Development of TSP warrants considering vehicular emissions and day-to to-day variability in traffic demand Zeeshan Abdy Supervisor: Bruce R. Hellinga Dec 2006
Introduction Signalized intersections Traffic Signal Warrants Volume (vehicular and pedestrian) School crossing Crash experience Signal coordination and roadway network
Introduction Highway Capacity Manual (HCM) and Canadian Capacity Guide (CCG) delay based procedure Signal design objective Volume and Delay relation Averag ge Delay (seconds) 120 100 80 60 40 20 0 Degree of Saturation (X) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 F E D C B A of Service Level Input Parameters Geometric Traffic Volume Signal Peak Hour Factor Saturation Flow rate Capacity v/c Performance Measures Delay LOS Queue Volume Distribution
Motivation Degree of Variability Impact on intersection delay Design and evaluation methods Performance variation Delay (sec c) 150 125 100 75 50 25 3-days averaged Delay Delay using 3-days averaged volume 0 0 1 2 3 4 Performance variation 0 1 2 3 4 Day
Motivation Transit Signal Priority (TSP) Preferential treatment to Transit vehicle (person delay) Expense to cross street customers Average performance TSP impact on delay with day to day variability TSP impact on vehicle emission Best Scenario to apply TSP
Objectives Overall Specific Develop a TSP warrant methodology based on the developed relationships that can be used by transit agencies to aid in TSP implementation decisions. 1. Quantify, on the basis of empirical data, the day to day variation in peak hour volume, PHF, and base saturation flow rate. 2. Develop an analytical relationship that reflects the day to day variation of peak hour demand, peak hour factor and base saturation flow rate. 3. Develop a signal analysis and design methodology that is based on the existing HCM and/ or CCG methods, but which explicitly considers day to day variability. This methodology will also address the implications for field data collection (e.g turning movement counts). 4. Examine the impact that TSP parameters, control strategies, signal timings, and variation in traffic demand have on TSP performance. 5. Develop a relationship between TSP performance measured in terms of delays and emission impacts. This relationship will help in assessing TSP impacts from an environmental perspective. 6. Develop a TSP warrant methodology based on the developed relationships that can be used by transit agencies to aid in TSP implementation decisions.
Transit Signal Priority Signal Controller C Check in Detector Check out Detector A simplified representation of Transit Priority at Traffic Signals Prioritized approach Non-prioritized approach No Modification Prioritized approach Non-prioritized approach Green extension Transit Check in Detector Transit Check out Detector Min TSP green interval Adjustable Green extension time on non-prioritized approach Green interval on prioritized approach All red interval Amber interval Figure 12: Illustration of Green extension process
Transit Signal Priority Analytical models Weighted delay model Sunkari et al., (1995) Spare green time Skabardonis (2000) Priority Strategies Offline Online (active, adaptive)
Transit Signal Priority Literature Review Traffic Volume Cycle length Signal Phasing Transit Arrival phase Transit demand TSP Strategies Bus Stop and Detector location Emissions
Day-to to-day variability Traffic Demand, Sullivan et al., (2006) Sullivan research for PHV COV for City of Malwauke was found (4.8% to 15.5%) mean of 8.9%. A 10% increase in design hour volume (above the average) would result in 15% failure rate. Specifically for intersection operating at LOS D, a 10 % increase in traffic volumes would cause deterioration to LOS E or LOS F, about 15% of the time. Tarko model for PHF > Results > Problems Delay (Sec/ veh) 70 60 50 40 30 20 10 0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 Degree of Saturation Kamarajugadda et al., (2003) Delay Variability 1.00 0.95 0.90 0.85 0.80 0.75 0.70 Morning Rush Population < 20000 Morning Rush Population > 20000 Afternoon Rush Population < 20000 Afternoon Rush Population > 20000 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 Tarko Model PHF
TSP Emissions Non Prioritized Approach Stop and go condition Waiting (Delay) times for non prioritized approach Reducing delay and Reducing emissions No guidance for TSP implementation in terms of emission. Prioritized Approach
Framework to establish TSP warrant methodology
Scenarios of Interest (2,3) TSP studies in literature > Scenario of Interest Component Parameters combination Pm Road way geometry and Traffic Signal system Transit System Signal operation Impact of Pedestrian movement, Variability in day-to-day traffic peak hour volume, peak hour factor and base saturation flow rate. Bus Headway or frequency Impact on emissions Cycle length, c/g ratio, v/c ratio, Control strategy and restriction on TSP calls, Impact on emissions.
Day-to to-day variability (1,4) Day-to-day variability in PHV, charts results and ddi discussions i 100% LOS A LOS B LOS C LOS D LOS E LOS F 90% 80% ve Probability (%) Cumulativ 70% 60% 50% 40% 30% 20% 10% Perfectly Correlated (ρ = 1.0) Uncorrelated (ρ = 0) Average Correlation (ρ = 0.3) High Level of Correlation (ρ = 0.55) 0% 0 20 40 60 80 100 120 140 160 Average Intersection Delay (seconds)
Day-to to-day variability (1,4) Day-to-day variability in PHV, charts results and ddi discussions i De elay (Sec/ veh) 300 250 200 150 100 50 160 Kamarajugadda (2003) Kamarajugadda 95% confidence limit 140 Proposed methodology Proposed Methodology 95% confidence limit Intersec ction Delay (Sec) 120 100 80 60 40 20 Perfectly Correlated (ρ = 1.0) 95% confidence limit (ρ = 1.0) 10) Uncorrelated (ρ = 0) 95% confidence limit (ρ = 0) 0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 Degree of Saturation 0 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 Degree of Saturation
Day-to to-day variability (1,4) Day-to-day variability in PHF, Base saturation flow rate 0.95 0.94 0.93 0.92 Average (Empirical) Empirical Data Estimated using Tarko's equation r Factor Peak Hou 0.91 0.9 0.89 0.88 Average (Tarko) 0.87 0.86 0.85 62 1 way 182-WB 184-WB 184-EB 290-WB 312-NB 313-NB 313-SB 484-NB 484-SB Station
Vehicle Actuated programming, batch and post processing (5,6,7) VAP, Batch processing, Post processing units
Vehicle Emission Model (8) Microscopic Vehicle Emission model Second by second activity data ( speed and acceleration) converted to emission i
Development of Warrant methodology(9,10) Statistical models TSP warrant methodology Step by step procedure for practitioners.
Future Tasks 1. Quantification of variability in base saturation flow rate estimation 2. Investigation and/ or development of HCM methodology based on day-to-day variability 3. Calibration and Validation of Simulation Model 4. Development and processing of Scenarios of Interest 5. Development of Comprehensive Vehicle Actuated Programming (VAP) code 6. Validation of Emission Model 7. Development of relationship for TSP in terms of delay and emission i 8. Development of TSP warrant methodology based on the developed relationships Frame work Component Task Research Objective Framework Component 1 1,2 1,4 2 3 1,4 3 4 5,6 4 4 2,3 7 5 4 5,6 6 5 8 7 5 9 8 6 10 2006 2007 2008 Fall Winter Summer Fall Winter