Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Research Home
Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-08-051
Date: June 2008

Surrogate Safety Assessment Model and Validation: Final Report

PDF Version (3.39 MB)

PDF files can be viewed with the Acrobat® Reader®

Chapter 3. Theoretical Validation (Continued)

CASE STUDIES

Various types of intersections have been implemented and evaluated in three simulation systems: VISSIM, TEXAS, and AIMSUN.

The goal of this portion of the validation effort was not to compare the results of the simulation model with traffic at a comparable real-world location. Hence, no calibration effort was necessary or performed in this study. Reasonable driver behavior was verified, and appropriate control measures were used to avoid gridlock during high-volume test cases. As such, for all the intersection designs, default driving behavior models and parameters were applied for each simulation model. The same underlying simulation parameters were used for each comparison case to maintain comparability.

Eleven comparison cases were executed among the three simulation systems as follows:

TEXAS Cases

  • Case 1: Signalized, four-leg intersection with permitted left turn versus protected left turn.
  • Case 2: Signalized, four-leg intersection with and without left turn bay.
  • Case 3: Signalized, four-leg intersection with and without right turn bay.

VISSIM Cases

  • Case 4: Signalized, four-leg intersection with leading left turns versus lagging left turns.
  • Case 5: Signalized, four-leg intersection versus offset T-intersection;
  • Case 6: Diamond interchange with three-phase timing versus four-phase timing.
  • Case 7: Single point urban interchange (SPUI) versus diamond interchange.

AIMSUN Cases

  • Case 8: Signalized, four-leg intersection with left turns versus signalized intersection with median U-turns.
  • Case 9: Signalized, four-leg intersection versus single roundabout.
  • Case 10: Signalized, three-leg, T-intersection versus single roundabout with three legs.
  • Case 11: Diamond interchange versus double roundabout.

Three sets of traffic volumes (low, medium, and high) were applied for each intersection design, and timing plans were designed to ensure no over-saturation would occur.

Case 1: Conventional Four-Leg Intersection with Permitted Left Turn Versus Protected Left Turn (TEXAS)

There are two basic alternative designs for left turns: protected and permitted. Protected left-turn design allocates an exclusive phase for left turn only, which will make the left-turn maneuvers have fewer conflict events with the opposing through traffic. Permitted left-turn design allows vehicles to make a left turn during the through traffic green phase and provides no specific green phase for left-turn only. This logic applies mostly to traffic conditions with low left-turn volumes. When the left-turn volumes become higher, there have been more conflict events when drivers begin accepting smaller gaps to cross the intersection.

According to the crash prediction models for four-leg signalized intersection, the existence of left-turn phase will result in lower crash frequency. Thus, it is hypothesized that protected left turn should have lower predicted conflict frequency than permitted left turn when other network parameters remain the same. Also, it would be reasonable to expect that severity values of the surrogate measures would be less critical for protected left turns versus permitted left turns.

Intersection Description for Case Study

The intersection used to test the alternative traffic control logic for left turns is four-legged intersections with three through lanes in the main travel directions and two through lanes on the side-street approaches to the intersection, as shown in figure 29. All left-turn bays are 76.25 m (250 ft) long.

Figure 29. Screen Capture. Intersection Geometry for Testing Control Logic. This is a screen capture of an intersection model in TEXAS. The intersection is four legged with three through lanes in the main travel direction and two through lanes in the side-street approaches. All approaches have left-turn bays that are 76.25 m (250 ft) long.

Figure 29. Screen Capture. Intersection Geometry for Testing Control Logic.

Table 8 lists the traffic volumes applied for each approach of the intersection. Fixed time traffic control is applied in this test. Figure 30 through figure 35 provide the key timing plan parameters for each testing scenario.

Table 8. Case 1 Service Flow by Each Approach.
Approach Southbound Northbound Eastbound Westbound
L TH R L TH R L TH R L TH R
Phase#
(Permitted)
4 4   8 8   2 2   6 6  
Phase#
(Protected)
7 4   3 8   5 2   1 6  
Low Volume 100 350 50 100 350 50 60 210 30 60 210 30
Medium Volume 240 400 160 240 400 160 180 240 180 180 240 180
High Volume 300 1050 150 300 1,050 150 240 840 120 240 840 120

Note: L, TH, and R correspond to vehicles proceeding left, through, or right at the intersection.

Figure 30. Illustration. Timing Plan for Permitted Left Turn in Low Volumes. This is a screen capture of the timing plan for the tested permitted left turn case in low volume condition. All through movements (phases 2, 4, 6, and 8) have 20 seconds as the split time.

Figure 30. Illustration. Timing Plan for Permitted Left Turn in Low Volumes.

Figure 31. Illustration. Timing Plan for Protected Left Turn in Low Volumes. This is a screen capture of the timing plan for the tested protected left turn case in low volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 21 seconds. The split time for E-W left turn movements (phase 1 and phase 5) is 8 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 11 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 31. Illustration. Timing Plan for Protected Left Turn in Low Volumes.

Figure 32. Illustration. Timing Plan for Permitted Left Turn in Medium Volumes. This is a screen capture of the timing plan for the tested permitted left turn case in medium volumes condition. The split time for the E-W movements (phase 2 and phase 6) is 20 seconds, and the split time for S-N movements (phase 4 and phase 8) is 25 seconds.

Figure 32. Illustration. Timing Plan for Permitted Left Turn in Medium Volumes.

Figure 33. Illustration. Timing Plan for Protected Left Turn in Medium Volumes. This is a screen capture of the timing plan for the tested protected left turn case in medium volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 21 seconds. The split time for E-W left turn movements (phase 1 and phase 5) is 15 seconds. The split time for the S-N left- turn movements (phase 3 and phase 7) is 19 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 33. Illustration. Timing Plan for Protected Left Turn in Medium Volumes.

Figure 34. Illustration. Timing Plan for Permitted Left Turn in High Volumes. This is a screen capture of the timing plan for the tested permitted left turn case in high volumes condition. The split time for the E-W movements (phase 2 and phase 6) is 20 seconds, and the split time for S-N movements (phase 4 and phase 8) is 25 seconds.

Figure 34. Illustration. Timing Plan for Permitted Left Turn in High Volumes.

Figure 35. Illustration. Timing Plan for Protected Left Turn in High Volumes. This is a screen capture of the timing plan for the tested protected left turn case in high volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 43 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 23 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 27 seconds. The split time for S-N through movements (phase 4 and phase 8) is 37 seconds.

Figure 35. Illustration. Timing Plan for Protected Left Turn in High Volumes.

Data Analysis and Comparison Results

Ten replications were performed for each design case and the resulting output trajectory data were analyzed by SSAM. F-test and t-tests were applied to compare surrogate measures of safety and the aggregations of those measures.

Table 9 through Table 13 list the values of all surrogate measures of safety and corresponding t test results for different types of aggregations with the low-speed events and crash data excluded (TTC ≤ 0 and MaxS ≥ 16.1 km/h (10 mi/h)).

Table 9. Case 1 Comparison Results for Total Conflicts.
Total
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  PER PRO PER PRO PER PRO
Low volume Mean
4.7
4.7
16.1
14
25.7
22.4
Variance
6.2
5.3
11.2
11.3
12.0
17.2
t-value(95%), difference (%)
0
1.399
1.932
Medium volume Mean
13.7
12.8
33.1
32.6
53.7
63.3
Variance
21.6
8.2
21.4
24.7
39.8
113.8
t-value(95%), difference (%)
-0.522
0.233
-2.45, -17.88%
High volume Mean
108.3
174.2
184.1
489.1
309.5
1208.6
Variance
138.7
183.1
332.8
189.2
408.1
1344.5
t-value(95%), difference (%)
-11.618 , -60.85%
-42.216, -165.67%
-67.916, -290.5%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table illustrates a counterintuitive result. The design with the protected left turn has, on average, more total conflicts than the case with the permitted left turn for medium- and high-traffic volumes. The following tables explain this result by breaking the total results into a result for each conflict type.

Table 10. Case 1 Comparison Results for Only Crossing Conflicts.
Crossing
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
PER PRO PER PRO PER PRO
Low volume Mean
3.1
0.3
5.2
0.4
6.3
1
Variance
3.4
0.5
4.4
0.5
4.2
1.1
t-value(95%), difference (%)
4.490, 90.32%
6.865, 92.31%
7.250, 84.13%
Medium volume Mean
5.2
0.6
8.1
0.8
10.1
1.8
Variance
7.3
0.7
5.0
0.8
10.1
0.8
t-value(95%), difference (%)
5.143, 88.46%
9.558, 90.12%
8.89, 92.808
High volume Mean
23.3
6.4
27.5
9.2
33.8
15
Variance
46.9
6.7
67.2
11.1
71.5
16.0
t-value(95%), difference (%)
7.299, 72.53%
6.543, 66.75%
6.355, 55.62%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table indicates what is expected to happen from adding a protected left-turn phase; that the total crossing conflicts are reduced for all levels of traffic volume. This indicates that SSAM, in its most basic form, is a valid indicator of safety.

Table 11. Case 1 Comparison Results for Rear-End Conflicts.
Rear End
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  PER PRO PER PRO PER PRO
Low volume Mean
1.3
3.9
8.6
11.4
15.4
17.7
Variance
0.7
2.3
12.5
7.6
21.2
10.0
t-value(95%), difference (%)
-4.747, -200.00%
-1.976
-1.303
Medium volume Mean
5.3
5.2
17.2
19.5
29.7
42.9
Variance
7.6
7.7
13.7
10.9
32.0
83.9
t-value(95%), difference (%)
0.081
-1.464
-3.878, -44.44%
High volume Mean
19.3
84.5
56.3
309.1
122.9
848.4
Variance
34.7
174.1
125.8
109.9
202.8
957.4
t-value(95%), difference (%)
-14.271 , -337.82%
-52.075, -449.02%
-67.357 , -590.32%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table indicates the large increase in rear-end conflicts for high and medium volumes that is generated by adding the protected left-turn phase. This large increase in rear-end events is the primary cause of the total conflicts being counter-indicative. Thus, it should be important to analyze all types of conflicts rather than just examining the total number of events when comparing designs.

Table 12. Case 1 Comparison Results for Lane Change Conflicts.
LC
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  PER PRO PER PRO PER PRO
Low volume Mean
0.3
0.5
2.3
2.2
4
3.7
Variance
0.5
0.9
2.0
4.0
2.4
4.5
t-value(95%), difference (%)
-0.535
0.129
0.361
Medium volume Mean
3.2
7
7.8
12.3
13.9
19.6
Variance
1.5
7.6
8.0
24.7
9.2
40.7
t-value(95%), difference (%)
-3.991 , -118.75%
-2.491 , -57.69%
-2.551, -41.01%
High volume Mean
65.7
83.3
100.3
170.8
152.8
345.2
Variance
101.1
26.2
142.0
148.0
93.3
381.1
t-value(95%), difference (%)
-4.932,-26.79%
-13.092,-70.29%
-27.935,-125.92%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table indicates a 40-percent increase in lane change events at medium volumes and a 125-percent increase at high volumes. These results are likely due to longer queues in the left-turn bay because there is no permitted portion of the phase.

Table 13. Case 1 Comparison Results for Average Surrogate Measures of Safety.
  TPER TPRO CPER CPRO REPER REPRO LPER LPRO
TTC (low)
0.92
0.91
0.6
0.99
1.03
0.9
1.01
0.92
t- value, diff(%)
0.299
-2.547, 65.00% 3.849, 12.62%
1.112
TTC (med)
0.89
0.97
0.59
0.29
0.97
1.06
0.94
0.81
t -value, diff(%)
-3.167, -8.99%
1.776
-3.256, -9.28% 2.639, 13.83%
TTC (high)
0.8
1.05
0.46
0.74
1.01
1.11
0.71
0.92
t -value, diff(%)
-25.245, -31.25%
-5.922, 60.87%
-8.067, -9.90%
-13.285, -29.58%
PET(low)
2.21
1.83
0.92
1.75
2.63
1.86
2.64
1.71
t -value, diff(%)
3.796, 17.19%
-1.991
7.580, 29.28%
3.912, 35.23%
PET(med)
1.81
1.85
0.92
0.15
2.13
2.2
1.75
1.17
t -value, diff(%)
-0.532
8.840, 83.70%
-0.769
4.046, 33.14%
PET(high)
1.42
2.08
0.58
0.73
2.07
2.41
1.08
1.33
t -value, diff(%)
-26.886, -46.48%
-2.218, 25.86%
-8.598, -16.43%
-8.445, -23.15%
MaxS(low)
26.7
33.22
34.5
38.64
24.29
33.3
23.69
31.35
t -value, diff(%)
-7.947, -24.42%
-3.091, 12.00%
-9.521, -37.09%
-3.601, -32.33%
MaxS(med)
32.4
27.94
37.41
37.73
30.56
26.37
32.68
30.98
t -value, diff(%)
8.321, 13.77%
-0.159
5.916, 13.71%
1.831
MaxS(high)
28.27
25.11
32.2
23.42
26.32
24.85
28.97
25.82
t -value, diff(%)
20.782, 11.18%
14.825, 27.27%
6.757, 5.59%
13.492, 10.87%
DeltaS(low)
29.16
31.83
47.48
37.92
23.66
32.44
21.51
27.25
t -value, diff(%)
-2.524, -9.16% 5.886, 20.13%
-8.739, -37.11%
-2.838, -26.69%
DeltaS(med)
31.07
19.54
48.5
37.26
27.06
18.46
27
21.2
t -value, diff(%)
13.651, 37.11%
3.089, 23.18%
8.425, 31.78%
4.461, 21.48%
DeltaS(high)
22.22
16.12
34.98
21.3
20.44
16.97
20.83
13.79
t -value, diff(%)
27.963, 27.45%
16.672, 9.11%
11.610, 16.98%
23.063, 33.80%
DR(low)
-5.54
-6.91
-0.91
-7.22
-7.12
-6.96
-6.73
-6.57
t -value, diff(%)
4.632, -24.73%
6.826, 93.41%
-0.555
-0.245
DR(med)
-5.12
-3.72
-1.7
-1.43
-6.15
-4.21
-5.4
-2.73
t -value, diff(%)
-6.947, 27.34%
-0.235
-8.651, 31.54%
-6.712, 49.44%
DR(high)
-2.97
-4.77
-0.88
-0.56
-4.93
-5.61
-1.85
-2.88
t -value, diff(%)
23.945, -60.61%
-1.309
5.818, -13.79%
10.026, -55.68%
MaxD(low)
-11
-15.98
-2.09
-17.26
-14.03
-16.45
-13.37
-13.39
t -value, diff(%)
9.844, -45.27%
16.899, 25.84%
5.388, -17.25%
0.019
MaxD(med)
-13.04
-11.43
-4.28
-1.62
-15.25
-13.11
-14.7
-8.14
t -value, diff(%)
-4.499, 12.35%
-2.490, 62.15%
-7.662, 14.03%
-9.440 44.63%
MaxD(high)
-7.97
-12.06
-2.02
-3.44
-12.97
-13.96
-5.27
-7.76
t -value, diff(%)
31.265, -51.32%
2.918, -70.30%
7.012, -7.63%
13.251, -47.25%
MaxDeltaV(low)
16.91
18
28.86
20.85
13.21
18.35
12.36
15.56
t -value, diff(%)
-1.594
5.112, 27.75%
-8.851, -38.91%
-2.636, -25.89%
MaxDeltaV(med)
18.02
11.2
29.22
21.68
15.5
10.47
15.24
12.38
t -value, diff(%)
12.935, 37.85%
2.312, 25.80%
8.262, 32.45%
3.647, 18.77%
MaxDeltaV(high)
12.67
9
20.35
11.82
11.67
9.44
11.78
7.79
t -value, diff(%)
27.676, 28.97%
15.233, 41.92%
12.307, 19.11%
22.270, 33.87%

Note: Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

Correlations with Predicted Crash Frequency

The predicted crash rates (crashes per year) for all scenarios in this test are listed in table 14 with the corresponding surrogate measures of safety (conflicts per hour). Rank orders for each category of data are also listed in the table. The Spearman rank correlation coefficients are calculated for each test.

Table 14. Case 1 Spearman Rank Correlations Between Conflicts and Crash Frequency.
AADT Low Medium High Rs
PER
PRO
PER
PRO
PER
PRO
Crash Frequency
M
5
3.3
7.5
5
12.6
8.4
1
R
2
1
4
2
6
5
Total Conflict
M
25.7
22.4
53.7
63.3
309.5
1,208.6
0.77
R
1
1
3
4
5
6
Crossing Conflict
M
6.3
1
10.1
1.8
33.8
15
0.89
R
3
1
4
2
6
5
Rear-End Conflict
M
15.4
17.7
29.7
42.9
122.9
848.4
0.74
R
1
1
3
4
5
6
LC Conflict
M
4
3.7
13.9
19.6
152.8
345.2
0.74
R
1
1
3
4
5
5

Note: Rows labeled "M" provide mean values and rows labeled "R" provide the ranking of each alternative. The Rs column provides Spearman rank correlation coefficients indicating agreement with theoretical crash estimates.

Findings and Conclusions

Based on the observation of the safety surrogate test data, the following conclusions can be drawn:

  • Total number of conflicts for protected left turn is significantly more than that of permitted left turn.
  • Protected left turn has less crossing conflicts but more rear-end and lane-change conflicts than permitted left turn.
  • The comparison results for all other safety measures show no distinct safety preference between protected left turn and permitted left turn.

In general, the surrogate measures present mixed results; however, an appropriate conclusion can be drawn with consideration of the differing severities of different conflict types. The addition of a protected left-turn phase tended to increase the total number of conflicts while the protected phase substantially decreased crossing conflicts, as expected. The increase in conflicts came primarily from rear ends and, under higher flows, from lane-changing maneuvers. This result is elucidated by considering that, in adding a protected left-turn phase, the cycle time was increased and the proportion of green time for through phases was decreased. This change in the timing has the effect of increasing the number of vehicle stops for through traffic. The number of vehicle stops is known to correlate with the number of rear-end crashes and thus with higher rear-end conflicts. It is also possible that with a greater proportion of vehicles arriving to a standing or dispersing queue, there is an increased tendency of drivers to change to a lane with a shorter queue, despite all lanes having stopped traffic. Drivers in free-flowing lanes may be relatively content to stay in their lane when all lanes are flowing at the same (nonzero) speed. Thus, the conflict frequency results appear reasonable and have provoked consideration of the effect of timing changes and driver behavior. However, the severity-related surrogate measures indicate that the protected left-turn case has improved average values (increased TTC and PET and decreased DeltaV, especially at high volumes), indicating that the protected left-turn phasing is safer than the permitted left-turn case, as would be expected.

The Spearman-rank correlation coefficients from all tests show a strong positive relationship between the rank orders of the surrogate measures of safety and the rank orders of the predicted crash rates. The relationships between the rank order of the totals of all conflict types and crossing conflict types are stronger than the relationship of rear-end and lane-change crossing conflicts. This, again, would be expected because it has been validated in the field that protected left turns reduce crossing crashes. TEXAS, however, shows a very high rate of rear-end and lane-change events per hour, indicating that the default driver behavior parameters may allow vehicles to perform maneuvers that allow closer proximity than the "rule of thumb" threshold of TTC = 1.5 would preclude in the real world.

Case 2: Left-Turn Bay Versus No Left-Turn Bay (TEXAS)

A left-turn bay on an approach to an intersection provides an independent lane for the storage and movement of the left-turn vehicles. With the left-turn bay, the conflict events between through movement vehicles (primarily traveling in the same direction as the turn vehicles) and left-turn vehicles is hypothesized to be significantly reduced. This has been tested over a range of traffic volume scenarios from light traffic to heavy traffic.

According to the crash prediction models for all conventional intersections, the existence of a left-turn bay will reduce the crash frequency under the same traffic conditions.(13)

Intersection Description

The intersection used to test the left-turns bay versus no left-turn bay is a four-legged intersection with two through lanes with shared right turn for all approaches to the intersection, as shown in figure 36 and figure 37. All left-turn bays are 76.25 m (250 ft) long. Table 15 indicates the traffic volumes arriving to each approach of the intersection. Fixed-time traffic control is applied in this test. The ring-diagrams from figure 38 through figure 43 show the timing plans for each testing scenario.

Figure 36. Screen Capture. Exclusive Left-Turn Lane. This is a screen capture of an intersection model in TEXAS. The intersection is four legged with two through lanes and shared right turn for all approaches to the intersection. All approaches have left-turn bays that are 76.25 m (250 ft) long.

Figure 36. Screen Capture. Exclusive Left-Turn Lane.

Figure 37. Screen Capture. Shared Use Left-Turn and Through Lane. This is a screen capture of an intersection model in TEXAS. The intersection is four legged with two through lanes and shared right turn as well as shared left turn for all approaches to the intersection.

Figure 37. Screen Capture. Shared Use Left-Turn and Through Lane.

Table 15. Case 2 Service Flow by Each Approach.
Approach Southbound Northbound Eastbound Westbound
L TH R L TH R L TH R L TH R
Phase#
(Permitted)
4 4   8 8   2 2   6 6  
Low Volumes 125 250 125 125 250 125 125 250 125 125 250 125
Medium Volumes 200 400 200 200 400 200 200 400 200 200 400 200
High Volumes 300 600 300 300 600 300 300 600 300 300 600 300

Note: L, TH, and R correspond to vehicles proceeding left, through, or right at the intersection.

Figure 38. Illustration. Timing Plan for Intersection with Left-Turn Bay in Low Volumes. This is the screen capture of the timing plan for intersection with a left-turn bay case in low volumes condition. All through movements (phases 2, 4, 6, and 8) have 20 seconds as the split time.

Figure 38. Illustration. Timing Plan for Intersection with Left-Turn Bay in Low Volumes.

Figure 39. Illustration. Timing Plan for Intersection without Left-Turn Bay in Low Volumes. This is a screen capture of the timing plan for intersection without a left-turn bay case in low volumes condition. All through movements (phases 2, 4, 6, and 8) have 20 seconds as the split time.

Figure 39. Illustration. Timing Plan for Intersection without Left-Turn Bay in Low Volumes.

Figure 40. Illustration. Timing Plan for Intersection with Left-Turn Bay in Medium Volumes. This is a screen capture of the timing plan for intersection with a left-turn bay case in medium volumes condition. The length of green phase for phases 2, 4, 6, and 8 are all 20 seconds.

Figure 40. Illustration. Timing Plan for Intersection with Left-Turn Bay in Medium Volumes.

Figure 40. Illustration. Timing Plan for Intersection with Left-Turn Bay in Medium Volumes. This is a screen capture of the timing plan for intersection with a left-turn bay case in medium volumes condition. The length of green phase for phases 2, 4, 6, and 8 are all 20 seconds.

Figure 41. Illustration. Timing Plan for Intersection without Left-Turn Bay in Medium Volumes.

Figure 42. Illustration. Timing Plan for Intersection with Left-Turn Bay in High Volumes. This is a screen capture of the timing plan for an intersection with a left-turn bay case in high volumes condition. All through movements (phases 2, 4, 6, and 8) have 20 seconds as the split time.

Figure 42. Illustration. Timing Plan for Intersection with Left-Turn Bay in High Volumes.

Figure 43. Illustration. Timing Plan for Intersection without Left-Turn Bay in High Volumes. This is a screen capture of the timing plan for an intersection without a left-turn bay case in high volumes condition. The split time for the E-W movements (phase 2 and phase 6) is 23 seconds, and the split time for S-N movements (phase 4 and phase 8) is 22 seconds.

Figure 43. Illustration. Timing Plan for Intersection without Left-Turn Bay in High Volumes.

Data Analysis and Comparison Results

Ten replications were performed for each simulation scenario, and the resulting output trajectory data were analyzed by SSAM. F-test and t-tests were applied to identify statistical significance. Table 16 through table 20 list the values of all surrogate measures of safety and corresponding t 'test results for different types of aggregations with the low speed events and crash data excluded (TTC ≤ 0 and MaxS ≥ 16.1 km/h (10 mi/h)).

Table 16. Case 2 Comparison Results for Total Conflicts.
Total
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NLB WLB NLB WLB NLB WLB
Low volume Mean
9.6
11.3
9.6
26.3
54
42.6
Variance
7.2
26.9
7.2
45.8
68.7
63.8
tvalue(95%), difference (%)
-0.921
-7.258, 173.96% 3.132, 21.11%
Medium volume Mean
15.8
19.7
58.2
47.6
210.7
98.5
Variance
18.8
19.3
166.2
32.3
704.5
53.6
t-value(95%), difference (%)
-1.996
2.38, 18.21%
12.887, 53.25%
High volume Mean
140.8
150.1
506.8
279.1
985.9
487
Variance
156.4
78.5
250.4
81.4
467.9
293.6
t-value(95%), difference (%)
-1.919
39.528, 44.93%
57.174, 50.6%

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay. Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

Table 17. Case 2 Comparison Results for Crossing Conflicts.
Crossing
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  NLB WLB NLB WLB NLB WLB
Low volume Mean
5.3 8.5 5.3 12.3 11.9 15.3
Variance
2.7 12.5 2.7 15.3 6.3 17.3
t-value(95%), difference (%)
-2.597, -60.38% -5.214, -132.08% -2.210, -28.57%
Medium volume Mean
6.1 13.4 9 18.6 11 27.1
Variance
5.9 6.9 7.8 13.4 7.1 23.2
t-value(95%), difference (%)
-6.45, -119.67% -6.6, -106.67% -9.246, -146.36%
High volume Mean
9.1 34 15 45.6 19.4 60.8
Variance
11.7 52.7 10.7 62.9 15.6 69.5
t-value(95%), difference (%)
-9.818, -273.63% -11.279, -204.00% -14.191, -213.4%

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay. Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

This table indicates that, for all traffic volumes, the number of severe-crossing conflicts is increased when the left-turn bay is added. This result likely reflects the increase in the number of available left-turn maneuvers due to the bay.

Table 18. Case 2 Comparison Results for Rear-End Conflicts.
Rear End
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NLB WLB NLB WLB NLB WLB
Low volume Mean
2.7
2.3
2.7
11.6
31.8
21.3
Variance
3.1
3.8
3.1
18.3
40.2
42.7
t-value(95%), difference (%)
0.481
-6.085, -329.60% 3.648, 33%
Medium volume Mean
5.6
4
38.4
22.2
175.5
54.8
Variance
10.933
6.222
129.378
17.956
637.611
23.067
t-value(95%), difference (%)
1.222
4.221, 42.19%
14.85, 68.77%
High volume Mean
110.4
24.2
434.1
94.8
855.8
223
Variance
131.8
44.6
209.0
183.7
377.3
194.7
t-value(95%), difference (%)
20.521, 78.08%
54.143, 78.16%
83.673, 73.97%

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay. Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

This table indicates a definite decrease in the number of rear-end conflicts when a left- turn bay is added to the intersection, as expected from field experience.

Table 19. Case 2 Comparison Results for Lane-Change Conflicts.
Lane Change
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NLB WLB NLB WLB NLB WLB
Low volume Mean
1.6
0.5
1.6
2.4
10.3
6
Variance
1.4
0.3
1.4
3.6
6.0
6.4
t-value(95%), difference (%)
2.703, 68.80% -1.134 3.853, 41.75%
Medium volume Mean
4.1
2.3
10.8
6.8
24.2
16.6
Variance
1.878
3.567
4.844
8.622
31.289
27.156
t-value(95%), difference (%)
2.439, 43.9%
3.447, 37.04%
3.144, 31.4%
High volume Mean
21.3
91.9
57.7
138.7
110.7
203.2
Variance
21.6
93.7
53.3
177.3
113.1
242.6
t-value(95%), difference (%)
-20.799, -331.46%
-16.864, -140.38%
-15.509, -83.56%

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay. Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

Table 20. Case 2 Comparison Results for Average Surrogate Measures of Safety.
  TNLB TWLB CNLB CWLB RENLB REWLB LCNLB LCWLB
TTC (low)
0.97
0.86
0.67
0.56
1.08
1.01
1.01
1.09
t -value, diff(%)
3.928, 11.34%
1.804
2.428, 6.48%
1.339
TTC (med)
1.17
0.97
0.54
0.66
1.23
1.11
1.02
1.04
t -value, diff(%)
12.404, 17.09% -2.048, -22.22%
7.646, 9.76%
-0.506
TTC (high)
0.99
0.84
0.64
0.58
1
1.07
0.96
0.68
t -value, diff(%)
18.245, 15.15%
1.484
-7.507,-7.00% 16.155, 29.17%
PET(low)
2.22
2.07
1.44
1.29
2.56
2.5
2.07
2.51
t -value, diff(%)
1.732
0.880
0.670
-1.866
PET(med)
2.75
2.27
1.26
1.47
3.03
2.71
1.45
2.11
t -value, diff(%)
9.464, 17.45%
-1.364
5.996, 10.56% -5.236, -45.52%
PET(high)
1.77
1.64
1.16
0.95
1.82
2.38
1.49
1.02
t -value, diff(%)
5.860, 7.34%
2.252, 18.10%
-18.904, -30.77% 12.073, 31.54%
MaxS(low)
30.75
29.65
33.6
34.07
29.54
27.06
31.18
27.6
t -value, diff(%)
1.833
-0.518
2.933, 8.40%
2.353, 11.48%
MaxS(med)
27.01
29.95
34.33
34.77
26.16
27.11
29.88
31.45
t -value, diff(%)
-8.429, -10.88%
-0.499
-2.248, -3.63%
-1.623
MaxS(high)
24.38
28.78
28.44
31.78
24.12
27.38
25.64
29.42
t -value, diff(%)
-31.808, -18.05%
-5.551, -11.74%
-17.846, -13.52%
-12.115, -14.74%
DeltaS(low)
28.57
30.05
38.86
39.96
25.03
24.93
27.59
22.95
t -value, diff(%)
-1.788
-0.732
0.104
2.848, 16.82%
DeltaS(med)
19.23
28.28
38.02
41.46
18.05
23.46
19.21
22.69
t -value, diff(%)
-18.607, -47.06%
-2.558, -9.05%
-11.112, -29.97%
-3.351, -18.12%
DeltaS(high)
18.97
23.24
32.31
34.25
18.74
21.36
18.45
22.02
t -value, diff(%)
-23.467, -22.51%
-2.555, -6.00%
-11.461, -13.98%
-10.732, -19.35%
DR(low)
-6.3
-5.32
-2.48
-1.47
-7.48
-7.49
-7.08
-7.43
t -value, diff(%)
-3.323, 15.56%
-2.022, 40.73%
0.032
0.476
DR(med)
-6.1
-5.65
-0.49
-1.57
-6.6
-7.55
-5.04
-6.02
t -value, diff(%)
-2.539, 7.38%
3.233, -220.41%
4.818, -14.39%
2.373, -19.44%
DR(high)
-4.79
-3.83
-0.03
-1.2
-5.1
-6.19
-3.27
-2.02
t -value, diff(%)
-13.525, 20.04%
8.067, -3900.00%
11.276, -21.37%
-9.530, 38.23%
MaxD(low)
-12.86
-11.36
-5.63
-5.52
-15.11
-14.95
-14.26
-13.52
t -value, diff(%)
-3.487, 11.66%
-0.135
-0.419
-0.838
MaxD(med)
-13.64
-11.91
-2.09
-5.21
-14.62
-14.86
-11.78
-13.13
t -value, diff(%)
-7.813, 12.68% 5.299, -149.28%
1.371
2.427, -11.46%
MaxD(high)
-13.56
-9.09
-1.98
-2.72
-14.11
-14.15
-11.34
-5.45
t -value, diff(%)
-41.995, 32.96% 2.288, -37.37%
0.398
-30.409, 51.94%
MaxDeltaV(low)
16.35
17.23
22.74
23.05
14.27
14.21
15.39
13.08
t -value, diff(%)
-1.706
-0.309
0.103
2.448, 15.01%
MaxDeltaV(med)
10.87
16.14
23.86
24.42
10.06
13.09
10.84
12.72
t -value, diff(%)
-17.353, -48.48%
-0.586
-10.731, -30.12%
-3.140, -17.34%
MaxDeltaV(high)
10.62
13.23
19.46
19.61
10.44
12.05
10.44
12.61
t -value, diff(%)
-24.135, -24.58%
-0.286
-11.999, -15.42%
-10.832, -20.79%

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay, and these abbreviations are prepended with T-, C-, RE-, and LC- to indicate data based on total, crossing, rear-end, and lane-change conflicts respectively. Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

In general, the data in the table 20 have some counter-indicative results. Some of the average surrogate measures of safety are better with the left-turn bay, and others are worse.

Correlations with Predicted Crash Frequency

The predicted crash rates (crashes per year) for all scenarios in this test are listed in table 21 with the corresponding average conflicts per hour. Rank orders for each category of data are also listed in the table. The Spearman rank correlation coefficients are calculated for each test.

Table 21. Case 2 Spearman Rank Correlations Between Conflicts and Crash Frequency.
AADT
Low
Medium
High
Rs
NLB WLB NLB WLB NLB WLB
Crash Frequency M
5.5
3.7
8
5.3
8.7
5.8
1
R
3
1
5
2
6
4
Total Conflict M
54
42.6
210.7
98.5
985.9
487
0.8
R
1
1
4
3
6
5
Crossing Conflict M
11.9
15.3
11
27.1
19.4
60.8
0
R
2
3
1
5
4
6
Rear-End Conflict M
31.8
21.3
175.5
54.8
855.8
223
0.8
R
1
1
4
3
6
5
LC Conflict M
10.3
6
24.2
16.6
110.7
203.2
0.6
R
2
1
3
4
5
6

Note: NLB indicates no left-turn bay and WLB indicates with left-turn bay. Rows labeled "M" provide mean values and rows labeled "R" provide the ranking of each alternative. The Rs column provides Spearman rank correlation coefficients indicating agreement with theoretical crash estimates.

Findings and Conclusions

Based on the observation on the total number of conflicts of various types and the average values of the surrogate measures obtained from the test, the following conclusions can be drawn:

  • The average number of total and rear-end conflicts for intersection with left-turn bay is less than that of intersection without left-turn bay. This is an intuitive result.
  • An intersection with a left-turn bay has more crossing and lane change (in high volumes) conflicts than an intersection without a left-turn bay. With an extra lane to change to, and increase in lane-change conflicts seems reasonable. Also, with exclusive left-turn bays on both sides of the street, drivers making left turns then face opposing vehicles traveling through the intersection at higher speeds since they are no longer periodically blocked by left-turning vehicles. It seems possible that the increased speed of opposing traffic increases the likelihood of conflicts. It is also possible that left-turning "blockers" in opposing traffic also created episodes of relatively safe crossing opportunities because there is one less oncoming lane of traffic to contend with.
  • Some average values of surrogate measures of safety indicate that adding the bay increases the severity of the conflicts that do occur, primarily by increasing speeds at the intersection. This fact makes it difficult to definitively determine the superiority of one design over the other.

In general, an intersection with a left-turn bay experiences fewer total and rear-end conflicts but more crossing and lane-change conflicts than an intersection without a left- turn bay. Rear-end conflicts constitute a major part of total conflicts (ranging from 60 to 80 percent) and have larger TTC and PET values (≥ 1.0).

The Spearman rank correlation coefficients resulted from all tests show a strong positive relationship between the rank orders of the surrogate measures of safety and the rank orders of the predicted crash rates, except for crossing conflicts, which shows no correlation with total crash rates. Perhaps a more normative comparison could be made by using rates of conflict occurrence by maneuver rather than total number of conflicts without relation to the number of other maneuvers that were executed by drivers without a conflict occurring.

TEXAS shows a very high rate of rear-end and lane-change events per hour, indicating that the default driver behavior parameters may allow vehicles to perform maneuvers that allow closer proximity than the "rule of thumb" threshold of TTC = 1.5 would preclude in the real world. Also, the existence of the turn bay requires more lane-changing maneuvers and thus a higher frequency of conflict events related to those necessary lane changes.

Case 3: Right-Turn Bay Versus No Right-Turn Bay (TEXAS)

A right-turn bay near the intersection provides an independent lane for the storage and movement of right-turn vehicles. With a right-turn bay near an intersection, the conflict events between through-movement vehicles (primarily traveling in the same direction as the turn vehicles) and right-turn vehicles is hypothesized to be reduced significantly. This reduction has been tested over a range of traffic volume scenarios, from light traffic to heavy traffic.

According to the crash prediction models for all conventional intersections, the existence of a right-turn bay will definitely reduce the crash frequency when all other roadway network factors remain the same.(13)

Intersection Description

The intersection used to test the right-turn bay versus no right-turn bay is a four-legged intersection with two through lanes and one left-turn lane for all approaches to the intersection, as shown in figure 44 and figure 45. All left-turn bays have are 76.25-meters (250-feet) long. Table 22 shows the traffic volumes applied for each approach of the intersection. Fixed time traffic control is applied in this test. Figure 46 through figure 51 provide the timing plans for each testing scenario.

Figure 44. Screen Capture. Intersection with Right-Turn Bay. This is a screen capture of an intersection model in TEXAS. The intersection is four legged with two through lanes and one left-turn lane for all approaches to the intersection. Each approach has an exclusive right-turn bay. All left turn-bays are 76.25 m (250 ft) long.

Figure 44. Screen Capture. Intersection with Right-Turn Bay.

Figure 45. Screen Capture. Intersection without Right-Turn Bay. This is a screen capture of an intersection model in TEXAS. The intersection is four legged with two through lanes and one left-turn lane for all approaches to the intersection. Right turns in all approaches share a through lane. All left-turn bays are 76.25 m (250 ft) long.

Figure 45. Screen Capture. Intersection without Right-Turn Bay.

Table 22. Case 3 Service Flow by Each Approach.
Approach
Southbound
Northbound
Eastbound
Westbound
L
TH
R
L
TH
R
L
TH
R
L
TH
R
Phase ID
7
4
 
3
8
 
5
2
 
1
6
 
Low Volumes
125
250
125
125
250
125
125
250
125
125
250
125
Medium Volumes
200
400
200
200
400
200
200
400
200
200
400
200
High Volumes
300
600
300
300
600
300
300
600
300
300
600
300

Note: L, TH, and R correspond to vehicles proceeding left, through, or right at the intersection.

Figure 46. Illustration. Timing Plan for Intersection with Right-Turn Bay in Low Volumes. This is a screen capture of the timing plan for an intersection with a right-turn bay case in low volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 21 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 12 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 12 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 46. Illustration. Timing Plan for Intersection with Right-Turn Bay in Low Volumes.

Figure 47. Illustration. Timing Plan for Intersection without Right-Turn Bay in Low Volumes. This is a screen capture of the timing plan for an intersection without a right-turn bay case in low  volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 21 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 12 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 12 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 47. Illustration. Timing Plan for Intersection without Right-Turn Bay in Low Volumes.

Figure 48. Illustration. Timing Plan for Intersection with Right-Turn Bay in Medium Volumes. This is a screen capture of the timing plan for an intersection with a right-turn bay case in medium volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 20 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 15 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 15 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 48. Illustration. Timing Plan for Intersection with Right-Turn Bay in Medium Volumes.

Figure 49. Illustration. Timing Plan for Intersection without Right-Turn Bay in Medium Volumes. This is a screen capture of the timing plan for an intersection without a right-turn bay case in medium volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 20 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 15 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 15 seconds. The split time for S-N through movements (phase 4 and phase 8) is 20 seconds.

Figure 49. Illustration. Timing Plan for Intersection without Right-Turn Bay in Medium Volumes.

Figure 50. Illustration. Timing Plan for Intersection with Right-Turn Bay in High Volumes. This is a screen capture of the timing plan for an intersection with a right-turn bay case in high volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 26 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 25 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 25 seconds. The split time for S-N through movements (phase 4 and phase 8) is 24 seconds.

Figure 50. Illustration. Timing Plan for Intersection with Right-Turn Bay in High Volumes.

Figure 51. Illustration. Timing Plan for Intersection without Right-Turn Bay in High Volumes. This is a screen capture of the timing plan for an intersection without a right-turn bay case in high volumes condition. The split time for E-W through movements (phase 2 and phase 6) is 40 seconds. The split time for E-W left-turn movements (phase 1 and phase 5) is 26 seconds. The split time for the S-N left-turn movements (phase 3 and phase 7) is 26 seconds. The split time for S-N through movements (phase 4 and phase 8) is 38 seconds.

Figure 51. Illustration. Timing Plan for Intersection without Right-Turn Bay in High Volumes.

Data Analysis and Comparison Results

Ten replications were performed for each simulation scenario, and the resulting output trajectory data was analyzed by SSAM. The F-test and t-test were applied to compare surrogate measures of safety and the aggregations of those measures. Table 23 through table 27 provide the values of all surrogate measures of safety and corresponding t-test results for different types of aggregations with the low speed events and crash data excluded (TTC ≤ 0 and MaxS ≥ 16.1 km/h (10 mi/h)).

Table 23. Case 3 Comparison Results for Total Conflicts.
Total
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NRB WRB NRB WRB NRB WRB
Low volume Mean
5.8
3.5
16.4
14.7
35.9
24.2
Variance
7.067
3.611
17.822
9.122
38.544
18.844
t-value(95%), difference (%)
2.226, 39.66%
1.036
4.884, 32.59%
Medium volume Mean
12
10.1
43.4
31.2
119
67.3
Variance
16.444
14.322
132.711
66.400
361.556
256.900
t-value(95%), difference (%)
1.083
2.734, 28.11%
6.574, 43.45%
High volume Mean
163
113.7
532.3
372.1
1259.5
883.8
Variance
207.333
86.233
1108.456
485.433
623.167
1494.622
t-value(95%), difference (%)
9.099, 30.25%
12.689, 30.1%
25.817, 29.83%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table indicates that the right-turn bay reduces the total number of conflict events for most levels of traffic volume and threshold for the TTC value.

Table 24. Case 3 Comparison Results for Crossing Conflicts.
Crossing
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NRB WRB NRB WRB NRB WRB
Medium volume Mean
0.5
0.6
0.7
0.8
1.3
0.8
Variance
0.944
0.933
1.122
1.511
1.567
1.511
t-value(95%), difference (%)
-0.231
-0.195
0.901
High volume Mean
6.5
7.3
9
10.8
14.3
15.3
Variance
6.278
3.122
7.778
7.067
11.789
11.789
t-value(95%), difference (%)
-0.825
-1.477
-0.651

As expected this table indicates that a right-turn bay would not reduce the number of crossing-conflict events.

Table 25. Case 3 Comparison Results for Rear-End Conflicts.
Rear End TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  NRB WRB NRB WRB NRB WRB
Low volume Mean
3.3
2.1
11.6
7.7
25.8
12.6
Variance
3.344
1.656
14.044
9.789
23.733
13.156
t-value(95%), difference (%)
1.697
2.526, 33.62%
6.873, 51.16%
Medium volume Mean
5
3.3
29.6
18.7
92.1
44.4
Variance
5.556
4.678
76.044
22.011
252.544
99.600
t-value(95%), difference (%)
1.681
3.481, 36.82%
8.038, 51.79%
High volume Mean
97.6
60
409.9
262.9
1024.9
675.6
Variance
108.044
81.111
922.322
360.322
468.767
1380.711
t-value(95%), difference (%)
8.645, 38.52%
12.98, 35.86%
25.685, 34.08%

Note: Shaded cells indicate statistically significant differences between the two alternatives.

This table indicates that adding a right-turn bay will statistically reduce the number of rear-end conflicts for all traffic volumes, as expected from field experience.

Table 26. Case 3 Comparison Results for Lane-Change Conflicts.
Lane Change
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
NRB WRB NRB WRB NRB WRB
Low volume Mean
2.5
1.4
4.7
7
9.6
11.4
Variance
2.056
1.600
2.233
5.556
5.156
13.156
t-value(95%), difference (%)
1.819
-2.606, -48.94%
-1.330
Medium volume Mean
6.5
6.2
13.1
11.7
25.6
22.1
Variance
6.722
8.622
15.656
28.456
30.044
66.544
t-value(95%), difference (%)
0.242
0.667
1.126
High volume Mean
58.9
46.4
113.4
98.4
220.3
192.9
Variance
82.322
4.711
170.044
20.267
235.122
169.656
t-value(95%), difference (%)
4.237, 21.22% 3.438, 13.23% 4.307, 12.44%

Note: Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

This table indicates that adding a right-turn bay will statistically reduce the number of lane change conflicts for some traffic volumes. Notably, at high volumes, the number of high-severity (indicated by considering only TTC values less than 0.5s and 1.0s) lane-change events is reduced, as expected from field experience.

. Case 3 Comparison Results for Average Surrogate Measures of Safety.
  TNRB TWRB CNRB CWRB RENRB REWRB LCNRB LCWRB
TTC (low)
1.03
0.97
N/A
N/A
1.05
0.96
0.94
0.97
t -value, diff(%)
1.887
N/A
2.213, 8.57%
-0.504
TTC (med)
1.11
1.02
0.82
0.49
1.17
1.09
0.92
0.9
t -value, diff(%)
4.625, 8.11%
1.378
4.115, 6.84%
0.445
TTC (high)
1.05
1.05
0.71
0.65
1.08
1.09
0.9
0.92
t -value, diff(%)
0.000
1.012
-1.842
-1.339
PET(low)
2.16
1.88
N/A
N/A
2.28
2.09
1.78
1.62
t -value, diff(%)
2.569, 12.96%
N/A
1.362
0.884
PET(med)
2.49
2.22
1.68
0.26
2.77
2.53
1.52
1.69
t -value, diff(%)
4.315, 10.84% 2.687, 84.52% 3.545, 8.66%
-1.462
PET(high)
2.12
2.14
0.85
0.75
2.3
2.4
1.35
1.32
t -value, diff(%)
-1.165
1.042
-5.378, -4.35%
0.918
MaxS(low)
30.74
32.3
N/A
N/A
31.1
31.34
29.99
33.61
t -value, diff(%)
-1.925
N/A
-0.212
-2.676, -12.07%
MaxS(med)
27.07
26.33
25.48
27.73
25.83
24.3
31.59
30.37
t -value, diff(%)
1.616
-0.553
2.946, 5.92%
1.380
MaxS(high)
24.67
25.15
21.32
23.43
24.75
25.28
24.55
24.84
t -value, diff(%)
-4.347, -1.95% -3.162, -9.90% -4.218, -2.14%
-1.181
DeltaS(low)
26.28
29.12
N/A
N/A
27.84
29.01
22.13
29.45
t -value, diff(%)
-3.004, -10.81%
N/A
-0.907
-4.854, -33.08%
DeltaS(med)
17.9
17.84
23.98
26.08
17.24
16.46
19.96
20.31
t -value, diff(%)
0.101
-0.515
1.054
-0.348
DeltaS(high)
18.29
18.58
19.35
21.89
18.95
19.44
15.14
15.29
t -value, diff(%)
-2.165, -1.59% -3.382, -13.13% -3.253, -2.59%
-0.515
DR(low)
-6.96
-6.48
N/A
N/A
-7.29
-6.59
-5.71
-6.22
t -value, diff(%)
-1.662
N/A
-1.925
1.010
DR(med)
-5.66
-4.69
-7.27
-3.97
-5.9
-4.99
-4.7
-4.12
t -value, diff(%)
-4.732, 17.14%
-1.052
-3.670, 15.42%
-1.489
DR(high)
-5.17
-5.16
-0.83
-0.93
-5.75
-5.96
-2.78
-2.7
t -value, diff(%)
-0.176
0.320
3.440, -3.65%
-0.667
MaxD(low)
-14.54
-15.33
N/A
N/A
-15.54
-16.13
-11.66
-14.39
t -value, diff(%)
1.980, 5.43%
N/A
1.164
3.866, -23.41%
MaxD(med)
-13.19
-12.28
-9.69
-4.48
-13.95
-13.41
-10.62
-10.29
t -value, diff(%)
-3.977, 6.90%
-1.558
-2.852, 3.87%
-0.558
MaxD(high)
-13.03
-12.72
-3.82
-3.14
-14.23
-14.33
-8.09
-7.86
t -value, diff(%)
-4.475, 2.38%
-1.158
1.924
-1.150
MaxDeltaV(low)
14.83
16.75
N/A
N/A
15.8
16.7
12.21
16.92
t -value, diff(%)
-3.278, -12.95%
N/A
-1.128
-5.199, -38.57%
MaxDeltaV(med)
10.11
10.07
13.59
14.13
9.72
9.28
11.32
11.52
t -value, diff(%)
0.115
-0.227
1.024
-0.342
MaxDeltaV(high)
10.19
10.37
10.89
12.52
10.54
10.82
8.54
8.62
t -value, diff(%)
-2.357, -1.77%
-3.428, -14.97%
-3.263, -2.66%
-0.478

Note: Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

Similar to the average value results for the left-turn bay, the existence of the right-turn bay tends to make the severity of conflicts worse for measures other than the primary measure TTC.

Correlations with Predicted Crash Frequency

The predicted crash rates for all scenarios in this test are listed in table 28 with the corresponding surrogate measures of safety. Rank orders for each category of data are also listed in the table. The Spearman rank correlation coefficients are calculated for each test.

Table 28. Case 3 Spearman Rank Correlations Between Conflicts and Crash Frequency.
AADT
Low
Medium
High
Rs
NRB WRB NRB WRB NRB WRB
Crash Frequency
M
5.5
5.2
8
7.5
8.7
8.3
1
R
2
1
4
3
6
5
Total Conflict
M
35.9
24.2
119
67.3
1,259.5
883.8
0.97
R
2
1
4
3
5
5
Crossing Conflict
M
0.5
0.2
1.3
0.8
14.3
15.3
0.91
R
1
1
3
3
5
5
Rear-End Conflict
M
25.8
12.6
92.1
44.4
1,024.9
675.6
0.97
R
2
1
4
3
5
5
LC Conflict
M
9.6
11.4
25.6
22.1
220.3
192.9
0.80
R
1
1
3
3
5
5

Note: Rows labeled "M" provide mean values and rows labeled "R" provide the ranking of each alternative. The Rs column provides Spearman rank correlation coefficients indicating agreement with theoretical crash estimates.

As seen in the table above, the correlation between the number of conflicts and the crash prediction model is very high.

Findings and Conclusions

Based on the observation on the safety surrogate data obtained from the test, the following conclusions can be drawn:

  • The number of total, rear-end, and lane-change conflicts for an intersection with a right-turn bay are less than that of an intersection without a right-turn bay. This is an intuitive result.
  • There is no significant difference for the number of crossing conflicts between an intersection with a right-turn bay and an intersection without a right-turn bay. This is an intuitive result.
  • There is no distinct difference for the average values of the surrogate measures of safety between an intersection with a right-turn bay and an intersection without a right-turn bay.

In general, an intersection with a right-turn bay experiences less total, rear-end, and lane-change conflicts than an intersection without a right-turn bay. Rear-end conflicts constitute the major part of the total conflicts (ranging from 50 to 80 percent) and have larger TTC and PET values (≥ 1.0) than other types of conflicts. There is no significant difference for the number of crossing conflicts between these two intersection designs. These results match what would generally be expected from field experience.

The Spearman-rank correlation coefficients resulted from all tests show a strong positive relationship between the rank orders of the surrogate measures of safety and the rank orders of the predicted crash rates, with only slight differences between the correlation coefficients for the various individual conflict types.

Case 4: Leading Left Turn Versus Lagging Left Turn (VISSIM)

Leading left turn and lagging left turn are two different control logics for protected left turns. A leading left turn allows the green phase of left-turn movements ahead of the green phase of through movements, while a lagging left turn places the green phase of the left turn after the green phase of through movements. At the beginning of the study, there is no determinate evidence that either type of left-turn operation has an appreciable effect on the safety of the intersection. Several combinations of the left-turn to through-volume ratio have been evaluated for each design alternative (leading versus lagging).

Until now, no considerations of control logic for protected left turn have been included in crash prediction models for conventional intersections, so no rank order comparison can be performed.

Intersection Description

The intersection used to test leading left turn versus lagging left turn is a four-legged intersection with one through lane having left- and right-turn bays in the main travel directions and one though lane on the side street, as shown in figure 52 and figure 53. All left- turn bays are 76.25 m (250 ft) long.

Figure 52. Screen Capture. Intersection with Leading Left Turn. This is a screen capture of an intersection model in VISSIM. The intersection is a four-legged intersection with one through lane having left- and right-turn bays in the main travel directions and one though lane on the side-street. All left-turn bays are 76.25 m (250 ft) long.

Figure 52. Screen Capture. Intersection with Leading Left Turn.

Figure 53. Screen Capture. Intersection with Lagging Left Turn. This is a screen capture of an intersection model in VISSIM. The intersection is a four-legged intersection with one through lane having left- and right-turn bays in the main travel directions and one though lane on the side-street. All left-turn bays are 76.25 m (250 ft) long.

Figure 53. Screen Capture. Intersection with Lagging Left Turn.

Table 29 shows the traffic volumes applied for each approach of the intersection. Fixed time traffic control is applied in this test. Figure 54 through figure 59 provide the timing plans used for each testing scenario (low to high volumes).

Table 29. Case 4 Service Flow by Each Approach.
Approach
Southbound Northbound Eastbound Westbound
L TH R L TH R L TH R L TH R
Phase ID
3
3
 
4
4
 
1
2
 
1
2
 
Low Volumes
25
75
25
50
100
50
150
400
50
150
400
50
Medium Volumes
25
75
25
50
100
50
125
650
75
125
650
75
High Volumes
100
250
50
100
250
50
150
700
150
150
700
150

Note: L, TH, and R correspond to vehicles proceeding left, through, or right at the intersection.

Figure 54. Illustration. Timing Plan for Intersection with Lag Left Turn in Low Volumes (Cycle: 80; Split: 22, 17, 22, and 19). This is a screen capture of the timing plan for an intersection with a lag left turn in low volumes condition. The cycle length is 80 seconds, and the split for phases 1 through 4 are 22, 17, 22, and 19, respectively.

Figure 54. Illustration. Timing Plan for Intersection with Lag Left Turn in Low Volumes (Cycle: 80; Split: 22, 17, 22, and 19).

Figure 55. Illustration. Timing Plan for Intersection with Lead Left Turn in Low Volumes (Cycle: 80; Split: 17, 22, 22, and 19). This is a screen capture of the timing plan for an intersection with a lead left turn in low volumes condition. The cycle length is 80 seconds, and the split for phases 1 through 4 are 17, 22, 22, and 19, respectively.

Figure 55. Illustration. Timing Plan for Intersection with Lead Left Turn in Low Volumes (Cycle: 80; Split: 17, 22, 22, and 19).

Figure 56. Illustration. Timing Plan for Intersection with Lag Left Turn in Medium Volumes (Cycle: 80; Split: 33, 17, 15, and 15). This is a screen capture of the timing plan for the intersection with a lag left turn in medium volumes condition. The cycle length is 80 seconds, and the split for phases 1 through 4 are 33, 17, 15, and 15, respectively.

Figure 56. Illustration. Timing Plan for Intersection with Lag Left Turn in Medium Volumes (Cycle: 80; Split: 33, 17, 15, and 15).

Figure 57. Illustration. Timing Plan for Intersection with Lead Left Turn in Medium Volumes (Cycle: 80; Split: 17, 33, 15, and 15). This is a screen capture of the timing plan for an intersection with a lead left turn in medium volumes condition. The cycle length is 80 seconds, and the split for phases 1 through 4 are 17, 33, 15, and 15, respectively.

57. Illustration. Timing Plan for Intersection with Lead Left Turn in Medium Volumes (Cycle: 80; Split: 17, 33, 15, and 15).

Figure 58. Illustration. Timing Plan for Intersection with Lag Left Turn in High Volumes (Cycle: 75; Split: 20, 11, 23, and 21). This is s screen capture of the timing plan for an intersection with a lag left turn in high volumes condition. The cycle length is 75 seconds, and the split for phases 1 through 4 are 20, 11, 23, and 21, respectively.

Figure 58. Illustration. Timing Plan for Intersection with Lag Left Turn in High Volumes (Cycle: 75; Split: 20, 11, 23, and 21).

Figure 59. Illustration. Timing Plan for Intersection with Lead Left Turn in High Volumes (Cycle: 75; Split: 11, 20, 23, and 21). This is a screen capture of the timing plan for an intersection with a lead left turn in high volumes condition. The cycle length is 75 second, and the split for phases 1 through 4 are 11, 20, 23, and 21, respectively.

Figure 59. Illustration. Timing Plan for Intersection with Lead Left Turn in High Volumes (Cycle: 75; Split: 11, 20, 23, and 21).

Data Analysis and Comparison Results

Ten replications were performed for each simulation scenario, and the resulting output trajectory data was analyzed by SSAM. The F-test and t-test were applied to compare the average number of conflict events and surrogate measures of safety from one scenario to the other.

Table 30 through table 33 provide the values of all surrogate measures of safety and corresponding t-test results for different types of aggregations with the low-speed events and crash data excluded (TTC ≤ 0 and MaxS ≥ 16.1 km/h (10 mi/h)).

Table 30. Case 4 Comparison Results for All Conflict Event Types.
Total
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
Lead Lag Lead Lag Lead Lag
Low volume Mean
N/A
N/A
N/A
N/A
9.90
8.80
Variance
N/A
N/A
N/A
N/A
5.43
14.84
t-value(95%), difference (%)
N/A
N/A
0.772
Medium volume Mean
N/A
N/A
N/A
N/A
16.30
13.40
Variance
N/A
N/A
N/A
N/A
12.23
23.60
t-value(95%), difference (%)
N/A
N/A
1.532
High volume Mean
N/A
N/A
8.10
8.10
35.60
32.30
Variance
N/A
N/A
5.88
11.88
78.27
33.34
t-value(95%), difference (%)
N/A
0
0.988
Table 31. Case 4 Comparison Results for Rear-End Conflicts.
Rear End TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
  Lead Lag Lead Lag Lead Lag
Low volume Mean
N/A
N/A
N/A
N/A
5.70
5.70
Variance
N/A
N/A
N/A
N/A
3.57
8.90
t-value(95%), difference (%)
N/A
N/A
0
Medium volume Mean
N/A
N/A
N/A
N/A
10.00
8.60
Variance
N/A
N/A
N/A
N/A
7.78
13.60
t-value(95%), difference (%)
N/A
N/A
0.958
High volume Mean
N/A
N/A
5.80
5.80
24.30
22.10
Variance
N/A
N/A
3.96
6.18
42.01
21.21
t-value(95%), difference (%)
N/A
0
0.875
Table 32. Case 4 Comparison Results for Lane-Change Conflicts.
Lane Change
TTC ≤ 0.5 TTC ≤ 1.0 TTC ≤ 1.5
 
Lead Lag Lead Lag Lead Lag
Low volume Mean
N/A
N/A
N/A
N/A
4.10
3.10
Variance
N/A
N/A
N/A
N/A
5.21
2.54
t-value(95%), difference (%)
N/A
N/A
1.136
Medium volume Mean
N/A
N/A
N/A
N/A
6.20
4.50
Variance
N/A
N/A
N/A
N/A
6.40
3.17
t-value(95%), difference (%)
N/A
N/A
1.738
High volume Mean
N/A
N/A
2.20
2.30
11.20
9.90
Variance
N/A
N/A
2.84
3.12
15.29
6.99
t-value(95%), difference (%)
N/A
-0.129
0.871
Table 33. Case 4 Comparison Results for Average Surrogate Measures of Safety.
  TLead TLag CLead CLag RELead RELag LCLead LCLag
TTC (low)
1.28
1.28
N/A
N/A
1.32
1.3
1.23
1.25
t -value, diff(%)
0.000
N/A
0.571
-0.305
TTC (med)
1.27
1.29
N/A
N/A
1.31
1.28
1.2
1.28
t -value, diff(%)
-0.868
N/A
1.281
-1.533
TTC (high)
1.2
1.2
N/A
N/A
1.2
1.19
1.2
1.24
t -value, diff(%)
0.000
N/A
0.46
-1.240
PET(low)
2.1
2.02
N/A
N/A
2.27
2.13
1.89
1.83
t -value, diff(%)
0.695
N/A
1.102
0.272
PET(med)
2.3
2.35
N/A
N/A
2.42
2.52
2.1
1.99
t -value, diff(%)
-0.497
N/A
-0.860
0.607
PET(high)
2.47
2.47
N/A
N/A
2.48
2.58
2.48
2.23
t -value, diff(%)
0.000
N/A
-1.467
2.154, 10.08%
MaxS(low)
6.69
6.78
N/A
N/A
6.37
6.72
7.16
6.9
t -value, diff(%)
-0.242
N/A
-0.763
0.405
MaxS(med)
7.52
7.19
N/A
N/A
6.94
7.53
8.47
6.6
t -value, diff(%)
0.964
N/A
-1.331
3.834, 22.1%
MaxS(high)
7.76
7.94
N/A
N/A
7.78
7.8
7.78
8.24
t -value, diff(%)
-0.734
N/A
-0.064
-1.036
DeltaS(low)
5.35
5.25
N/A
N/A
5.28
5.25
5.45
5.26
t -value, diff(%)
0.576
N/A
0.195
0.481
DeltaS(med)
5.37
5.35
N/A
N/A
5.24
5.21
5.56
5.59
t -value, diff(%)
0.135
N/A
0.184
-0.1015
DeltaS(high)
4.47
4.61
N/A
N/A
4.22
4.3
4.22
5.23
t -value, diff(%)
-0.986
N/A
-0.462
-4.079, -23.93%
DR(low)
-2.48
-2.24
N/A
N/A
-2.62
-2.23
-2.28
-2.25
t -value, diff(%)
-1.781
N/A
-2.357, 14.89%
-0.129
DR(med)
-2.28
-2.44
N/A
N/A
-2.31
-2.34
-2.22
-2.51
t -value, diff(%)
1.095
N/A
0.230
0.871
DR(high)
-1.9
-1.97
N/A
N/A
-1.72
-1.85
-1.72
-2.21
t -value, diff(%)
0.624
N/A
1.057
2.371, -28.49%
MaxD(low)
-3.42
-3.05
N/A
N/A
-3.2
-2.9
-3.65
-3.32
t -value, diff(%)
-1.573
N/A
-1.177
-0.743
MaxD(med)
-3.54
-3.79
N/A
N/A
-3.26
-3.83
-3.94
-3.61
t -value, diff(%)
1.117
N/A
2.379, -17.48%
-0.737
MaxD(high)
-4.13
-4.02
N/A
N/A
-4.02
-4.02
-4.02
-4
t -value, diff(%)
-0.805
N/A
0.000
-0.077
MaxDeltaV(low)
2.78
2.68
N/A
N/A
2.78
2.62
2.79
2.77
t -value, diff(%)
1.027
N/A
1.801
0.094
MaxDeltaV(med)
2.74
2.7
N/A
N/A
2.66
2.6
2.87
2.86
t -value, diff(%)
0.452
N/A
0.675
0.051
MaxDeltaV(high)
2.28
2.36
N/A
N/A
2.15
2.2
2.15
2.68
t -value, diff(%)
-1.089
N/A
-0.559
-4.107, -24.65%

Note: Shaded cells indicate statistically significant differences between the two alternatives. The tan and blue colors indicate extreme values to the right and left columns respectively.

These tables indicate no significant differences between either the number or severity of conflict events for leading and lagging protected left turns, as expected from field experience.

Correlations with Predicted Crash Frequency

Since no consideration for leading or lagging left turns has been incorporated into a crash prediction model to date, any comparisons do not have meaning. Results for crossing conflicts have been excluded from the table below because there were not enough events to analyze (left turns were protected only).

Table 34. Case 4 Spearman Rank Correlations Between Conflicts and Crash Frequency.
AADT
Low Medium High Rs
Lead Lag Lead Lag Lead Lag
Crash Frequency
M
3.9
3.9
4.8
4.8
6.3
6.3
1
R
1
1
3
3
5
5
Total Conflict
M
9.90
8.80
16.3
13.4
35.60
32.30
1
R
1
1
3
3
5
5
Crossing Conflict
M
 
 
 
 
 
 
N/A
R
 
 
 
 
 
 
Rear-End Conflict
M
5.7
5.7
10
8.6
24.30
22.10
1
R
1
1
3
3
5
5
LC Conflict
M
4.1
3.1
6.20
4.50
11.20
9.90
1
R
1
1
3
3
5
5

Note: Rows labeled "M" provide mean values and rows labeled "R" provide the ranking of each alternative. The Rs column provides Spearman rank correlation coefficients indicating agreement with theoretical crash estimates.

Findings and Conclusions

Based on the observation on the safety surrogate data obtained from the test, the following conclusions can be drawn:

  • There is no significant difference for any of the surrogate measures of safety between leading left turns and lagging left turns. Note that in both cases the left turns were protected only.

In general, there is no significant difference for any of the surrogate measures of safety between leading protected left turns and lagging protected left turns. This result matches the intuitive expectation. There may, however, be some difference between the two if a permitted-protected operation was considered instead of a protected-only operation.

Previous | Table of Contents | Next

ResearchFHWA
FHWA
United States Department of Transportation - Federal Highway Administration