Modeling the Response to Parking Policy

Transcription

1 Modeling the Response to Parking Policy Yoram Shiftan Transportation Research Institute Dep. of Civil and Environmental Eng. Technion, Israel Institute of Technology Visiting Professor IVT, ETH IVT, June 2015

2 Outline Parking in Tel-Aviv The activity-based (ABM) approach to model travelers response to policy (parking and other) The recent stated preference parking survey and model Integrating the parking model into the ABM Congestion pricing study (if times permit)

3 Some Facts About Parking in Tel Aviv

4 Parking Spaces per 1000 CBD Jobs

5 Parking Lot Occupancy Distribution 200% בית מונטיפיורי מוזס 1 רבניצקי חניון המגדל דובנוב 2 מפעל הפיס התרבות עמגר מגדלי רובינשטיין רבע חינם בית הדר סעדיה גאון MMN הצפירה חומה ומגדל גיא רובע לב העיר לב ת"א ע. לירון 180% 160% 140% 120% 100% 80% 60% 40% 20% 0%

6 Street Parking Occupancy Distribution 160% 150% 140% 130% % 110% 100% 90% 80% 70%

7 7 Focus Area Study The study focuses on an area that includes quarter 6 and the eastern parts of quarters 3 and 5 of the city of Tel Aviv-Jaffa.

8 Current Parking Characteristics Respondents parking characteristics (mean): Parking location parking lot street Total Parking search time (min.) These data are based on: Walking to destination time (min.) Parking cost per hour (NIS) A recent survey of more than 1,000 parkers in the area The survey was conducted between 7AM and 7PM, according to a distribution of sub-areas in parking lots and specified streets in the research area. Respondents were recruited in parking lots/on the street and the survey was completed via internet/phone.

9 Distribution of Parking Search Time (Min.) parking lot street

10 Distribution of Parking Walk Time (min.) parking lot street

11 Distribution of Hourly Parking Cost Paid parking lot street

12

13 Congestion Patterns at AyalonHighway hourly congestion levels monthly average congestion levels Journey Time (sec) 18:00 May June July Day Of week Time 08:00

14

15 Zurich Munich Copenhagen Sydney Average Speed of Public Transport Average Speed of Public Trasnport Mexico city Santiago Cairo Lyon Barcelona Rome Brussels Toronto Viena Berlin London Helsinki Mumbai Prague Chicago Oslo Paris New York Tel Aviv Amsterdam Rio de Janeiro Km/h

16 The Activity Based Approach to Model Response to Policies

17 Why Activity-Based Approach?

18 Complex Responses to Pricing Pre-Toll Schedule Space (Home) Car Work (a) Change Mode & Pattern Potential Responses to Toll Space (b) Change Time & Pattern Car Space Bus Work Work (c) Work at Home Car Space Shop Time Shop Car Time Shop Time Shop Time = Peak Period

19 Basics of Activity-Based Travel Theory Travel demand is derived from demand for activities. People face time and space constraints that limit their activity schedule choice. Activity and travel scheduling decisions are made in the context of a broader framework: Conditioned by outcome of longer term processes. Scheduling process interacts with the transportation system.

20 Activity Schedule Approach Dinner Secondary Tour Evening Home AM peak Shopping Primary Tour PM peak Work

21 The Tel-Aviv Model Structure car availability main activity work education shopping other no tour Time periods Morning AM peak midday PM peak evening Detailed Time Slices 34 half-hour slices early AM late PM

22 Model Structure (cont.) main destination destination 1 destination 2 destination 3 destination destination 1219 Tour Main Mode Revealed Preference NTHS and Rail Corridor Survey Stated Preference New SP Survey and NTA Survey taxi driver passenger bus rail P&R K&R Walk P&R K&R Walk Bus

23 intermediate stops no stops before after before & after work education shopping other destinations of intermediate stops destination 1 destination 2 destination 3 destination destination 1219 mode switching same mode switch

24 Parking Analysis Parking variables in the mode choice models: Cost Search time Walk time to destination Search time = f (parking supply, demand, other) Demand by time of day (occupancy) estimated endogenously from the model Supply data from the Tel-Aviv GIS data

25 The Tel Aviv GIS Data Public parking garages by statistical zones Private parking garages by statistical zones Curb side parking Parking supply by statistical zones were distributed to TAZs in proportion to population and employment data

26 Parking Data (Only in Tel Aviv) Public parking lots by statistical zone Private parking lots by statistical zone Curb side parking

27 Parking Search Time Model ( β + β x + β x + ) Search Time = Demand Supply Variable B t-value Explanation (Constant) Constant RMALED Male Dummy (1 if person is Male). RAGE Young age Dummy (1 If person < 24 years) RDENSITY Density of Area (Population/Area in km 2 ) REMPDENSITY Employees Density (Employees/Area in km 2 ) RAMPEAKD AM Peak Period Dummy (1 if Trip is in AM Peak Period) RPMPEAKD PM Peak Period Dummy (1 if Trip is in PM Peak Period) RAMOFFPD AM Off-Peak Period Dummy (1 if Trip is in AM Off-Peak Period) µ 1.0 µ No significant improvement obtained for nonlinear models Time of Day Minimum Search time Maximum Search Time (min) (min) a.m. Peak a.m. Off-peak p.m. Off-peak p.m. Peak

28 Base list of persons Zonal data POPULATIONGE NERATOR Persons by zone Networks TOUR GENERATOR Estimated models Level of Service Full list of tours Initial demand TRIP ASSIGNMENTS OD matrices TRIP GENERATOR

29 Extending the Framework Urban development Household decisions Residential choice Kids arrangs Work place Parking Transit Auto ownership ABA Shopping behavior Activity participation (location, sequence, scheduling, mode) Driver s decisions (route, parking) Transportation system performance

30 The Recent Parking Stated Preference Study

31 Parking Stated Preference Survey Data asked in the survey: Socioeconomic: age, gender, work status, etc. Who pays for car expenses, parking and gas? Travel & activity behavior: Tours done by person. Parking scenarios: Price: + 0/5/10 NIS Walk time: + 0/5/10 MIN Search time: +0/5/10 MIN

32 Stated Preference by Trip Purpose to TLV Center continue to use private car private car as passenger public transportation walk bicycle cab change my departure time change my destination not center TLV unknown other shop. leisure errands edu. home work Trip purpose

33 Stated Preference by Hourly Price Increase continue to use private car private car as passenger public transportation walk bicycle cab change my departure time change my destination

34 Models Estimation Results Binary model (deciding between private car and other mode): Value of Time Parking search time Parking walk time NIS per hour

35 Probability of Continuing to Drive

36 A Recent Congestion Stated Preference Study

37 Congestion Pricing Area Tel Aviv City

38 Congestion Hours חלופת Daily Hours יממה 06:30-19:00 שעת שיא בקר בלבד Morning Peak Hours 06:39-09:30 Morning Peak Hours שיאבקרלכניסה ושיאערבליציאה Afternoon Peak Hours 06:30-09:30 16:00-19:00

39 Response to Congestion Pricing Choice Distribution: Change Trip Time 20% Drive to Another Destination Cancel Trip 9.8% 4.2% Pay the Toll 44% Use Public Transit 20%

40 70% Response by Toll Level 60% 50% Pay the Toll 40% 30% 20% 10% 0% Public Transport Change Trip Time Change Destination Cancel Trip

41 Response by Income Level Pay Toll Distribution: Low Income High Income

42 Response by Main Activity Pay toll distribution vs. purpose: 50.0% 46.0% 42.0% 38.0% 34.0% Other Education Shopping & Maintenance Discretionary Work 30.0%

43 Workers Responses Workers 12.0% Percentage 10.0% 8.0% 6.0% 4.0% 2.0% Pay Toll Public Transit Change TT Cancel Trip Change Dest 0.0% Toll

44 Non-Workers Responses NonWorkers 5.0% 4.5% 4.0% 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% Toll Pay Toll Public Transit Change TT Cancel Trip Change Dest

45 Multinomial Logit Results Variable Pay Toll Public Transit Choice Utility Change Time Cancel Trip Change Destination Constant 4.5 [t-test] [21.1] Constant 1.96 [t-test] [15.87] Constant 1.63 [t-test] [16.44] Constant 1.1 [t-test] [9.88]

46 Multinomial Logit Results: Continued Socioeconomic Variables Variable Pay Toll Public Change Cancel Transit Time Trip Change Destination Religious Dummy (1 if Person is Not Hiloni ) [t-test] [2] [2] [2] [2] Single Dummy (1 if Status is Single ) [t-test] [1.41] Male Dummy (1 if Gender = Male ) [t-test] [-3.22] Middle Age Person (1 If Person is Years Old) [t-test] [0.85] Employee Dummy ( 1 if Employment is Employee ) [t-test] [3.25] Freelancer Dummy ( 1 if Employment is Self- Employed ) [t-test] [3.57]

47 Multinomial Logit Results: Continued Scenario Variables Variable Pay Toll Public Transit Car Travel Time Saved When Toll Applied [Minutes] [t-test] [3.15] High Complexity Tour Dummy (1 if Tour is not Simple i.e. S, H-P-H, H-P-O-H ) [t-test] [1.3] Public Transit "Bus" Cost [Shekels] [t-test] [-3.09] Toll Paid in the Scenario for Average Income [Shekels] [t-test] [-11.36] Toll Paid in the Scenario for High Income [Shekels] [t-test] [-9.44] Toll Paid in the Scenario for Low Income [Shekels] [t-test] [-13.67] Change Time Cancel Trip Change Destination

48 Multinomial Logit Results: Continued Purpose Variables Variable Pay Toll Public Transit Main Activity Work Dummy (1 if Main Tour Activity is Work) [t-test] Main Activity Discretionary Dummy (1 if Main Tour Activity is Discretionary) Change Time Cancel Trip [t-test] Main Activity Education Dummy (1 if Main Tour Activity is Education) [t-test] 2.73 Main Activity Maintenance Dummy (1 if Main Tour Activity is Maintenance) [t-test] 3.64 Live in Congestion Pricing Area Dummy (1 if Person lives in Congestion Toll Area) [t-test] Change Destination

49 Value of Time High Income: 13.0 NIS/Hr Mid Income: 10.3 NIS/Hr Low Income: 9.2 NIS/Hr

50 Scenario Analysis Percent change for morning peak hour, 2015, all the metro area toll of 15 NIS for entering, 3 NIS for residents. Small Ring Trips -4 Speed (base 30.5) KMH +4 Travel Hours -7 Travel KM -3 Medium Ring Large Ring Small Area Medium Area Large Area The large ring don t add much benefits as many of those entering the medium ring live in the large ring.

51 Speeds by Ring: No Toll Scenario Analysis 11.9 Small Ring Medium Ring Large Ring Small Ring % % % Medium Ring % % % Large Ring % % % Small Area % % % Medium Area % % % Large Area % % %

52 Scenario Analysis Morning Only Period Base Auto hourly Pass (1000s) Base Transit Pass hourly (100s) % Auto Change % Transit Change 06:00-07: :00-08: :00-09: :30-12: :00-16: :00-19: :00-20: Daily Total Daily both modes -0.05

53 Thank You For Your Attention!