1. Presented by: Ms. KAMINI GUPTA CSIR-Central Road Research Institute
PROVISION OF SUSTAINABLE ROAD TRANSPORT INFRASTRUCTURES:
A CASE STUDY OF URBAN CORRIDOR IN DELHI
Presented by: Ms. KAMINI GUPTA
CSIR-Central Road Research Institute
New Delhi, India
Authored by: Ms. Kamini gupta, Dr. Ch. Ravi sekhar, Dr. B. Kanaga Durai
& Dr. Ravindra Kumar
15th International conference foe women Engineers & Scientists, Adelaide Convention Centre, Adelaide, Australia, 19th-22nd July, 2011

2. Outline of the Presentation
Introduction
Sustainable Transportation System & its necessity
Study area Characteristics
Identified Problems in the Study Area
Methods of Data Collection
Data Analysis and Discussion
Provision of Various Infrastructure facilities
Evaluation of sustainability under various strategies
Driving Cycle Patterns & Fuel Consumed
Conclusions and Recommendation

3. Introduction on Sustainable Transportation Measures
Social:
Reduce Congestion: improve mobility and reliability
Enhance Safety: reduce crash rates and traffic incidents
Economic:
Expand Economic Opportunity: Optimise land use and improve vehicular movement
Preserve Value of Transportation Assets: maintain pavement condition, capacity addition etc
Environmental:
Improve Air Quality: reduce green house gas emissions
Ref: NCHRP 8-74 and Zietsman Josias et.al 2008

4. Necessity of the Present Study
No quantitative research and implementation of sustainable transportation infrastructures concept for road networks.
To evaluate sustainability of road transportation system for current and future transportation system scenarios.

5. Objective of the Study
To evaluate sustainability of road transportation system by Provision of Various Road Infrastructures.

6. Applications of Sustainable Measures
Route Guidance
Travel Time Information
Sustainable Transportation system Measures
Commuters
Transportation System Planners/ Policy Makers
Freight Operators
Road Network Operations
Road Network Planning
Routing
Scheduling

7. Review of Travel Time Reliability Measures
Statistical Measures
Standard deviation , and Travel time window.
Coefficient of variation
Reliability Measures
95th or 90th % Travel Time: Travel time on some of the heaviest traffic days.

8. Case Study - Urban Corridor on NH 2
Study Area: Urban Corridor on NH 2 ( Ashram Jn. to Apollo Hospital)

9. Study Area Characteristics
The total Traffic Volume (both directions) of this corridor near CRRI is about 1,37,569 vehicles per day. ( CRRI Report 2009).
Badarpur to Ashram direction was observed bit more than the other direction ( vehicles).
On an average 7% of the total traffic was observed during morning peak hours and evening peak hours.
This indicates that during the periods from 9.00 am to 8.00 pm the travel time is very uncertain because of high amount of traffic.
In this study mainly emphasized for measuring the travel time variability for the morning hours 8.30am to 10.30am and 4.00pm to 6.00pm.
Traffic Composition
(Badarpur to Ashram )
Traffic Composition
(Ashram to Badarpur)
Time versus Percentage of Total Traffic

10. Definition of Problems in the Study Area
Section 1: From Ashram Chowk to Grandly Junction ( about 1.50 km length)
Intersection Spacing: Two signalized junctions are with in km road stretch. The spacing between these signals is about km.
Traffic Incidents: Signal failures, vehicle breakdown, cleaning, marking etc during rush hours.
Right turning traffic: Traffic behavior at Matamandir Intersection and Grand lay Intersection.

11. Problem definitions in the study Area
Section 2: From Grandly Junction to CRRI Bus Stop ( 1.40 Km length)
Merging and Diverging behavior: at Modi Mill flyover
Pedestrian crossing: Crossing at un pedestrian signal
Bus Stops: Unauthorized bus stop near merging of the ramp on NH2
U turn: Medium and Large size buses taking U turn in front of CRRI median opening from inner lane during rush our.

12. Problem definitions in the study Area
Section 3: From CRRI Bus Stop to Apollo Hospital ( Length 1.4 Km)
Pedestrian crossing: High number of Pedestrian crossing during peak hours at signalized junctions in front of CRRI (often manually operated at this signal)
Pedestrian signal : Specially at Okhla tank, traffic stops for 12 sec creating congestion during peak hours
Location of Bus Stops: Location of CRRI Bus stop is immediately after junction.

13. Structure of Travel Time (TT) Fluctuation for Study Area
Demand Side Factors
With in day and day-to-day demand fluctuation
External Factors
TT Fluctuation
adverse weather
Supply Side Factors
vehicle Break down, traffic accident, signal failure,
roads works, pedestrian crossing at unsignalised location etc.

14. Examining Travel Time Reliability on Urban Corridor of NH 2 (Ashram to Apollo)
Traffic Flow
Traffic Flow

15. Data Collection Travel Time Estimation (Direct Measurements)
Moving Car Methods (Traditional) : 25th Aug 2009 and 26th Aug 2009
Licence Plate Matching Techniques: 24th Aug 2009 to 28th Aug 2009 and 12th October 2009 to 16th October 2009
Probe Vehicle techniques: (GPS fitted Vehicle): 31Aug 2010 to 2 Sep 2010 and 23rd Jan 2011 to 27th Jan 2011
Traffic Incident Survey: 24th Aug 2009 to 28th Aug 2009
Traffic Accidents
Road Works
Vehicle Breakdown
Cleaning and Falling objects

16. Travel Time Estimation - Moving Car Method
Delay due to events (seconds)
Time
Travel Time
(min)
Signal
Signal failure
Merging and pedestrian C/s
Day 1
8.30am
7.00 64
Day 2
6.00 24
8.54am
10.00
155
9.00am
9.00 30 90 51
9.45am
11.00
160
10.05
140 13
Moving car data useful estimate average travel time and delay when there is no uncertainty conditions.
The limitation of moving car method is sample size, therefore the travel time estimation by this method are unable to explain travellers experience on the study area.

17. Travel Time Estimation- License Plate Method
Minimum travel time on all the time periods is around 7 minutes.
Travel Time highly varies between the morning peak hour 9.00 am and am.
Explains that the need travel time uncertainty analysis

18. Statistical Measures- SD and Cv
Morning Hours:
SD values are higher (around 7minutes) for the vehicles enter after 9.00 am in the study corridor.
Cv values becomes larger for the vehicles enter after am and before am in the morning peak hour

19. Statistical Measures- SD and cv
Evening Hours
SD value during evening peak hours varies between 5 and 7 minutes.
In the evening hours the Cv varies approximately 0.4 to 0.55 for the entire period.

20. Travel Time Reliability Measures-PT&PTI
Morning hours, before 9.00 am the planning time is around 12 minutes where as between 9.00 am to am is around 25 minutes.
Vehicle enter after 9am has to plan double the travel time to reach the destination in time.
In the evening hours the planning time is approximately 20 minutes for all the time intervals

21. Travel Time Reliability Measures-BT&BTI
Clear peak values in BT and BTI was observed between am and am.
Highest BI value i.e. 1.0 observed during morning period between 9.00 am and 9.15am.
Travellers should budget an additional 12 minutes buffer for 25 minutes planning to ensure 90% on time arrival at the destination.

22. Statistical Summary of Travel Time
(Time units in minutes)
Just before peak
(830 am to 900am)
(Sample Size 94)
Peak Period
(900 am to 1000am)
(Sample Size 217)
Just After Peak Period
(10am-1030am)
(Sample Size 56)
Minimum Travel Time 6
Mean Travel Time
8.63
13.13
12.13
Median Travel Time 7 10 11
Maximum Travel Time 29 37 33
Variance
14.73
50.40
20.51
Std. Deviation
3.84
7.10
4.53
Coef. of Variation
0.45
0.54
0.37

23. Travel Time Distribution
Just before morning peak period
830 am to 900 am
Just after peak
10.00 am to 1030 am
Morning peak period
9.00 am to am
Goodness-of-fit summary for various distributions
Sl.No
Type of Distribution
Just before peak period
830am to 900 am
During peak Period
900 am to 1000 am
Just after peak period
1000 am to 1030 am
Anderson-Darling test 1
Burr
4.346
2.898
0.713 2
Gen. Extreme Value (GEV)
5.528
4.541
1.256 3
Log-Logistic 
4.291
4.146
0.757 4
Lognormal 
4.555
4.872
1.304 5
Gamma
11.322
14.620
3.199 6
Normal
13.809
22.711
4.403
From AD test it can be identified that Log-Logistic is the best fit for just before peak period and Burr distribution is best fit for peak period and just after peak period.

24. Operational Efficiency of Urban Corridor on NH2
Time Period
Sample Size
Statistical Measures
Reliability Measures
Congestion Measure
Std
Cov
PT (95%) PT
(90%)
PTI BT
BTI
TTI
Operational Efficiency of NH2 (From Km 8 to Km 12) - Urban Corridor 43
2.92
0.36
16.1
11.9
2.0
3.7
45%
1.37 51
4.46
0.50
17.6
12.8
2.1
3.9
43%
1.54 61
7.63
0.64
32.1
24.4
4.1
12.6
106%
1.85 76
6.90
0.53
26.7
24.9
4.2
11.8
90%
1.70 45
6.16
0.44
27.0
10.8
77%
1.80 35
7.59
0.54
28.1
26.3
4.4
12.2
87%
1.93 28
6.01
0.47
19.0
3.2
6.3
49%
1.90
2.31
0.20
13.9
13.8
2.3
2.4
21%
1.89
95th % travel time for urban corridor is varies between 3.45 to 8 min per km during morning peak hour.
This value for rural corridor varies between 1.5 to 1.74 min per km in rural section on NH2.

25. Reliability Measures Vs Congestion Measures
Severity order of congestion represented by the congestion index (TTI) may not be consistently followed with that of reliability index.
Travel time reliability measures are capable of measuring the variability in congestion level.

26. Speed Profile during Morning Hours (Survey Period : 31st Aug 2010 to 2nd Sep 2010)
Speed Profile of the Urban Study Corridor at 915 am
Speed Profile of the Urban Study Corridor at 830 am
Mata Mandir Jn
Grand lay
Okhla mode
Okhla Tank
Apollo
CRRI
Mata mandir Jn
Grand lay
Okhla mode
CRRI
Okhla Tank
Apollo
Speed Profile of the Urban Study Corridor at am
Mata Mandir Jn
Grand lay
Okhla mode
CRRI
Okhla Tank
Apollo

27. Speed and Flow profile on working days and non working day at 8.15 am
Speed and Fuel Flow Profile using Probe Vehicle (survey Period: 23rd Jan 2011 to 27th Jan 2011)
Speed and Flow profile on working days and non working day at 8.15 am
The travel time on working as well as non working day for the direction of Ashram to Apollo and the opposite direction at 8.15am is varies between 7.3 minutes to 7.7 minutes.
The average fuel consumption for each direction is about 390 ml.(diesel)

28. Speed and Fuel Flow Profile using Probe Vehicle
Speed and Flow profile on working days and non working day at 9.10 am
The travel time on non working day for the direction of Ashram to Apollo and the opposite direction at 9.10am is about 6.5 min and 8.5 min and the fuel consumption is about 380ml and 390 ml respectively
The travel time on working day as for the direction of Ashram to Apollo and the opposite direction at 9.10am is about 11.2min and 14min and the fuel consumption is about 435 ml and 477ml respectively

29. Speed and Fuel Flow Profile using Probe Vehicle
Speed and Flow profile on working days and non working day at 10.00am
The travel time on non working for the direction of Ashram to Apollo and the opposite direction at 10.00am is about 7.0 min and 9.4 min and the fuel consumption is about 355ml and 410 ml respectively
The travel time on working as for the direction of Ashram to Apollo and the opposite direction at 10.00am is about 13.0 min and 20 min and the fuel consumption is about 465 ml and 570 ml respectively

30. Modeling Travel Time Variability
Traffic Flow
Traffic Flow

31. Modeling Travel Time Variation
Variation of travel time during the study period is mainly due to traffic flow variation, delay due to traffic signal and pedestrians interruption to main traffic at un-signalized pedestrian crossing near Okhla Mode Bus Stop.
Time Period
Average delay due to Traffic Signals (sec) 44
167
212
183
about 146 pedestrians were crossing in every 15 minutes interval at this location during morning hours.
on an average about 3minutes delay in 15minutes interval of time

32. Modeling Approach- Stochastic Simulation
Stochastic Response Surface Method (SRSM)
Y = a0 + a1ξ1 + a2ξ2 + a3(ξ12-1) + a4(ξ22-1) + a5ξ1ξ2

33. Table Uncertainty ranges of model parameter and response variable
SRSM Analysis
Continuous random variable such as Traffic Volume and Pedestrian Volume was considered for SRSM to model travel time distribution.
Table Uncertainty ranges of model parameter and response variable
Parameter
Traffic volume (PCU/15min)
Pedestrian Volume (persons/15min)
Travel time (seconds)
Minimum value
748 83
480
Maximum value
930
206
1080
Average Value
865
140
729
Standard deviation 49 32
135.6
Distribution type
Lognormal
Distribution parameter

34. SRSM Analysis Contd..
Lognormal Distribution for Traffic Volume
Lognormal Distribution for Pedestrian Volume
Lognormal Distribution Parameters and 0.058
KS Statistic   , AD Statistic   
Lognormal Distribution Parameters and 0.236
KS Statistic   , AD Statistic
To check validity of the assumption Goodness of fit test was done by considering Kolmogorov-Smirnov (KS) test and Anderson-Darling (AD) test.
Results indicates that no difference between empirical and theoretical cumulative distributions.
Lognormal distribution was considered for generating the random data for traffic volume and pedestrian volume respectively in SRSM model.

35. Modeling Travel Time Variation
Model Coefficients Estimated by SRSM a0 a1 a2 a3 a4 a5
Coefficie nts
754.83
75.33
7.41
2.22
0.722
0.1

36. Provision of sustainable Infrastructure
Main contribution factor for travel time uncertainty during study period is demand side variation i. e. about 70% and remaining is due to capacity side variation.
The goal/ target for provision of sustainable infrastructure is to reduce traffic volume impact on the study area by providing suitable infrastructure facilities.
Following are the provisions of infrastructure to improve travel time uncertainty which provides the sustainable system for the corridor

37. Different options
Option 1: Providing following Provisions
Provision 1: Provision of subway at two locations (Modi mill flyover and Okhla Tank)
Provision 2 : Provision of underpass at Grandly Junction and Provision of signal closing at Mata Mandir.
Provision 3 : Provision of additional Lane from on ramp to ashram as space is available (lane extension)
Option 2 : Introducing New Mode (at grade )in the Median ( From ….. To Apolo Metro Station)

38. Provision of Subway at Okhla Mode
Providing Subway at Okhla mode the average speed at this section( 200mts) will improve from 10 kmph to 35 kmph. This leads to saving of 30 sec to 60 each of each passenger during morning and evening peak hours.
This also saves the fuel consumption approximately 700 lts per hour. ( approximately saves 10.5 crors per year )
Travel time reliability of the corridor will also be improved ( to be estimated)

39. Provision of Underpass
Providing underpass (facility providing for NMT) at Okhla Tank and closing pedestrian signal will improve the average speed at this section (200 mts) from 20 kmph to 40 kmph. This will leads to saving of an average of 30 sec of each passenger during morning and evening peak hours.
This also saves the fuel consumption approximately 280 lts per hour. ( approximately saves 4.5 crors per year )
Travel time reliability of the corridor will also be improved ( to be estimated)
Further banning of U turn at CRRI Signal will be improved the aveage section speed from 10 kmph to 30 kmph

40. Provision of Slopped Subway
Providing underpass for straight traffic from Badrpur to ashram will improve the average speed at this section (200 mts) from 10 kmph to 40 kmph. This will leads to saving of an average of 60 sec of each passenger during morning and evening peak hours.
This also saves the fuel consumption approximately 480 lts per hour ( approximately saves 24 crors per year )
Travel time reliability of the corridor will also be improved ( to be estimated)
Further banning of U turn at CRRI Signal will be improved the aveage section speed from 10 kmph to 30 kmph

41. Provision of additional lane contd..
Provision 1: Provision of Under passes at locations (Modi mill and Okhla Tank)
Provision 2 : Provision of Subway at Grandly Junction and Provision of signal closing at Mata Mandir.
Provision 3 : Provision of additional Lane, ( starting from on ramp to ashram free left turn ,Space is available)

42. Guidance Provided to various Students
Related Publications
Dr.Ch.Ravi Sekhar, Dr. B.K Durai, A.Mohan Rao and K.Sitaramanjaneyulu “Toll System Design based on Travel Time Reliability for High Speed Corridors” , IRC PPP seminar, Indian Habitate Center, New Delhi, 28th &29th Aug 2009
Dr.S.Gangopadyay, Dr. Ch.Ravi Sekhar and Dr.B.Kanaga Durai “ Travel Time Reliability Measurement for an urban Corridor-Acase Study”, Indian Highways , May 2010.
Dr.Ch.Ravi Sekhar, Dr.B.Kanaga Durai and Dr.S. Gangopadhyay” Modeling Travel Time Variation under Uncertainties: A Case Study of Urban Corridor in Delhi, Conference on Infrastructure, Sustainable Transportation and Urban Planning 2010) October 18-20, 2010 (CD ROM)
Dr.Ch.Ravi Sekhar, Dr.B.Kanaga Durai and Dr.S. Gangopadhyay “Examining Travel Time Distribution of Urban and Rural Corridor of National Highway in India”, Easter Asia Socity for Tansportation Studies 2011, submited on 7th October 2010 for Journal section.
Ms Kamini Gupta, Dr.Ch.Ravi Sekhar, Dr.Ravider Kumar and Dr.B.Kangadurai “ Road Transport Infrastructures provisions for Sustainabile Transportation System” Abstract Accepted for Women conference
Guidance Provided to various Students
In-Plant Summer Training (Two M.E Students )
Master Thesis (Association with Anna University - On going )
Extension of this Work
Proposal selected for CSIR EMPOWER Project.

43. Thank You