An Analysis of TDM Impacts on a Corridor Segment

Transcription

1 An Analysis of TDM Impacts on a Corridor Segment Phil Winters, CUTR TDM Program Director Liren Zhou, Graduate Research Assistant Sachin Rai, ITS Research Associate Nevine Georggi, TDM Research Associate Presentation Overview Study Background Review TDM Measures of Effectiveness Review Selected Transportation Operations Performance Measures Study Objective and Hypothesis Study Methodology Simulate impacts of employer-based programs affecting a segment of a corridor. Data available from CTR Data needed for CORSIM analysis With TDM Scenario The Without TDM Scenario CORSIM Analysis Preliminary Results Study Background This study seeks to establish the relationship between employer-based strategies and transportation system performance. The ultimate goal is the development of some methods or indices that measure the impacts of such programs on the overall transportation system and clearly communicate these impacts to policy makers and transportation decision makers. 1

2 TDM MOEs Florida CAP Evaluation Measurement Number of commuters requesting assistance Number of commuters switching modes Number of vans in service (where applicable) Number of vehicle trips eliminated Vehicle miles eliminated Employer contacts Parking spots saved/parking needs reduced Commuter costs saved Major accomplishments Gasoline saved Emissions reduced Information materials distributed Special events Media/community relations Transportation Operations MOEs National Transportation Operations Coalition few good measures Customer satisfaction Extent of congestion - spatial Extent of congestion - temporal Incident duration Non-recurring delay Recurring delay Speed Throughput - person Throughput - vehicle Travel time-link Travel time reliability (buffer time) Travel time trip Study will use for analysis of TDM impacts Study Objective This study will develop a methodology for measuring the impacts of employer-based TDM programs on transit system ridership and transportation system performance, in particular from the management and operations perspective. 2

3 Study Hypothesis The hypothesis of the study is that a wide scale adoption of employerbased strategies is likely to have a noticeable difference on the transportation system at the local, corridor and regional levels. Study Methodology Simulate impacts of employerbased programs affecting a segment of I-5 in Seattle. Simulate using CORSIM Employer-based programs from the Washington State CTR program database Now... to the nitty-gritty How to create network scenarios with TDM? without TDM? How to estimate impacts on network traffic with TDM? Without TDM? How to compare the scenarios? 3

4 The CTR Program The CTR program was created by the 1991 legislature to reduce the economic and environmental degradation caused by the increasing number of commute trips made by employees in Washington State. To accomplish this, the CTR program works with employers to encourage employees to commute without driving alone every day. The program also encourages transportation service providers to expand the opportunities available to employees for commuting in ways other than driving alone. The results of these efforts are demonstrated in the daily choices made by more than 56, employees at the 1,114 worksites participating in CTR. The CTR Program (cont.) Employers affected by the CTR law are required to submit an Employer Annual Report & Program Description form to report the summarized information of the programs they implemented. The affected employers are also required to measure employee commute behavior every two years to measure their progress toward their CTR goals. The employer annual report and employee biennial survey composed two databases that comprise information of employer s TDM performance and employee s travel behavior change. The CTR Program (cont.) Information contained in the annual employer report includes: Worksite description Employee information Program promotion information, include list of CTR programs implemented or promoted by the employer Worksite characteristics Worksite parking information and parking management Financial incentive and subsidies Site amenities Work schedule policy Other programs availability, include GRH, internal match, fleet vehicle, etc. 4

5 The CTR Program (cont.) Information contained in the individual employee survey includes: work schedule, commute trip mode split, compressed work week schedule, teleworking schedule, travel distance, employee job title, and home zip code. The databases are maintained and managed by Department of Public Transportation, Washington State DOT. Transportation Network Seattle Downtown Area High Density of employers participating in TDM programs in the area Number of employers Number of employees - 63 Planned improvements for the transportation network by the DOT Pavement reconstruction and bottleneck improvement project on I-5 Map of Transportation Network Study Area 5

6 Traffic Analysis Tool FHWA Traffic Analysis Toolbox Volume II: Decision Support Methodology for Selecting Traffic Analysis Tools The decision toolbox s recommendation: Microscopic Simulation Software FHWA Traffic Analysis Toolbox Decision Toolbox Criteria Analysis Context Geographic Scope Facility Type Travel Mode Management Strategy/Applications Traveler Response Performance Measures Tool/Cost Effectiveness WEIGHTED TOTAL Analyti Sketch TDM cal Plan (HCM) Analysis Tools Traffic Opt Macro Sim Meso Sim Micro Sim Microscopic Simulation Software Advantages of using Microscopic Second-by-second simulation for each vehicle Customize vehicle fleet type and driver type to represent field conditions Simulate interaction of traffic flow with roadway geometry Calibrate the model to match existing traffic conditions Graphical interface to visually observe and compare different scenarios Requires large amount of data collection effort 6

7 Traffic Analysis Tool : CORSIM CORSIM is a stochastic microsimulation tool which is being used by practitioners and researchers for the past 3 years to simulate and conduct various transportation related evaluations. Development of CORSIM being supported by FHWA and it is used by several state agencies including FDOT and WSDOT Comparable studies conducted in the past using CORSIM CORSIM Data was available for the Seattle region Research team has past experience and training in CORSIM CORSIM Analysis Evaluate the impact of TDM programs on the transportation network by comparing the performance of the network with TDM programs and without TDM programs Existing Traffic Volumes on the network = E Estimate traffic reduced by TDM programs = R Scenario A (with TDM) Performance of the transportation network with existing traffic volume E Scenario B (without TDM) Performance of the transportation network with traffic volume E + R Comparison of Scenario A and Scenario B TDM Impact Area 7

8 TDM Impact Area 3hr AM peak Volume : 19,9 3hr PM peak Volume : 2,6 SR Mile Segment 67 lane-mile 16 on-ramps and 19 off-ramps I-5 I-9 3hr AM peak Volume: 22,5 3hr PM peak Volume : 19,8 Comparing a segment of I-5 with and without TDM Scenario WITH TDM is the existing traffic conditions of a selected segment of I-5 (CORSIM Files from WSDOT) Scenario WITHOUT TDM will add the reduced trips, from the CTR data, to the existing traffic of the I-5 selected segment. WITH TDM Scenario Data from WSDOT CORSIM simulation data for the year 24 AM peak and PM peak period traffic volumes on freeway mainline On-ramps and Off-ramps volumes and location Existing Carpool volume Segment roadway geometry information Origin-Destination of the existing volume Calibration of the network 8

9 The WITHOUT TDM Scenario Introduction TDM Trip Reduction Estimation Step 1: Evaluate the impact of Telecommuting and Compressed Work Week Step 2: Estimate the impact of Employer TDM Support Strategies Step 3: Estimate the impact of Parking and Financial subsidies Step 4: Estimate the number of trips shifted out of peak period due to the Flexible Work Hours Trip Distribution Trip Assignment Preliminary Results WITHOUT TDM Scenario Introduction Objective Provide link traffic flow change due to TDM Programs as input to CORSIM micro traffic simulation model Methodologies regarding Estimating TDM reduced vehicle trips at employer level Distributing the estimated reduced vehicle trips to the Worksite- Home pairs Assigning the reduced vehicle trips to the highway network Output includes Number of reduced vehicle trips for each worksite Traffic flow change for each on and off ramp of I-5 within the study area On and off ramp ID for each reduced vehicle trip Models Review TDM Trip Reduction Estimation Literature Review Air Resources Board (California) Cost-Effective Model Washington State TDM Effectiveness Estimation Methodology (TEEM) Model Environmental Protection Agency (EPA) COMMUTER Model CUTR Worksite Trip Reduction Model Major Characteristics of These Models Focus on predicting or estimating the possible travel behavior change currently or in the future based on projected social economic data and TDM programs 9

10 TDM Trip Reduction Estimation Summary Our Research Goal Quantitative analysis of the effect of the TDM programs Our Data Washington State Commuter Trip Reduction Database provides Detailed description of TDM strategies implementation Corresponding travel behavior survey data Our Methodology Tracing back the situation without TDM by estimating the travel behavior without TDM Comparing the scenarios with and without TDM to analyze the quantitative impact of TDM programs Based on EPA s COMMUTER model TDM Trip Reduction Estimation Washington State Commute Trip Reduction Database Washington State Commute Trip Reduction Database Employer Annual Report Employer basic information Alternative work hours strategies Employer TDM support strategies Parking and financial subsidies or incentives Employee Biennial Travel Behavior Survey Work schedule Commute trip mode Job type Home ZIP code TDM Alternative Work Hours: Reduce person trip Employer TDM Support Strategies: Reduce the drive alone trip Parking and Financial Subsidy: Reduce the drive alone trip Flexible Work Hours: Shift vehicle trip out of peak period With TDM TDM Trip Reduction Estimation Definition Some or all of the TDM programs are implemented by the employer Vehicle trips with TDM can be calculated directly from Washington state employee biennial survey data Without TDM Some or all of the TDM programs are not implemented by the employer Develop a procedure to calculate vehicle trips without TDM based on EPA COMMUTER model and Washington CTR data 1

11 TDM Trip Reduction Estimation Vehicle Trips and Mode Shares With TDM Treat all non-respondents as drive alone, calculating the Vehicle Trips with TDM Vehicle Trip WithTDM = Non-respondents + DriveAlone + Motorcycle + Carpool 2 /2 + Carpool 3 /3 + Carpool 4 /4 + Carpool 5 /5 + Vanpool/6 Treat all non-respondents as drive alone, calculating the Mode Shares with TDM TDM Trip Reduction Estimation Vehicle Trips Without TDM Step 1: Evaluate the impact of Telecommuting and Compressed Work Week The Telecommuting and Compressed Work Week are treated as direct results of TDM. The employees on Telecommuting and CWW are added back to calculate the number of person trips if they were not allowed Revised Person Trip = Non-respondents + DriveAlone + Motorcycle + Carpool + Vanpool + Transit + Bicycle + Walk + CWW + Telecommuter TDM Trip Reduction Estimation Vehicle Trips Without TDM Step 2: Estimate the impact of Employer TDM Support Strategies Starting Mode Share (With TDM) Carpool: 11% Vanpool: 5% Transit: 15% Bicycle: 3% Carpool TDM support program Program Level Δ Share -1.5% -2% -3% -5% Vanpool TDM support program Program Level Δ Share -.5% -1% -1.5% -2% Transit TDM support program Program Level Δ Share -1% -2% -4% -5% Bicycle TDM support program Program Level Δ Share -.1% -.3% -.5% -1% Carpool: -2% Vanpool: -1% Transit: -2% Bicycle: -1% Revised Mode Share (Without TDM) Carpool: 9% Vanpool: 4% Transit: 13% Bicycle: 2% 11

12 TDM Trip Reduction Estimation Vehicle Trips Without TDM Step 3: Estimate the impact of Parking and Financial subsidies Assumption Without TDM program, parking for all modes is free Subsidies or incentives for alternative modes are direct results of TDM Travel time change is Calculating the changed mode share due to the travel cost change p' ( m) U ( m [ e (1 p' ( m)) ] + p' ( m) p( m) = ) where: p(m) = the new share of mode m ( Without TDM ) p (m) = original share of mode m ( With TDM ) U(m) = the change in disutility of mode m ( U(m) = a*(changed travel time)+ b*(changed travel cost) ) TDM Trip Reduction Estimation Vehicle Trips Without TDM Step 4: Estimate the number of trips shifted out of peak period due to the Flexible Work Hours Number Daily Vehicle Trips Shifted, Peak to Off-peak = Total Affected Employment x Percent of Employment that is Office x 22% Participate in Program x Percent of Trips Shifted (based on length of peak period) x Current Private Vehicle Mode Share TDM Trip Reduction Estimation TDM Reduced Vehicle Trips Vehicle Trips without TDM = Revised person trip * (R-DriveAlone + R-Motorcycle + R-Carpool2/2 + R-Carpool3/3 + R-Carpool4/4 + R-Carpool5/5 + R-Vanpool/6) + Vehicle Trips Shifted out of Peak Period Total Number of Reduced Vehicle Trips = Vehicle Trips without TDM _ Vehicle Trips with TDM 12

13 Trip Distribution Objective Find origin and destination for each reduced trip Assumption The reduced trips are derived from the non drive alone trips Distribute the reduced trips to Worksite-Home pairs proportionally according to the percentage of non drive alone trips Non Drive Alone Total Non Drive By Reduced 1*Distributed Worksite HomeZip Trips By HomeZip Alone Trips HomeZip Trips Reduced Trips 1 E % E % E % E % E % E % E % 1 1 Trip Assignment Create a transportation network of five counties, including King, Island, Snohomish, Kitsap, and Pierce, based on the Census Tiger/Line road data Geocode the geographical location of worksites based on their street address Use the centroid of the ZIP to represent the home location Find the shortest path between the worksite-home pairs based on the link length Assign the traffic based on the so called All-or-Nothing traffic assignment approach Final results include the changed traffic flow on each on and off ramp of I-5 and the on and off ramp ID for each reduced trip 13

14 Total number of worksites Total number of affected employees Total number of reduced vehicle trips Average percentage of TDM vehicle trip reduction Average Share of Driving Alone Average Share of Transit Total number of person trips Total number of vehicle trips Total number of reduced vehicle trips takes I-5 Total I-5 ramp traffic flow Average percentage of I-5 ramp traffic flow change Preliminary Results 56.91% 28.8% AM Peak Period % With TDM % 64.49% 2.39% PM Peak Period 3815 Without TDM % Preliminary Results Ramp traffic flow for AM peak period (Home to Worksite) On Ramp I-5 NB South End I-5 SB North End Corson NB Spokane NB I-9 NB University NB Oliver NB Mercer NB SR52 NB Harvard NB Spokane SB 6th SB 4th SB Yale SB Mercer SB Boylston SB SR52 SB 45th SB Total Original Flow Flow Change % Off Ramp I-5 SB South 6.42% End I-5 NB North 1.13% End 6.49% Spokane NB 2.1% I-9 NB 4.37% 4th NB.64% Seneca NB % Oliver NB.24% Mercer NB % Lakeview NB % SR52 NB % 45th NB.15% Corson SB % Spokane SB.7% Forest SB 1.44% I-9 SB 8.61% 6th SB 2.1% Union SB % Stewart SB Mercer SB SR52 SB Boylston SB Total Original Flow Flow Change % % %.65%.39% 48.8% 39.3% 1.9% % 4.46%.24% % %.47%.2%.45% 6.22% 39.36% 19.82% 5.48% 2.41% % 14

15 Preliminary Results Ramp traffic flow for PM peak period (Worksite to Home) On Ramp NB South End SB North End Corson NB Spokane NB I-9 NB University NB Oliver NB Mercer NB SR52 NB Harvard NB Spokane SB 6th SB 4th SB Yale SB Mercer SB Boylston SB SR52 SB 45th SB Total Original Flow Flow Change % % % %.32%.73% 55.83% 1.82% 7.48%.77% 4.18% 2.29% 23.82% % 11.44% 1.21% 1.67%.6% % Off Ramp SB South End NB North End Spokane NB I-9 NB 4th NB Seneca NB Oliver NB Mercer NB Lakeview NB SR52 NB 45th NB Corson SB Spokane SB Forest SB I-9 SB 6th SB Union SB Stewart SB Mercer SB SR52 SB Boylston SB Total Original Flow Flow Change % 1.55% %.3% %.36% % % 14.99% 2.72% % 6.15%.8% % 1.2%.13%.4% %.8% % % 3.71% I-5 SB North On Original Flow Flow Change % Boylston SB On Original Flow 1312 Flow Change 113 Original Flow 8.61% Flow Change I-5 SB North End (On) Stewart SB Off Original Flow 2477 Flow I-5 Change SB North 491 End (On) 19.82% Original Flow Union SB Off Original Flow 2726 Flow Change 173 Flow Change 39.36% % Seneca NB Off Original Flow % Flow Change % I-5 SB North End (On) Corson NB On Original Flow 2358 Flow Change Original Flow % Flow Change I-5 NB South On Original Flow 2981 Flow Change % % 4 th NB Off Original Flow 443 Flow Change % Oliver NB On Original Flow 4712 Flow Change % Yale SB On Original Flow 499 Flow Change 469 Original Flow 11.44% I-5 SB North End (On) I-5 NB North Off Original Flow Flow Change I-5 SB North 1867End (On) 1.55% Original Flow Flow Change 1884 Flow Change % 1.13% University NB On Original Flow 2192 Flow Change Original Flow % Flow Change 6 th SB On Original Flow 4278 Flow Change % I-5 SB North End (On) % I-5 SB South Off Original Flow Flow Change % 15

16 CORSIM Analysis g1 Existing Traffic Volumes on the network: WSDOT Estimate traffic reduced by TDM : CTR database Comparison of Scenario A and Scenario B: CORSIM CORSIM Analysis Scenario A (with TDM) Performance of the transportation network with existing traffic volume Scenario B (Without TDM) Performance of the transportation network with existing traffic volume + Traffic volume reduced by TDM programs Comparison of Scenario A and Scenario B Performance Measures Delay in vehicle-hours and seconds per vehicle Average Speed in mph Fuel Consumption in gallons Emissions in grams/mile Extent of Congestion Spatial Temporal 16

17 Performance Measures Spatial Extent of Congestion Length of roadway within a predefined area and time period for which average travel time are 3% longer than unconstrained travel time Temporal Extent of Congestion The time duration during which more than 2% of the roadway sections in a predefined area are congested Source: NTOC Performance measurement Initiative Final Report Performance Measures Delay in vehicle-mins 35, 3, 25, Delay (veh-mins) 2, 15, 1, 5, Time Period Scenario A (with TDM) Scenario B (without TDM) Performance Measures Delay in secs/vehicle Delay (sec/vehicle) Links on SB I-5 Freeway Mainline Scenario A (with TDM) Scenario B (without TDM) 17

18 Performance Measures Average Speed in mph Speed (mph) Time Period Scenario A (with TDM) Scenario B (without TDM) Performance Measures Average Speed in mph Speed (mph) Links on NB I-5 Freeway Mainline Scenario A (with TDM) Scenario B (without TDM) Performance Measures Fuel Consumption 5 Fuel Consumption (gallons/lane-mile) Links on SB I-5 Freeway Mainline Scenario A (with TDM) Scenario B (without TDM) 18

19 CORSIM DEMO 19