory or discretionary. The TIF ordinance is, in short, precisely the kind of detailed and uniformly applied legislative assessment scheme that courts in other states hold does not fall within the express reach or the implicit rationale of Dolan’s heightened scrutiny test.60 A similar challenge was at issue in Dudek, a case involving application of an ordinance to road widening. This time, the court applied the Nollan/Dolan analysis because there was an increased risk that the governmental entity could abuse its discretion in applying the legislation.61 The court reasoned: As we discussed in Rogers Machinery, Inc., a significant consideration in the determination of whether Dolan’s rough proportionality test applies to a particular government action is whether the action is taken pursuant to a legislatively adopted scheme that applies to a broad class of property and whether the action involves the exercise of discretion. Also pertinent is whether the government action requires adjudication to determine whether and how a government regulation applies to particular property. As we explained in Rogers Machinery, Inc., one of the main reasons those considerations are pertinent is that, when government action is taken pursuant to a legislatively adopted standard that was not adopted to apply to particular properties or development but rather to a broad class of property, and when there is no need for adjudication or the exercise of discretion at the time that the standard is applied to a particular property, there is far less danger of a governmental entity attempting to use its power to extort unconstitutional conditions from persons seeking governmental approval of a specific proposal…. [T]he practical reality is that application of this ordinance to 57 Jane Needleman, Exploring Exactly When Nollan and Dolan Should be Triggered, 28 Cardozo L. Rev. 1563 (2006). 58 Rogers Machinery, Inc. v. Washington County, 45 P.3d 966 (Or. Ct. App. 2002). 59 Dudek v. Umatilla County, 69 P.3d 751 (Or. Ct. App. 2003). 60 Id. at 982. 61 See Needleman, supra note 57.

F-10 a particular case requires a significant exercise of discretion. It effectively requires an adjudication in each case.62 Thus, the court focused on whether there was potentially a significant amount of discretion even if the ordinance was legislative. Common sense would dictate that a court would find it difficult to look through the layers of broad, discretionary application of legislative action and instead would prefer clear-cut delineations of application. For those courts that distinguish Dolan on this basis, it is the assumption that Dolan applies to adjudicative, discretionary regulations. Other state courts, though, including Illinois, Ohio, Oregon, Texas, and Washington apply the Nollan/Dolan heightened scrutiny to monetary exactions.63 As well, some state courts have determined that impact fees, regardless of whether they came about through legislation or other means, trigger the Nollan/Dolan heightened scrutiny analysis. For example, in the case of Town of Flower Mound v. Stafford Estates Ltd. P’ship,64 the Texas Supreme Court, in deciding whether there was a compensable taking where a town conditioned its approval of the development of a residential subdivision on the developer’s rebuilding an abutting road, noted: Conditioning government approval of a development of property on some exaction was a compensable taking unless the condition: (1) bore an essential nexus to the substantial advancement of some legitimate government interest and (2) was roughly proportional to the projected impact of the proposed development. There was no important distinction between a dedication of property to the public and a requirement that property already owned by the public be improved. The town failed to relate discounted traffic impact fees to the impact of developments on traffic. Conditioning development on rebuilding the road with concrete and making other changes was simply a way for the town to extract from the developer a benefit to which the town was not entitled. The exaction was thus a taking for which the developer was entitled to be compensated.65 In Home Builders Ass’n of Dayton & the Miami Valley v. City of Beavercreek,66 the Supreme Court of Ohio addressed the issue of whether an ordinance that allows impact fees payable by developers of real estate to aid in the cost of new roadway projects is constitutional. The Court had consistently held that under specific legislative language, municipalities had the authority to impose exactions “provided that the municipality is not statutorily forbidden from doing so, and the exactions meet constitutional standards.”67 The Court reviewed Nollan/Dolan requirements, the requirements noted in other states’ court opinions, and those articulated in its own decisions and held that the impact fee ordinance did not violate the Ohio Constitution or the United Station Constitution. 62 69 P.3d at 756. 63 See, e.g., Daniel J. Curtin, Jr., and W. Andrew Gowder, Jr., Exactions Update: When and How Do the Dolan/Nollan Rules Apply?, 35 Urb. Law. 729, 733–38 (2003). 64 Town of Flower Mound v. Stafford Estates Ltd. P’ship, 135 S.W.3d 620, 641 (Tex. 2004). 65 Id. 66 Home Builders Ass’n of Dayton & the Miami Valley v. City of Beavercreek, 729 N.E.2d 349, 356 (Ohio 2000). 67 Id.

F-11 In Benchmark Land Co. v. City of Battle Ground,68 the developer applied to the City of Battle Ground for a development permit. As a condition of approving the application, the City required the developer to make half-street improvements to a street adjoining the development. The Court applied the Nollan/Dolan analysis to the condition and held that the City failed to show that the condition was proportional to the development’s impact on the street. The Court noted that “the condition advanced a legitimate state interest, improving the public roads, and the condition did not deny the developer all economically viable use of its land…. The City did not restrict development of property, but required developer to address a problem that existed outside the development property, an adjoining street in need of improvement. There was a necessary connection between the condition and the public problem.”69 Courts in Arizona, Colorado, Kansas, and Maryland have held that Nollan/Dolan standards do not apply to monetary exactions. For example, in Krupp v. Breckenridge Sanitation Dist.,70 a case heard by the Supreme Court of Colorado, the developers of a townhouse project contended that the required “plant investment fee” (PIF) that was assessed for their project was an unconstitutional taking. The Court found that the PIF was “a valid, legislatively established fee that was reasonably related to respondent district’s interest in expanding its infrastructure to account for new development, and that [the Sanitation District’s] specific PIF assessment on [the developer’s] project was fairly calculated and rationally based. As such, the PIF did not fall into the narrow category of charges that were subject to a Nollan/Dolan takings analysis.”71 The distinction, then, was that the PIF was legislatively mandated as opposed to an adjudicative, discretionary determination. This distinction has played a crucial role in several of the cases already discussed, and the courts have taken note of the Ehrlich rationalization in which a greater threat of extortion exists with discretionary determinations. The Court also held that the PIF was “purely a monetary assessment rather than a dedication of real property for public use.”72 In Homebuilders Ass’n of Central Arizona v. City of Scottsdale, 73 the Arizona Supreme Court focused on the legislative versus adjudicative distinction and held that the ordinance in question was legislative act and therefore “came to the court cloaked with a presumption of validity…. Land use regulations of general application will be overturned by the courts only if a challenger shows the restrictions to be arbitrary and without a rational relation to a legitimate state interest…. Development or impact fees are presumed valid as exercises by legislative bodies of the power to regulate land use.”74 The Court held that Dolan was inapplicable for two reasons. First, the reasonableness of the amount of the fee was not an issue, so no question arose as to whether the fee was roughly proportional to the burden imposed on the community; therefore, a Dolan test was not applicable.75 68 Benchmark Land Co. v. City of Battle Ground, 14 P.3d 172, 175 (Wash. Ct. App. 2000). Second, Dolan is distinguishable because Dolan involved a city’s adjudicative decision to “impose a condition tailored to the particular circumstances of an individual case” as the Chief Justice took care to explain. The present case involved a legislative 69 Id. 70 Krupp v. Breckenridge Sanitation Dist., 19 P.3d 687, 695-98 (Colo. 2001). 71 Id. 72 Id. at 697. 73 Home Builders Ass’n v. City of Scottsdale, 187 Ariz. 479 (1997). 74 Id. 75 Id.

F-12 decision by the city. Although agreeing that the question has not been settled by the U.S. Supreme Court, the court referred to Ehrlich v. City of Culver City,76 stating that the case: dramatically illustrates the differences between the two exactions. In Ehrlich, the city had imposed an individually tailored $280,000 mitigation fee as a condition of approving a rezoning request. On remand from the United States Supreme Court for reconsideration in light of Dolan, the California Supreme Court held the record insufficient to show that the fee was roughly proportional to the public burden of replacing recreational facilities that would be lost as a result of rezoning Ehrlich’s property. The California court suggested that the Dolan analysis applied to cases of regulatory leveraging that occur when the landowner must bargain for approval of a particular use of its land…. The risk of that sort of leveraging does not exist when the exaction is embodied in a generally applicable legislative decision.77 The case was appealed to the U.S. Supreme Court, but the Court denied the writ of certiorari.78 That is unfortunate because at least two members of the Supreme Court have noted in a dissenting opinion in denying a writ that the legislative versus adjudicative distinction is not valid. The case, Parking Ass’n of Georgia v. City of Atlanta,79 came before the U.S. Supreme Court at about the same time as the Homebuilders decision by Arizona’s Supreme Court. In Parking, the Atlanta City Council passed an ordinance requiring certain existing surface parking lots to include landscaped areas “equal to at least 10% of the paved area and to have at least one tree for every eight parking spaces. The ordinance covers some 350 parking lots; petitioners estimate that compliance with the landscaping requirements will cost approximately $12,500 per lot, for a total of $4,375,000. Additionally, parking lot owners will lose revenue due to lost parking spaces and lost advertising dollars: the trees allegedly will obscure existing advertising signs and cause petitioners to lose contracts worth about $1,636,000.”80 Justice Thomas, joined by Justice O’Conner, wrote in the dissent denying the writ that a taking may occur when there is a legislative act. Justice Thomas maintained that the case should have been heard: It is not clear why the existence of a taking should turn on the type of governmental entity responsible for the taking. A city council can take property just as well as a planning commission can. Moreover, the general applicability of the ordinance should not be relevant in a takings analysis. If Atlanta had seized several hundred homes in order to build a freeway, there would be no doubt that Atlanta had taken property. The distinction between sweeping legislative takings and particularized administrative takings appears to be a distinction without a 76 Id. (citing Ehrlich v. City of Culver City, 12 Cal. 4th 854, 911 P.2d 429 [Cal. 1996], cert. denied, 136 L. Ed. 2d 218, 117 S. Ct. 299 [1996]). 77 Home Builders, supra note 73 at 486. See also Jonathan M. Block, Limiting the Use of Heightened Scrutiny to Land-Use Exactions, 71 N.Y.U. L. Rev. 1021, 1024 n.154 (1996). 78 Homebuilders Ass’n v. City of Scottsdale, supra note 73. 79 Parking Ass’n of Georgia v. City of Atlanta, 264 Ga. 764, 450 S.E.2d 200 (Ga. 1994), cert. denied, 515 U.S. 1116, 132 L. Ed. 2d 273, 115 S. Ct. 2268, 2269 (1995). 80 Id.

F-13 constitutional difference…. The lower courts should not have to struggle to make sense of this tension in our case law. In the past, the confused nature of some of our takings case law and the fact-specific nature of takings claims has led us to grant certiorari in takings cases without the existence of a conflict.81 Where, as here, there is a conflict, the reasons for granting certiorari are all the more compelling.82 In the federal Circuit Courts, the Ninth and Tenth both apply a narrower Nollan/Dolan standard. In an interesting twist, the Ninth Circuit includes California.83 Thus, California has two opposing rules, once again demonstrating the confusion and conflict involved in exactions and the regulatory takings issues. These cases point out that it is clear there is no clarity. Determinations of applicability of the Nollan/Dolan standards will be adjudicated on a case-by-case basis. However, what does seem to reveal some clarity is the fact that if an impact fee or other monetary exaction is legislatively mandated, then it may not fall within the development exactions that triggered the Nollan/Dolan requirements. How do these oftentimes inconsistent, conflicting, and erratic decisions apply to governmental actions such as providing for safe highways and, in particular, left-turn accommodations? Who should bear the cost when a government seeks to fulfill a broad public objective such as safety and, in this project, left-turn accommodation—the developer who would be adding traffic to the roadway network or the general public? The answers can apparently only be derived by looking at specific jurisdictions, and even then, as in the case of California and the Court of Appeals in Oregon, there may be conflicting applications. Rather than having the consistency that is anticipated and expected from the judicial system, in the area of development conditions cases, courts appear to engage in ad hoc, individualized, factual determinations of whether a governmental entity has exceeded constitutional limitations. Thus, the burden of proof has shifted so that the government must jump through an additional hurdle; the justification for exaction must be clear. No longer is the presumption of validity of legislatively mandated exactions unassailable. This burden could result in potentially negative results, for example, for environmentalists who propose strong, enforceable regulations to protect cities and rural areas from expansive development or for those who see that government actions are critical in providing safety for its citizenry. The resolution of this conflict is critical, particularly when growth management schemes are becoming increasingly commonplace with government entities searching for ways to fund such programs. Shifting the cost of development including “smart growth” or “sustainable growth” to property owners through exactions such as impact fees may appear a logical alternative. From the mishmash of court opinions, however, a governmental entity 81 Id (citing Tigard, 512 U.S. at ___ [slip op., at 8] observing that certiorari was granted because the Oregon Supreme Court allegedly had misapplied Nollan v. California Coastal Comm’n, 483 U.S. 825, 97 L.Ed. 2d 677, 107 S. Ct. 3141 [1987]). 82 Parking Ass’n of Georgia v. City of Atlanta, 515 U.S. 1116. 83 See, e.g., Lockhard, supra note 40 at 742.

F-14 should, at the very least, have these fees buttressed by clear legislative language. Who should pay? Unless the facts of the case are clear cut, the outcome is unpredictable even within specific jurisdictions. Results will be more predictable if agencies have the documented authority to manage access. This authority could be achieved in a number of different ways (municipal codes for access management, administrative rules, etc.). Otherwise, there are too many hazy cases and multiple factors so that predicting the outcome is impossible. The U.S. Supreme Court must articulate an applicable rule to bring consistency and predictability in order to answer the question of “who should bear the burden, who should pay.”