Efficiency Assessment of Extended Change and Clearance Intervals on Signalized Intersections and Corridors

To investigate the impact of lengthening the change and clearance intervals timing on intersections and arterial efficiency, three major parameters were measured for the three signal timing patterns. The performance measures are determined. Finally, the equations of dilemma and option zones were utilized to study the impact of the three signal timing patterns of extending change and clearance phases on intersections' safety.

3.1 Signal timing plans

The delay, 95th percentile queuing length of the critical left turn movement, and V/C ratio were measured to investigate the extended change and clearance intervals on the signalized intersections and corridors.

3.1.1 Signal timing pattern 1

It was implemented for SR 50, SR 535, and SR 536 intersections before FDOT carried the signal system retiming project. Generally, the cycle's length of SR 50 corridor was ranged between 150 seconds and 170 seconds for morning peak hour and ranged between 180 seconds and 190 seconds for the evening peak hour.

3.1.2 Signal timing pattern 2

It is the current signal timing system implemented recently on central Florida signalized intersections. It was recently approved by the FDOT and performed on a number of corridors in Florida. The signal retiming system extended moderately the signal cycles by lengthening the change and clearance intervals. For pattern 2, the cycle lengths signalized intersections show longer cycles length with usually between 5 to 10 seconds more than in pattern 1 that led to an increase in delays of traffic passing through the corridors intersections.

3.1.3 Signal timing pattern 3

It is a proposed signal timing designed with a perception-response time (PRT) of 2 seconds. In this research the signal timing pattern 3 was theoretically proposed to predict the efficiency and safety risks expected on signalized intersections.

The cycle lengths of pattern 3 are optimized with longer cycle's length of usually 3 more seconds comparing to pattern 2. Pattern 3 signal timing composes of PRT of 2 seconds and clearance interval as same as pattern 2.

3.1.4 Signal timing patterns evaluation

The main purpose of signal retiming effort performed by FDOT is to minimize the RLR frequencies and improve the level of safety for signalized intersections. However, potential issues are considered to be obvious of extending the change and clearance intervals. Both adopted and proposed patterns of signal timing during morning and evening peak hours were investigated using proper traffic measurements to evaluate each pattern and how it varies from the before signal timing pattern 1. Results are graphically illustrated to complete the picture of the analysis.

(1) The estimated delay per vehicle for signal timing patterns

The overall intersection delay was used to evaluate signal timing patterns based on the Synchro 8 simulation of the signalized intersections along the three corridors.

Figures 5 and 6 show the delay of the signal timing patterns for SR 50. From pattern 1 to patterns 2 and 3, it is evident that the general trend of delay rises. But from pattern 1 to patterns 2 and 3, the delay is more noticeable, whereas from pattern 2 to pattern 3, the delay is marginally acceptable. The reason is that pattern 1 has a shorter clearing interval duration than patterns 2 and 3, whereas patterns 2 and 3 have the same clearance interval time. Additionally, pattern 3, which have a longer cycle length and a longer PRT of two seconds. Some intersections were observed to have lower delay comparing with pattern 2. Certain approaches to those crossings have large volumes and short green intervals, which explains why there is less delay at those intersections with higher PRT and cycle duration. As a result, Synchro 8 views a longer PRT of change interval as a component of the green interval, which allowed for the discharge of more traffic and reduced the overall delay at intersections 3, 5, and 7.

5.png

Figure 5. Estimated signal timing pattern delays during SR 50's morning peak

6.png

Figure 6. Estimated signal timing pattern delays during SR 50's evening peak

Figures 7 and 8 show the estimated delays for the signal timing patterns on SR 535 and SR 536. During both morning and evening peak hours, the overall delay increases from pattern 1 to patterns 2 and 3. Intersection delay is notably higher when comparing pattern 1 to pattern 2 or pattern 1 to pattern 3, while the difference between pattern 2 and pattern 3 is relatively minor. Synchro 8 treats the longer PRT of the change interval as part of the green interval, which allows more vehicles to discharge and helps minimize overall delay at certain intersections, such as intersections 11, 14, 16, and 19.

7.png

Figure 7. Estimated delays of signal timing patterns during SR 535 and SR 536's morning peak hour

8.png

Figure 8. Estimated delays of signal timing patterns during SR 535 and SR 536's evening peak hour

(2) The estimated 95th percentile queue length for signal timing patterns

Queue length can accurately assess signal timing efficiency. To find out how effective it is to lengthen the change and clearance intervals on the examined signalized junctions and arterials, the queue length measurement for left turn movement was measured for three signal timing patterns. Actually, the FDOT traffic engineers thoroughly examined the timing of the left turn movement signal in order to reduce the possibility of RLR on the main thoroughfares in Central Florida. This study will use the three signal timings throughout morning and evening peak hours to examine the 95th percentile queue duration of the critical left turn movement of such an intersection.

9.png

Figure 9. Queue lengths of signal timing patterns during SR 50's evening peak

Figure 9 shows the 95th percentile queue lengths for the three signal timing patterns implemented on SR 50. The trend indicates that queue lengths generally increase as the durations of the change and clearance intervals increase. When comparing patterns 2 and 3 to signal timing pattern 1, most signalized intersections along the SR 50 corridor exhibit longer queue lengths. However, for several intersections under pattern 3, increasing the clearance interval appeared beneficial during morning and evening peak hours. The longer PRT of 2 seconds, included in the change interval of pattern 3, is treated by Synchro 8 as part of the green interval. This treatment allows for greater vehicle discharge and helps reduce queue lengths at critical intersections where green time is limited and traffic volume is high.

Most signalized intersections along SR 535 and SR 536 exhibited longer queue lengths as a result of the additional seconds added to the change and clearance intervals in signal timing patterns 2 and 3. Figure 10 illustrates the queue lengths for the three signal timing patterns across ten signalized intersections on SR 535. The results show that queue lengths increase with longer change and clearance intervals. The queue lengths observed under signal timing patterns 2 and 3 were significantly higher than those under pattern 1.

10.png

Figure 10. Queue length of signal timing patterns during SR 535 and SR 536's evening peak

(3) Intersection V/C ratio for signal timing patterns

V/C ratio is used with other traffic measurements to distinguish between the different signal timing systems. It is found that extending the change and clearance intervals increases the intersection V/C ratio.

The intersection V/C ratio for three signal timings at the signalized intersections on SR 50 during evening peak hours is displayed in Figure 11. As the cycle length grows, so does the average V/C ratio of the three signal timings? It is demonstrated that the mean V/C ratio of signal timing patterns 2 and 3 is higher than the average of pattern 1.

11.png

Figure 11. V/C ratio of signal timing patterns during SR 50's evening peak

12.png

Figure 12. V/C ratio of signal timing patterns for SR 535 and SR 536

Figure 12 shows the intersection V/C ratio for the signalized intersections located on SR 535 and SR 536 arterials during evening peak hours for three signal timings. To conclude, it was observed that the average V/C ratio for signal timing pattern 2 and pattern 3 was found significantly more saturated than signal timing pattern 1.

3.1.5 Signal timing patterns impact on a corridor

(1) Arterial total delay

Total delay (seconds per vehicle) is a measurement to estimate the arterial delay for a vehicle traveling with signal timing patterns system as shown in Figure 13.

It was observed that the additional time of change and clearance intervals significantly increases the total delay per vehicle traveling along the studied arterials. The new and proposed signal timing patterns 2 and 3 caused more total delay for a motorist traveling through the studied SR 50, SR 535, and SR 536 corridors.

13.png

Figure 13. Arterial total delays per vehicle during morning peak

(2) Total travel time for a corridor

Travel time along a corridor is a key performance measure for evaluating signal timing patterns. Both total delay and travel time were assessed for the three signal timing patterns during the evening peak hour. The microsimulation results across the three corridors revealed significantly higher delays and longer travel times for signal timing patterns 2 and 3. Specifically, patterns 2 and 3 resulted in 3% to 10% additional delay and increased travel time on the arterials compared to pattern 1, as shown in Figure 14.

14.png

Figure 14. Arterial total travel time

3.2 Red light running statistics

RLR is a safety risk issue in most metropolitan areas. The Florida States Department of Transportation, FDOT admitted this issue and investigates it widely for potential solutions. To mitigate this issue, the FDOT adopted the practice of prolonging the change interval to minimize running red light. In this part of the research, a naïve statistical method was performed to investigate the potential impact of additional change interval on red light running and if that reduces RLR. Based on limited data provided by Orange County FDOT for eight approaches located in six intersections during ten months, a naïve statistical method was performed as shown in Figure 15.

15.png

Figure 15. Average monthly red light running

Based on the naïve statistical method, it can be concluded that the new signal timing does not significantly reduce the frequency of red light running. However, the new signal timing shows a significant increase of RLR in congested approaches. To sum up, the limited data is not reliable enough to state such conclusion.

3.3 Dilemma and option zones study for SR 50 intersections

In order to assess how the new signal timing project would affect the potential safety risk of signalized intersections, a subset of signalized intersections on SR 50 for signal timing patterns 1, 2, and 3 were measured using Eqs. (4) and (5), which are described by Gazis et al. [5] and estimate dilemma and option zones.

$c=V 0 \delta 2+\frac{V 0^2}{2 a 2}$                          (4)

$\mathrm{X} 0=\mathrm{V} 0 \tau-\mathrm{W}+\frac{1}{2} \mathrm{a} 1(\tau-\delta 1)^2$                            (5)

where,

V0 = the approach speed (ft/s);

$\delta 2$ = the driver's perception-reaction time (sec);

a2 = the maximum vehicle's deceleration rate (ft²/s);

$\delta 1$ = the driver's perception-reaction time for running (sec);

a1 = the constant of the vehicle's acceleration rate (ft²/s);

τ = the duration of change interval (sec);

W = the summation of intersection width and the length of vehicle.

To put it simply, the area of the dilemma and option zones was quantified by obtaining the Xc and X0. The dilemma zone can be measured by subtracting X0 out of Xc. Otherwise, the option zone exists when the results of Xc-X0 are negative as shown in Table 4.

Table 4. Dilemma and option zones identification of intersection 2

Table 4 reveals meaningful implications for intersection safety. According to crash risk literature, particularly studies by Bonneson and Zimmerman [27] and Rakha et al. [28], a well-defined and minimized dilemma zone reduces the likelihood of indecisive driver behavior, which is a key contributor to red-light running and rear-end collisions. In this study, signal timing pattern 2 effectively reduced the size of the dilemma zone across all approaches, suggesting improved decision-making conditions for drivers as they approach the intersection at the onset of yellow. Pattern 3 further shortened the dilemma zone while extending the option zone, providing drivers with a larger buffer for safe manoeuvring decisions.

These outcomes imply that both patterns 2 and 3 have the potential to lower crash risks associated with late stopping or unsafe acceleration at the end of green intervals. As an actionable insight, agencies could consider implementing signal timing patterns that balance both reduced dilemma zones and extended option zones, especially at intersections with high-speed approaches or elevated red-light running histories. The use of vehicle trajectory data and high-resolution signal monitoring can further help traffic engineers identify critical approaches where targeted interval adjustments would yield the highest safety benefits.

3.4 Statistical analysis

This study evaluates the effects of prolonged change and clearance intervals on the operational efficiency and safety of signalized intersections and corridors. The change and clearance intervals function as independent variables, whereas the primary dependent variables include intersection delay, vehicle queue length, V/C ratio, and safety indicators, such as RLR frequency and dimensions of dilemma/option zones.

The following hypotheses were proposed, providing a way to test the effects of FDOT’s retiming strategy using simulations and statistical analysis:

1. The null hypothesis (H₀) posits that extending the change and clearance intervals at signalized intersections does not lead to statistically significant differences in intersection performance or safety, as assessed by metrics such as intersection delay, queue length, V/C ratio, and frequency of red-light violations.

2. Alternative hypothesis (H₁): Extending the change and clearance intervals at signalized intersections leads to statistically significant alterations in intersection performance and safety.

3. Prolonged cycle lengths result in heightened intersection delay and extended queue lengths.

4. Elevated V/C ratio signifies greater saturation levels.

5. Shortened dilemma zone lengths and extended option zone lengths may affect driver decision-making and potentially decrease the frequency of RLR.

A paired t-test was run to determine if the variations shown in the simulation results were statistically significant, therefore allowing an evaluation of the operational and safety consequences of the three signal timing patterns. The comparisons consisted of pattern 1 (pre-retiming) vs. pattern 2 (current FDOT implementation), pattern 1 vs. pattern 3 (suggested with 2.0 sec PRT), and pattern 2 vs. pattern 3.

Three main performance indicators were compared using these: delay at intersections (sec/vehicle), length of 95th percentile queue (ft), and the ratio of intersection volume to capacity (V/C).

Tables 5-8 summarize the results of the paired t-tests for both morning and evening peak hours.

Table 5. SR 50 paired t-test of signal timing patterns during morning plan

As shown in Table 5, significant differences were observed among the three signal timing patterns during the morning peak period along SR 50. Specifically, the intersection delay showed a statistically significant difference between pattern 1 and pattern 3 at the 95% confidence level, while the difference between pattern 1 and pattern 2 was significant at the 90% confidence level. No significant difference was found between the delays associated with pattern 2 and pattern 3.

The paired t-test does not illustrate any significant difference between the 95th queue lengths signal timing patterns. In contrast, the paired t-test showed a slight increase in the V/C ratio in pattern 2 and a significant increase in pattern 3 compared to pattern 1.

Also, it applied on SR 535 & SR 536 signal timing patterns seem to behave as the SR 50 signal timing during morning peak hour. Table 6 shows a significant increase in delay in pattern and pattern 3 at a 90% confidence interval while the delay mean deference was found insignificant between pattern 2 and pattern 3.

On the other hand, the paired t-test applied between pattern 1 vs pattern 2, pattern 1 vs. pattern 3, and pattern 2 vs. pattern 3 at 95th percentile queue length does not show any significant difference. The paired t-test of intersection V/C ratio provides significant differences between all signals timing pattern in 95% confidence interval.

Table 6. SR 535 and SR 536 paired t-test of signal timing patterns during morning plan

Table 7. SR 50 paired t-test of signal timing patterns during evening plan

Table 7 states significant differences between the means of the signal timings different measurements during evening plan along SR 50 corridor. It can be concluded that there are significant differences between the intersection delay means of signal timings pattern 1 vs. pattern 2 using a 90% confidence interval and pattern 1 vs. pattern 3 using a 95% confidence interval. Moreover, the 95% percentile queue length shows significant differences between signal timings pattern 1 vs. pattern 3 and pattern 2 vs. pattern 3. The intersection V/C ratio shows significant differences between all the signal timing patterns 95% confidence interval.

Table 8 shows that there are no significant differences between the traffic measurements pattern 1 vs. pattern 2, pattern 1 vs. pattern 3, and pattern 2 vs. pattern 3 for intersection delay and 95th queue length during the evening plan along SR 535 & SR536. However, it shows significant differences between all signal timing patterns for the intersection V/C ratio.

Table 8. SR 535 and SR 536 paired t-test of signal timing patterns during evening plan

3.4.1 Delay at intersections

Significant increases in delay were observed from pattern 1 to patterns 2 and 3, particularly along the SR 50 and SR 535 corridors during morning and evening peak times. The rise from pattern 2 to pattern 3 was modest and usually not statistically relevant.

This suggests that longer cycle lengths result from prolonging change and clearance intervals, hence lowering green time efficiency and raising the time every car spends at the junction. Although greater change intervals can potentially enhance safety by lowering red-light running, the related delay might affect travel time dependability, particularly during peak traffic. The non-significant difference between patterns 2 and 3 implies declining benefits when raising PRT beyond 1.4 seconds.

Observed increases in wait length, especially between pattern 1 and pattern 3. Most variations, however, were not statistically relevant.

Longer lines indicate lower junction throughput and possible spillback into upstream intersections. Although the rises are not always statistically significant, they may still have practical consequences in busy urban areas with limited queue space. The outcome indicates that while the longer intervals noticeably affect latency, their effect on queue lengths is more unpredictable and reliant on the situation.

3.4.2 Volume-to-capacity (V/C) ratio

All corridors and peak times exhibited very significant variations in all comparisons.

Pattern 1 had the lowest V/C ratio, followed by pattern 2, while pattern 3 had the highest.

The V/C ratio indicates how saturated an intersection is. A higher V/C ratio shows that crossings are functioning nearer to their capacity, hence raising the likelihood of congestion and operational failures. The results imply that especially under peak demand, prolonged change and clearing intervals might impair capacity, thereby driving junctions closer to oversaturation situations.

Traffic engineers need this knowledge especially. Although longer intervals may provide safety advantages, trade-offs in junction capacity have to be carefully considered, especially on arterials with large traffic volumes.

Table 9. Summary of operational implications