CALIFORNIA CENTER FOR INNOVATIVE TRANSPORTATION
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
I- 880 Corridor Management Plan
Demonstration
Thomas West
Director, California Center for Innovative Transportation
CCIT Research Report
UCB- ITS- CWP- 2010- 1
ISSN: 1557- 2269
The California Center for Innovative Transportation works with
researchers, practitioners, and industry to implement transportation
research and innovation, including products and services that improve
the efficiency, safety, and security of the transportation system.
CALIFORNIA CENTER FOR INNOVATIVE TRANSPORTATION
INSTITUTE OF TRANSPORTATION STUDIES
UNIVERSITY OF CALIFORNIA, BERKELEY
I- 880 Corridor Management Plan Demonstration
Thomas West, Director, CCIT
CCIT Research Report
UCB- ITS- CWP- 2010- 1
This work was performed by the California Center for Innovative Transportation,
a research group at the University of California, Berkeley, in cooperation with the
State of California Business, Transportation, and Housing Agency’s Department
of Transportation, and the United States Department of Transportation’s Federal
Highway Administration.
The contents of this report reflect the views of the authors, who are responsible
for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official views or policies of the State of California. This
report does not constitute a standard, specification, or regulation.
January 2010
I- 880 CORRIDOR MANAGEMENT
PLAN DEMONSTRATION
Task Order 1015 - Final Report
January 2010
California Center for Innovative Transportation
System Metrics Group, Inc.
Braidwood Associates
CALIFORNIA CENTER FOR INNOVATIVE TRANSPORTATION
UNIVERSITY OF CALIFORNIA BERKELEY
2105 BANCROFT WAY, SUITE 300 · BERKELEY, CA 94720- 3830
PHONE: ( 510) 642- 4522 · FAX: ( 510) 642- 0910 · HTTP:// WWW. CALCCIT. ORG
Project Fact Sheet
Title: I- 880 Corridor Management Plan Demonstration
Sponsor: Caltrans Division of Research and Innovation
Executing organization: California Center for Innovative Transportation
2105 Bancroft Way, Berkeley, CA 94720
Phone: ( 510) 642- 4522. Fax: ( 510) 642- 0910
Execution period: 6/ 1/ 2006— 9/ 30/ 2009
Contract amount: $ 855,458
Principal Investigator: Hamed Benouar, PhD
Center Director: Thomas West
Project Manager: Thomas West
Dedication
We dedicate this Corridor Management Plan Demonstration to the memory of Patricia “ Pat”
Weston ( 1951- 2009), Chief, Caltrans Office of Advanced System Planning, whose seemingly
limitless energy and passion for transportation system planning in California has been an
inspiration to countless transportation planners and engineers within Caltrans and its partner
agencies. Pat’s efforts elevated the importance of corridor- based system planning, performance
measurement for system monitoring, and the blending of long- range planning with near- term
operational strategies. This has resulted in stronger planning partnerships with Traffic Operations
in Caltrans and led directly to the requirement to conduct comprehensive corridor planning through
Corridor System Management Plan ( CSMP) documents. This is but one of a long list of major
achievements in Pat’s lengthy Caltrans career. She generously shared her knowledge, wisdom,
and guidance with us over the years. She will be sorely missed as a planner, mentor, and friend.
Executive Summary
It is clear that transportation infrastructure expansion will continue to fall behind the pace of
demand. If conditions are to improve, or at least not deteriorate as fast, a new approach to
transportation decision making and investing is needed. The Corridor system Management
Plan for the Nimitz ( I- 880) Freeway corridor in the Bay Area is a “ first cut” template that
integrates the overall concept of system management into Caltrans’ planning and decision-making
process.
System Management is the wave of the future and is being touted at the federal, state,
regional and local levels. Understanding how a corridor performs and why it performs the way
it does is critical to crafting the appropriate strategies. From the research, it is found that
congestion leads to lost productivity in the form of bottlenecks. Expanding existing
infrastructure, however, is not always the best route to go, especially in today’s economic
climate. The system management philosophy begins by defining how the system is performing,
understanding why it is performing that way, and then evaluating different strategies to address
deficiencies.
In 2004, under sponsorship from the California Department of Transportation, the California
Center for Innovative Transportation ( CCIT) at the University of California, Berkeley began the
process to evaluate the performance of a heavily congested major urban transportation
corridor in the San Francisco Bay Area and to model and assess the benefits of a variety of
transportation investments upon the corridor. Systems Metrics Group ( SMG), a subcontractor
to CCIT and responsible party to conduct the overall evaluation, modeling, and investment
review has returned with a comprehensive and scientifically justifiable assessment of
Interstate 880, the selected corridor with boundaries that include the SR- 237 interchange in
Fremont to the Grand Avenue Interchange in Oakland. Through extensive performance
monitoring, SMG was able to conduct and document a comprehensive performance
assessment of the corridor and through the use of sophisticated microscopic traffic simulation
modeling tools and techniques, to evaluate the validity of a variety of investment scenarios.
While not intended to replace other studies, this analysis represents the first attempt by the
California Department of Transportation to address existing travel conditions and mobility
challenges though the integration of operational analyses, traditional planning management
strategies, and capital improvements all based upon a strong and scientific assessment of
existing conditions and potential scenarios. In summary, results of this study produced a
return- on- investment ranking for a variety of improvement opportunities for the Interstate 880
corridor, primarily located in bottle- neck related problem areas. In addition, the study identified
advanced ramp metering as highest performing investment included in the study and
proposes, among other recommendation, that Caltrans and its partners focus on a properly
implemented advanced ramp metering systems along the Interstate 880 corridor.
I- 880 Corridor Management Plan Demonstration
Page i of 88
System Metrics Group, Inc.
Table of Contents
Table of Contents ............................................................................................................. i
List of Exhibits .................................................................................................................. ii
I. INTRODUCTION .......................................................................................................... 4
What is a Corridor Management Plan? .................................................................... 5
What is System Management? ................................................................................ 5
What is Productivity? ................................................................................................ 7
Study Approach...................................................................................................... 10
Document Organization ......................................................................................... 11
II. SCENARIO DEVELOPMENT FRAMEWORK ........................................................... 12
Scenario Development Process ............................................................................. 13
III. SCENARIO RESULTS ............................................................................................. 17
Scenario Analysis Approach .................................................................................. 17
Model Output Summaries ...................................................................................... 18
IV. POST MODEL ANALYSIS ....................................................................................... 23
Scenario Costs ....................................................................................................... 24
Scenario Benefits ................................................................................................... 24
Scenario Benefit Cost Ratios ................................................................................. 26
V. CONCLUSIONS........................................................................................................ 27
A1. CORRIDOR DESCRIPTION ................................................................................... 30
Freeway ................................................................................................................. 30
Transit .................................................................................................................... 31
Intermodal Facilities ............................................................................................... 33
Special Event Facilities .......................................................................................... 36
Land Use ................................................................................................................ 36
Government Lands ................................................................................................. 37
Parks and Recreational Areas................................................................................ 38
Schools .................................................................................................................. 39
Hazardous Material Sites ....................................................................................... 40
Wetlands ................................................................................................................ 41
A2. COMPREHENSIVE PERFORMANCE ASSESSMENT .......................................... 43
Corridor- wide Performance Measures and Trends ................................................ 44
Corridor- Wide Mobility Results - Delay ............................................................... 45
Corridor- Wide Mobility Results – Travel Time and Reliability of Travel Time ..... 52
Corridor- Wide Productivity Results ........................................................................ 56
Corridor- Wide Safety Results ................................................................................. 57
Bottleneck Analysis ................................................................................................ 61
Northbound Bottlenecks ..................................................................................... 63
Southbound Bottlenecks ..................................................................................... 71
Bottleneck Areas .................................................................................................... 81
A3. SCENARIO DRAWINGS ........................................................................................ 82
I- 880 Corridor Management Plan Demonstration
Page ii of 88
System Metrics Group, Inc.
List of Exhibits
Exhibit I- 1: 880 Corridor Study Boundaries and Detection Stations ______________________ 4
Exhibit I- 2: California Freeway Traffic Congestion Growth Last 20 Years __________________ 5
Exhibit I- 3: System Management Pyramid __________________________________________ 6
Exhibit I- 4: Productivity Loss during Severe Congestion _______________________________ 8
Exhibit I- 5: 2003- 2007 Lost Productivity in District 4 __________________________________ 9
Exhibit I- 6: Study Approach ____________________________________________________ 10
Exhibit II- 1: Fully Funded Near Term Corridor Projects _______________________________ 13
Exhibit III- 1: Example Model Output _____________________________________________ 17
Exhibit III- 2: 2006 Base Year Model Delay Scenario Results __________________________ 18
Exhibit III- 3: Peak Periods Percent Delay Reductions Compared to 2006 Base Year________ 19
Exhibit III- 4: Percent Delay Reductions by Direction Compared to 2006 Base Year _________ 19
Exhibit III- 5: Base Year 2006 and Do Minimum 2020 Horizon Year Corridor Delays ________ 20
Exhibit III- 6: 2020 Horizon Model Scenario Results __________________________________ 21
Exhibit III- 7: Percent Delay Reductions Compared to 2020 Do Minimum Scenario _________ 21
Exhibit IV- 1: Scenario Component Costs as Provided ( in mil. $) ________________________ 24
Exhibit IV- 2: Scenario Costs Summary ( in 2007 mil. $) _______________________________ 24
Exhibit IV- 3: Percent Delay Reductions Compared to 2020 Do Minimum Scenario _________ 25
Exhibit IV- 4: Aggregated GHG Emission Benefits by Scenario _________________________ 25
Exhibit IV- 5: Aggregated GHG Emission Reductions by Scenario ______________________ 26
Exhibit IV- 6: Benefit Cost Ratios for Scenario Components ___________________________ 26
Exhibit V- 1: Summary of Planned and Recommended Projects Related to Bottlenecks ______ 28
Exhibit A1- 1: Bay Area Rapid Transit Map ________________________________________ 31
Exhibit A1- 2: Port of Oakland Aerial _____________________________________________ 34
Exhibit A1- 3: Oakland Airport Passenger Volume Trends _____________________________ 35
Exhibit A1- 4: Oakland Airport Cargo Volume Trends ________________________________ 35
Exhibit A1- 5: McAfee Coliseum and Adjacent Sports Arena Aerial ______________________ 36
Exhibit A1- 6: Government- Owned Land __________________________________________ 38
Exhibit A1- 7: Recreational Areas ________________________________________________ 39
Exhibit A1- 8: Educational Facilities ______________________________________________ 40
Exhibit A1- 9: Primary Hazardous Waste Sites ______________________________________ 41
Exhibit A1- 10: Wetland Locations _______________________________________________ 42
Exhibit A2- 1: PeMS Connectivity to TMCs and Example Screens_______________________ 44
Exhibit A2- 2: I- 880 Study Area Average Daily Delay by Time Period ____________________ 45
Exhibit A2- 3: Northbound Average Daily Delay by Time Period ________________________ 46
Exhibit A2- 4: Southbound Average Daily Delay by Time Period ________________________ 47
Exhibit A2- 5: Northbound Average Monthly Daily Delay by Time Period__________________ 48
Exhibit A2- 6: Southbound Average Monthly Daily Delay by Time Period _________________ 48
Exhibit A2- 7: Average Severe Congestion by Day of Week ___________________________ 50
Exhibit A2- 8: Average Non- Severe, Other Congestion by Day of Week __________________ 50
Exhibit A2- 9: Average Northbound Weekday Hourly Delay ____________________________ 51
Exhibit A2- 10: Average Southbound Weekday Hourly Delay __________________________ 51
Exhibit A2- 11: Average Northbound Travel Times by Hour ____________________________ 53
Exhibit A2- 12: Average Southbound Travel Times by Hour ___________________________ 53
Exhibit A2- 13, A2- 14, A2- 15: Northbound Travel Time Variability _______________________ 55
Exhibit A2- 15, A2- 16, A2- 17: Southbound Travel Time Variability ______________________ 55
Exhibit A2- 18: Average Northbound Lost Lane Miles ________________________________ 56
Exhibit A2- 19: Average Southbound Lost Lane Miles ________________________________ 56
Exhibit A2- 20: Daily CHP Incidents Reported ______________________________________ 57
I- 880 Corridor Management Plan Demonstration
Page iii of 88
System Metrics Group, Inc.
Exhibit A2- 21: Week Total CHP Incidents Likely to Impact Congestion __________________ 58
Exhibit A2- 22: Daily CHP Collisions Reported ______________________________________ 59
Exhibit A2- 24: AM Congestion Breakdown Estimates by Cause in 2006 _________________ 60
Exhibit A2- 25: PM Congestion Breakdown Estimates by Cause in 2006 _________________ 60
Exhibit A2- 26: 2006 HICOMP Report Maps ________________________________________ 62
Exhibit A2- 27 – Example Speed Contour Plot ______________________________________ 63
Exhibit A2- 28, A2- 29, A2- 30: Northbound Speed Contour Plots ________________________ 64
Exhibits A2- 31, A2- 32, A2- 33, A2- 34: Northbound AM Peak Speed Contour Plots _________ 66
Exhibit A2- 35: Tennyson Merge Bottleneck ________________________________________ 67
Exhibit A2- 36: Traffic Clearing after Tennyson On- ramps _____________________________ 68
Exhibits A2- 41, A2- 42, A2- 43: Southbound Speed Contour Plots _______________________ 72
Exhibit A2- 44: Five- Minute Speeds by Detector for January, 2006 ______________________ 73
Exhibit A2- 45: Speed Contour Plot for January, 2006 ________________________________ 73
Exhibit A2- 46: 98th On- Ramp Bottleneck __________________________________________ 74
Exhibit A2- 47: On- Ramp from Eastbound 98th ______________________________________ 74
Exhibit A2- 48: 29th Off- Ramp Bottleneck __________________________________________ 75
Exhibit A2- 49: 5- Minute Speeds for Oak On- Ramp Bottleneck _________________________ 76
Exhibit A2- 50: Oak On- Ramp Bottleneck __________________________________________ 76
Exhibit A2- 51 – Smaller Bottlenecks ( Southbound AM) _______________________________ 77
Exhibit A2- 52 – Winton On- Ramp Occasional Bottleneck _____________________________ 77
Exhibit A2- 53: 5- Minute Speeds by Detector SB at/ near Mission Blvd - October 2004 _______ 79
Exhibit A2- 54: 5- Minute Speeds by Detector SB at/ near Mission Blvd - March 2005 ________ 79
Exhibit A2- 55: 5- Minute Speeds by Detector SB at/ near Mission Blvd - October 2005 _______ 79
Exhibit A2- 56: 5- Minute Speeds by Detector SB at/ near Mission Blvd – March 2006 ________ 79
Exhibit A2- 57 – Southbound Bottleneck at the Fremont Interchange ____________________ 80
Exhibit A2- 58: Southbound Bottleneck at the Mission/ Rte 262 Interchange _______________ 80
Exhibit A2- 59: Dividing Corridors into Bottleneck Areas ______________________________ 81
Exhibit A3- 1: Scenario 7A Changes Coded at 23rd Avenue ___________________________ 82
Exhibit A3- 2: Scenario 7A Changes Coded at 29th Avenue ___________________________ 83
Exhibit A3- 3: Scenario 8A Changes Coded - Aux Lanes Paseo Grande to Winton Avenue ___ 83
Exhibit A3- 4: Scenario 8A Changes Coded - Aux Lanes Whipple Road to Ind Pkwy West____ 84
Exhibit A3- 5: Scenario 8A Changes Coded - I- 880/ Whipple Road Interchange ____________ 84
Exhibit A3- 6: Scenario 8A Changes Coded - I- 880/ West A Street Interchange ____________ 85
Exhibit A3- 7: Scenario 8A Changes Coded - I- 880/ West Winton Avenue Interchange _______ 85
Exhibit A3- 8: Scenario 9A Changes Coded - Hegenberger to 98th Avenue Interchange _____ 86
Exhibit A3- 9: Scenario 9A Changes Coded - 98th Avenue Interchange __________________ 86
Exhibit A3- 10: Scenario 9A Changes Coded - 98th Avenue to Davis Street Interchange _____ 86
Exhibit A3- 11: Scenario 9A Changes Coded - 98th Avenue to Davis Street Interchange ____ 87
Exhibit A3- 12: Scenario 9A Changes Coded - Davis Street Interchange _________________ 87
Exhibit A3- 13: Scenario 9A Changes Coded - Davis Street to Marina Blvd Interchange _____ 87
Exhibit A3- 14: Scenario 9A Changes Coded - Marina Blvd Interchange __________________ 88
Exhibit A3- 15: Scenario 9A Changes Coded - Marina Blvd to Continuation of HOV Lane ____ 88
I- 880 Corridor Management Plan Demonstration
Page 4 of 88
Corridor Management Plan Page - 4 System Metrics Group, Inc.
I. INTRODUCTION
This document presents the Model Corridor Management Plan for the Nimitz ( I- 880)
Freeway corridor in the Bay Area from the SR- 237 Interchange in Fremont to the Grand
Avenue Interchange in Oakland. This project was intended to demonstrate the concept
of corridor management, including conducting and documenting the comprehensive
performance assessment and evaluating improvements. This was done by using
advanced micro simulation tools to duplicate corridor performance conditions
documenting and projecting the benefits of different improvement strategies on traffic
flow and overall mobility on the corridor. The project was not intended to replicate or
replace other studies or previous decisions.
This plan represents the first attempt by the California Department of Transportation
( Caltrans) to develop a phased strategy that integrates operational analysis with more
traditional system planning based on a foundation of comprehensive performance
assessment and evaluation. The corridor was selected by Caltrans District 4 ( Bay Area)
and its stakeholders, partly based on the availability of detection data needed for the
critical performance assessment efforts. Exhibit I- 1 below shows the corridor
boundaries ( identified by the arrows) and its detection stations.
Exhibit I- 1: 880 Corridor Study Boundaries and Detection Stations
• Southbound Detectors
• Northbound Detectors
I- 880 Model Corridor System Management Plan
I- 880 Corridor Management Plan Demonstration
Page 5 of 88
Corridor Management Plan Page - 5 System Metrics Group, Inc.
What is a Corridor Management Plan?
A Corridor Management Plan is a document that identifies the recommended system
management strategies for a given State Highway System facility based on
comprehensive performance assessment and evaluation. The strategies are phased
and include both operational and more traditional longer range capital expansion
strategies. The strategies take into account transit usage and projections and
interactions with the arterial network. As such, this corridor management plan serves as
a “ first cut” template that integrates the overall concept of system management into
Caltrans’ planning and decision- making processes. Moving away from the traditional
approach that often focuses on expensive capital improvements to localized freeway
problem areas, this project follows a corridor management plan approach, which
emphasizes performance assessments and operational strategies that yield higher
benefit to cost results.
What is System Management?
With the rising cost and complexity of construction and right of way acquisition, the era
of building new facilities is coming to an end. From 1998 through 2007, California, like
so many other states, expanded its freeway transportation infrastructure by less than
one half percent annually. However, demand for transportation during the same period,
as measured by freeway vehicle miles traveled, rose by an average of 2.5 percent,
which is five times the rate of infrastructure growth. As indicated in Exhibit I- 2,
congestion continues to generally increase at a rate higher than demand except during
periods of economic stagnation
Exhibit I- 2: California Freeway Traffic Congestion Growth Last 20 Years
-
50
100
150
200
250
300
350
400
450
500
550
600
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
YEAR
Congestion ( Avg Daily Veh- Hrs of Delay in 1000s)
Veh- Miles Traveled ( VMT in Billions)
-
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
60.0
Directional Miles ( in 100s)
Statewide Population ( in millions)
Daily Vehicle- Hours of Delay ( in 1,000s)
Caltrans State Highway System VMT ( billions)
Congested Directional Miles ( in 100s)
Total Directional Urban Freeway Miles ( in 100s)
Statewide Population ( millions)
I- 880 Corridor Management Plan Demonstration
Page 6 of 88
Corridor Management Plan Page - 6 System Metrics Group, Inc.
It is clear that infrastructure expansion will continue to fall behind the pace of demand.
If conditions are to improve, or at least not deteriorate as fast, a new approach to
transportation decision making and investing is needed.
Caltrans recognized this emerging need as it adopted a “ One Vision/ One Mission”
statement to improve mobility across California. It specifies a revised set of goals to
help guide the State towards that new approach: productivity, reliability, flexibility,
safety, and performance. The first three goals are new and call for improving the
efficiency of the transportation system, reducing traveler delays due to incidents and
road work, and making transit a more practical travel option. The last two goals are
traditional but critical, ensuring the public’s safety and delivering the projects efficiently.
System Management ( SM) is the wave of the future and is being touted at the federal,
state, regional and local levels. The SM “ pyramid” shown in Exhibit I- 3 illustrates how
we need to address both transportation demand and supply to maximize system
performance. In the end, it is critical that the productivity of our system increases to
make up with the past and likely future difference ( deficiency) between supply and
demand increases.
Exhibit I- 3: System Management Pyramid
I- 880 Corridor Management Plan Demonstration
Page 7 of 88
Corridor Management Plan Page - 7 System Metrics Group, Inc.
Ideally, Caltrans and its regional partners would develop a regional system
management plan that addresses all components of the SM pyramid for an entire region
comprehensively. However, because SM is new to Caltrans and its regional and local
partners, it is prudent to practice SM at the corridor level first.
The foundation of system management is system monitoring and evaluation ( the base
of the pyramid) through comprehensive performance assessment and evaluation.
Understanding how a corridor performs and why it performs the way it does is critical to
crafting the appropriate strategies. Two entire sections of the appendix to this
document ( Sections A- 2 and A- 3) are dedicated to performance assessment. A
relatively new, sometimes controversial measure merits a discussion here since it
explains the increased emphasis on operational strategies. This measure is
productivity.
What is Productivity?
A critical goal of System Management is to “ get the most out” of the existing system, or
maximize system productivity. One would think that a given freeway is most productive
during peak commute times. This is true for freeways not experiencing congestion.
However, for California’s urban freeways which have been experiencing growing
congestion, the opposite is true.
Exhibit I- 4 illustrates how congestion leads to lost productivity. The exhibit represents
speeds in red and flow rates in blue on one section of the 405 freeway in Los Angeles.
It shows that once severe congestion starts ( at around 2 pm) and speeds dip to 20
miles an hour, flow rates ( the number of vehicles passing through the segment per
hour) dip to below 750 per lane per hour. Given that design capacities for freeways are
around 2,000 vehicles per hour per lane, actual flow rates during the congested period
can represent a loss of more than 50 percent of this capacity ( i. e., 750 actual flow rates
versus 2,000 design capacity). This loss, shown as the shaded area in the exhibit, is
referred to as lost productivity and can be presented in terms of “ Lost Lane Miles”.
The cause of lost productivity can almost always be linked to bottlenecks ( or pinch
points). These bottlenecks sometimes occur on a regular basis ( e. g., at certain
interchanges) and sometimes occur as a result of special circumstances ( e. g.,
incidents).
In both cases though, bottlenecks occur when the overall demand at a particular
location exceeds the effective capacity of that location. In this case, demand refers to
vehicular demand that is actually either on the freeways or is allowed on the freeways
( e. g., from on- ramps). It does refer to the total number of vehicles who want to get on
the freeway, but may still be on the ramps or on the arterials. Conversely, effective
capacity refers to the maximum throughput ( e. g., number of vehicles per hour per lane)
that can be sustained at a certain location.
I- 880 Corridor Management Plan Demonstration
Page 8 of 88
Corridor Management Plan Page - 8 System Metrics Group, Inc.
When demand exceeds the effective capacity, unstable traffic flow occurs and any
additional merging and weaving lead to queues building behind the bottleneck. The
flow rates are generally lower in bottleneck queues. This in turn leads to productivity
losses. To the extent that operational strategies can be implemented to eliminate the
bottleneck altogether or to reduce the severity of the bottlenecks and the queues,
productivity can be increased without major facility expansion.
Exhibit I- 4: Productivity Loss during Severe Congestion
-
10
20
30
40
50
60
70
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
TIME
SPEED
( 50)
200
450
700
950
1,200
1,450
1,700
1,950
2,200
FLOW RATE ( VPHPL)
Speed < 35 mph
Speed < 35mph
As speeds drop, flow rates drop significantly
-
10
20
30
40
50
60
70
2: 00 PM
3: 00 PM
4: 00 PM
5: 00 PM
6: 00 PM
7: 00 PM
8: 00 PM
TIME
SPEED
( 50)
200
450
700
950
1,200
1,450
1,700
1,950
2,200
FLOW RATE ( VPHPL)
Speed < 35 mph
SR- 99 NB
Sacramento County
Turnbridge Drive
PM= 20.18
October 17, 2006
VDS ID#: 312513
Speed < 35mph
As shown in Exhibit I- 5, the lost productivity aggregated for District 4 was estimated to
be equivalent to exceed 100 lane- miles during the afternoon peak commute periods in
2007. Total lost productivity for the district in 2007 ( i. e., adding up lost lane miles for all
time periods) added up to almost 200 lane- miles. Therefore, just when the region
needed the most capacity, its freeways performed in a less productive manner.
I- 880 Corridor Management Plan Demonstration
Page 9 of 88
Corridor Management Plan Page - 9 System Metrics Group, Inc.
Exhibit I- 5: 2003- 2007 Lost Productivity in District 4
34
42
54
63
58
7
17
29
34
29
47
61
88
107 100
11 11
17
13 11
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
Equivalent Lost Lane Miles
AM Midday PM Early AM/ Night
2003
2004
2005
2006
2007
Losing 100 lane- miles in the afternoon peak periods effectively means that previous
investments in the region were not fully productive when demand was at its highest.
Clearly, the District and the State aim to leverage these past investments to the extent
possible, which can be done to some extent by implementing targeted operational
strategies.
Infrastructure expansion, although still an important strategy, cannot be the only
strategy for addressing the mobility needs of Californians. System Management is
needed to get the most out of the current system and must be an important
consideration as we evaluate the need for facility expansion investments. Simply
stated, the System Management philosophy begins by defining how the system is
performing, understanding why it is performing that way, and then evaluating different
strategies, including operations centric strategies, to address deficiencies. These
strategies can then be evaluated using different tools to allow for estimation of the
benefits and an evaluation of whether the benefits are worthy of the associated costs.
I- 880 Corridor Management Plan Demonstration
Page 10 of 88
Corridor Management Plan Page - 10 System Metrics Group, Inc.
Study Approach
The study approach and its steps are shown in Exhibit I- 6 and include the important
data sources or tools used for each task ( data needs and sources are discussed in the
appendix section). Note that the base performance assessment relied on the
Performance Measurement System ( PeMS) developed by Caltrans and the Traffic
Accident and Surveillance Analysis System ( TASAS), also developed and maintained
by Caltrans. These systems are invaluable for mobility, reliability, productivity, and
safety analyses. Also note that throughout the study, stakeholders from all jurisdictions
were involved to ensure acceptance of the final recommendations.
Exhibit I- 6: Study Approach
Base Performance Assessment
( PeMS, TASAS, Other)
Bottleneck Identification
( HICOMP, PeMS, Aerial
Photographs, CHP Logs, Other)
Future
Performance
( Micro- Simulation,
Regional Models)
Improvement
Scenarios
Planned,
Programmed and
other
Improvements
Scenario Performance
Evaluation
( Micro Model)
Recommendations
and Performance
Improvement
Estimates
I- 880 Corridor Management Plan Demonstration
Page 11 of 88
Corridor Management Plan Page - 11 System Metrics Group, Inc.
Document Organization
This document focuses on the scenario development and evaluation process.
However, for reference purposes, previous documents and sections thereof are
included as an appendix section. The remainder of this final report is organized as
follows ( Section I is this Introduction):
Section II – discusses the scenario development framework ( i. e., how the
scenarios where developed and why)
Section III – presents the model results of the scenario performance evaluation
process
Section IV – presents the “ post model” evaluation results, which include benefit
cost analysis results as well as Green House Gas ( GHG) emission reduction
estimates
Section V – outlines the conclusions of this study and how these conclusions
may impact ongoing or future corridor management planning efforts.
Appendix A
­Section
A1 – Presents the corridor description section
­Section
A2 – Presents the comprehensive performance assessment,
including corridor- wide performance measures updated through 2007 and the
bottleneck identification and causality findings
­Section
A3 – Presents exhibits with drawings of the different scenarios tested
Also note that there are two additional technical appendices under a separate cover.
The first is the technical model calibration report and the second is the technical
scenario analysis report. Both focus on the modeling aspects of the corridor. Electronic
copies of all models ( base year, horizon year, and scenarios) have been submitted to
Caltrans and can be made available.
I- 880 Corridor Management Plan Demonstration
Page 12 of 88
Corridor Management Plan Page - 12 System Metrics Group, Inc.
II. SCENARIO DEVELOPMENT FRAMEWORK
This section describes the logic behind developing the scenarios that were evaluated
using the microsimulation model. Ideally, one would wish to evaluate each project on its
own and in combination with others. Realistically, that is not possible due to resource
and schedule constraints.
For instance, consider a case where 10 projects are candidates for evaluation. To
evaluate each possible combination, one would need to run the microsimulation model
over 1,000 times. Given the time it takes to run the model and check the results, this is
not currently feasible. As computer power and the ability to streamline such testing
improve, this may become possible. But for now and for the near future, this
comprehensive evaluation approach is not pragmatic.
Therefore, projects have to be combined to the extent possible. This is why the study
focused on developing scenarios that make logical sense. It is also important to note a
couple of important factors upfront:
Scenario testing in this study is different from traditional “ alternatives evaluation”
generally undertaken for Major Investment Studies ( MIS) or Environmental
Impact Reports ( EIRs). The latter types of studies focused on identifying
alternative solutions to addressing current and/ or projected corridor problems.
So each alternative is evaluated separately and results are compared. At the
end, a locally preferred alternative is defined. For this study, scenarios build on
each other ( as detailed later). So a given scenario generally equates to a
previous one plus one or more projects. This difference is important since
corridor management studies are new and are often confused with alternative
studies.
For horizon year 2020, we started with a “ do minimum” model which does not
include any improvements scheduled to be delivered before 2020. This way, we
could evaluate the expected benefits from fully programmed improvements as
part of this study. This is somewhat different from other studies that start off with
a “ baseline” horizon year that includes all projects programmed and to be
completed before the horizon year. These types of studies look for projects over
and beyond the programmed ones. However, we wanted to evaluate
programmed improvements first so we can estimate their benefits and then later
on compare real benefits versus estimates ones.
I- 880 Corridor Management Plan Demonstration
Page 13 of 88
Corridor Management Plan Page - 13 System Metrics Group, Inc.
Scenario Development Process
Developing the first set of scenarios involved several steps. First, a list of programmed
and planned projects was compiled for the corridor. This was an iterative process partly
due to the delays encountered in this study.
Using this list of programmed and planned projects, we identified all projects that were
fully programmed and scheduled to be delivered in the short term ( i. e., by 2012 or
sooner). The reason we distinguished between projects to be delivered by 2012 and
projects to be delivered afterwards is that the first group were candidates to be tested
by both the 2006 Base Year Model and the 2020 Do Minimum Model. This would allow
us to estimate the benefits expected from these projects in the near term as well as the
longer term.
From that list, we then combined those projects related to our performance analysis,
specifically to bottlenecks identified and discussed in the appendix section. Other
projects, such as sound walls, were discarded since microsimulation models cannot
evaluate them. The list of projects and selected ones for testing are shown in Exhibit II-
1 below. Note only three projects met the two criteria ( to be delivered by 2012 and
related to mobility on the corridor). These three projects represented Scenario ( 1A).
Scenario 1A ( 2006) = Base Year 2006 + Mobility Related and Fully Funded
Programmed Projects to be delivered by 2012
Scenario 3AA ( 2020) = No Project Horizon Year - 2020 ( also referred to by the
modeling firm as the Do Minimum Horizon Year 2020)
Scenario 4A ( 2020) = Scenario 3AA + Mobility Related and Fully Funded
Programmed Projects to be delivered by 2012
Exhibit II- 1: Fully Funded Near Term Corridor Projects
Work Description
Capital Cost
( x1000) 2006 2007 2008 2009 2010 2011 2012
ALA 238 Widening $ 85,772.00
880 Seismic Retrofit - 5th Avenue $ 107,840.00
92/ 880 Interchange Reconstruction $ 110,994.00
ALA 880 Oakland High Street Retrofit $ 84,994.00
ALA 580 Seismic Retrofit Phase II Bent $ 1,110.00
ALA 580 MacArthur On- Ramp Partial Widening $ 9,742.00
ALA 880 Interchange Improvement $ 2,583.00
ALA 880 Structure Rehabilitation $ 8,946.00
ALA 880 Route 262/ I- 880 I/ C Construction $ 70,818.00
SCL 880/ 87 at Coleman Avanue $ 59,700.00
ALA 580 Pavement Structure Rehabilitation $ 35,742.00
ALA/ SCA 880 Bridge Widening $ 33,893.00
ALA 880 Improve Median for Relinquishment $ 12,281.00
BART to Airport Connector $ 50,000.00
SC 880 AC Overlay RT 280 $ 4,000.00
ALA 92 Rehabilitation of the Existing Roadway $ 3,000.00
ALA 238 Roadway Rehabilitation $ 19,522.00
I- 880 Corridor Management Plan Demonstration
Page 14 of 88
Corridor Management Plan Page - 14 System Metrics Group, Inc.
Note that all three projects relate to more than the I- 880 Corridor. For instance the
Alameda I- 238 widening improves the I- 880/ I- 238 Interchange, but also improves I- 238
and I- 580. So when evaluating these projects, especially in terms of benefit cost
analysis, the benefits derived from microsimulation will understate total benefits since
they represent the I- 880 Corridor only.
Once scenario 1A was evaluated, the team looked for additional, inexpensive projects
that could be implemented before 2012. The only realistic one was an improvement in
ramp metering. Generally speaking, changes in ramp metering can be implemented
reasonably quickly and inexpensively ( at least compared to other physical
improvements).
First, we tried to make manual adjustments to the ramp metering rates at specified
bottleneck locations. However, the results from the microsimulation analysis showed
increased congestion. Therefore, we discarded this scenario and looked for more
advanced ramp metering as a substitute.
Scenario 2 ( 2006) = Scenario 1A plus Selected Ramp Meter Rate
Adjustments - discarded
Next, we would have liked to test the Systemwide Adaptive Ramp Metering ( SWARM)
algorithm developed by Delcan Corporation and deployed on a test basis in Southern
California. However, an application that emulates the current SWARM algorithm for the
microsimulation model does not exist and the details of the algorithm were not readily
available for the team. Therefore, another algorithm called ALINEA was used. ALINEA,
is a more advanced adaptive ramp metering algorithm that has been deployed on many
freeways internationally. We therefore used the available ALINEA API as a proxy for
more advanced algorithms. ALINEA however, is locally adaptive and therefore its
benefits probably understate the potential of a well calibrated corridor- wide ramp
metering algorithm. This scenario therefore represented scenario 1A plus ALINEA and
was tested for both the base year and the horizon year.
Scenario 3A ( 2006) = Scenario 1A plus ALINEA
Scenario 5A ( 2020) = Scenario 4A plus ALINEA
The next scenario attempted to evaluate improvements in traveler information by 2020
with en- route and pre- route applications that provide the traveler with real time traffic
information. This proved to be very difficult. Microsimulation models sometimes have a
variable called “ familiarity” that attempts to represent how familiar drivers are with
alternative routing. The higher the percent familiarity, the more knowledgeable the
drivers are assumed to be in terms of alternative routing. By increasing the percent
familiarity we could hypothetically simulate improved information provided to the
traveler. However, as will be discussed in the next section, the study model was limited
to the I- 880 Corridor, major interchanges and a limited set of arterials.
I- 880 Corridor Management Plan Demonstration
Page 15 of 88
Corridor Management Plan Page - 15 System Metrics Group, Inc.
As a result, this scenario led to degradation of performance as drivers tried to bypass
one bottleneck only to create another one downstream. Therefore, although the results
are shown in the next section, we believe them to be incorrect. Were the model
significantly more extensive to allow for more re- routing, we believe the results would
have been superior.
Scenario 6A ( 2020) = Scenario 5A + Traveler Information - discarded
Finally, three additional scenarios were tested. These built on priorities defined by the
Alameda County Central Freeway Study ( ACCFS), which all showed incremental
improvement in performance. The three scenarios were defined as follows:
Scenario 7A added to Scenario 5A the recently approved Trade Corridor
Improvement Fund ( TCIF) project. This project will remove and reconstruct the
29th Avenue overcrossing and the two 23rd Avenue overcrossings of I- 880,
which is the major truck route in the Bay Area. Reconstruction of the
overcrossings will provide room to widen the existing I- 880 mainline lanes to the
Caltrans standard width of 12 feet. In addition, the proposed project will widen
the mainline outside shoulders and lengthen existing auxiliary lanes.
Note that our original bottleneck analysis did not identify the 29th Avenue
overcrossing as a major mobility issue. The close proximity of the on- ramps is
the main reason for this bottleneck. Nevertheless, we tested the entire project.
Exhibits A3- 1 and A3- 2 in the appendix section illustrate the changes coded into
the model for this scenario at 23rd and 29th Avenue respectively.
Scenario 8A added to Scenario 7A a number of high priority projects identified
by the ACCFS. These included a number of interchange improvements and
auxiliary lanes as defined by Technical Memorandum: Task 8.2 by the ACCFS.
Exhibits A3- 3 through A3- 7 in the appendix section illustrate the changes coded
into the model for this scenario, including:
­I-
880 Auxiliary Lanes, Paseo Grande to Winton Avenue - This project
would add auxiliary lanes in both the northbound and southbound directions
between Winton Avenue and West A Street by widening the freeway and
reconfiguring the lane layout. A northbound auxiliary lane would be added
between West A Street and Paseo Grande to effectively extend the auxiliary
lane to the south limit of the northbound auxiliary lane portion of the SR- 238
Widening Project.
­I-
880 Auxiliary Lanes, Whipple Road to Industrial Parkway West - This
project would add auxiliary lanes by widening the freeway and reconfiguring
the lane layout to provide the minimum lane widths identified by Caltrans.
This assumes the existing I- 880 bridge over Alameda Creek would be
widened to accommodate the new cross- section.
I- 880 Corridor Management Plan Demonstration
Page 16 of 88
Corridor Management Plan Page - 16 System Metrics Group, Inc.
­I-
880/ Whipple Road Interchange - This project would expand the on ramp
from Whipple Road to I- 880 northbound to provide two lanes, including one
HOV bypass lane. Construction of this project requires expanding the existing
bridge over the Union Pacific Railroad and some right- of- way acquisition.
­I-
880/ West A Street Interchange - This project was defined in concept by
the City of Hayward and would involve widening A Street between the foot- of-ramp
intersections. This required reconstructing the I- 880 overcrossing. This
project would involve intersection and signalization modifications.
­I-
880/ West Winton Avenue Interchange - This project was defined in
concept by City of Hayward and would involve reconstructing ramps to create
a partial cloverleaf with signalized foot- of- ramp intersections. It would also
include reconfiguration of the eastbound West Winton to southbound I- 880
on- ramp and a new connection to Southland Mall Drive opposite the I- 880
southbound off- ramp intersection with West Winton Avenue.
Scenario 9A added to scenario 8A added an HOV extension from Hegenberger
Street to Marina Boulevard. In addition to the HOV lane on the southbound
mainline, a dedicated HOV on- ramp lane has been added at the 98th Avenue
Interchange. . Exhibits A3- 8 through A3- 15 in the appendix section illustrate the
changes coded into the model for this scenario.
Scenario 7A ( 2020) = Scenario 5A + Trade Corridor Improvement Fund ( TCIF)
Scenario 8A ( 2020) = Scenario 7A + Aux Lanes and Interchange
Improvements defined in the ACCFS
Scenario 9A ( 2020) = Scenario 8A + HOV Extension and related Interchange
Improvements
It is certainly important to note that this study benefited from the ACCFS in several
ways. First and foremost, it provided our modelers with specific details of all of the
operational improvements tested ( e. g., interchange modifications, auxiliary lanes). In
other corridor studies, these details would not have been available and would have
been left to the study team to draw conceptually.
Second, and as importantly, the conclusions of the ACCFS reflected local input and
priorities. So even though some of the improvements would not have been critical from
a bottleneck relief perspective, we believe the local consensus make these projects
easier to implement. Without such input, we may have excluded one or two projects or
changed the parameters of others from a pure technical perspective. But in the end, as
can be seen in the next section, all of the ACCFS projects do indeed improve corridor
performance ( as shown in the next section) and the sometimes tough work of selling
projects to the local stakeholders has already been done.
I- 880 Corridor Management Plan Demonstration
Page 17 of 88
Corridor Management Plan Page - 17 System Metrics Group, Inc.
III. SCENARIO RESULTS
This section first discusses how scenarios were evaluated and then presents the model
output summaries for the different scenarios.
Scenario Analysis Approach
For every model run, output statistics were provided and divided by major segment,
direction, and time of day ( i. e., AM Peak, PM Peak). An example of an output is shown
below under Exhibit III- 1. The statistics included Delay ( measured as the difference
between free flow and actual travel speeds), Vehicle Miles Traveled ( VMT), and Vehicle
Hours Traveled ( VHT). Note that the statistics are also broken down by hour as well as
by mainline, on- ramp, off- ramp, and arterial).
Exhibit III- 1: Example Model Output
Northbound Section 1 SR- 237 to SR84 Northbound Directional freeway distance 13.1 miles Directional freeway Freeway On Ramp Off Ramp Arterial*
Delay
06: 00 - 07: 00 101.37 2.72 6.18 8.91
07: 00 - 08: 00 175.48 14.14 12.68 23.28
08: 00 - 09: 00 154.89 7.23 11.25 23.92
Total Peak Period 431.74 24.08 30.11 56.12
VHT
06: 00 - 07: 00 924.82 42.00 30.71 37.30
07: 00 - 08: 00 1203.41 75.81 50.92 76.44
08: 00 - 09: 00 1185.43 69.15 51.73 76.55
Total Peak Period 3313.66 186.96 133.36 190.29
VMT
06: 00 - 07: 00 50021.03 1959.09 1222.51 1050.59
07: 00 - 08: 00 62502.24 2973.91 1857.46 1947.78
08: 00 - 09: 00 62634.67 2993.49 1947.92 1927.42
Total Peak Period 175157.93 7926.49 5027.89 4925.78
When such results were provided for the aforementioned scenarios, they were first
evaluated for reasonableness. In several cases, the models had to be adjusted and
rerun to address concerns voiced by reviewers.
Second, the results were compared to the appropriate base model results as well as
preceding scenario results. For example, Scenario 1A ( programmed projects to be
delivered before 2012) were compared against the 2006 Base Year model. Scenario
3A ( Scenario 1A plus ALINEA) was compared against Scenario 1A.
I- 880 Corridor Management Plan Demonstration
Page 18 of 88
Corridor Management Plan Page - 18 System Metrics Group, Inc.
The results were then aggregated to derive overall congestion reduction. Then, these
aggregated results were used to derive other benefits using the Caltrans Cal- B/ C
model1. GHG emissions were also estimated. Finally, we computed the benefit cost
ratios of each scenario.
Model Output Summaries
This subsection presents the evaluation results of the different scenarios. First, Exhibit
III- 2 presents the delay comparisons of the different 2006 model runs and includes the
Base Year 2006, Scenario 1A ( the three programmed projects to be delivered by 2012),
and Scenario 3A ( Scenario 1A plus the implementation of the ALINEA ramp metering
algorithm). The delay numbers are the sum of mainline, ramps, and arterial delays.
Exhibit III- 2: 2006 Base Year Model Delay Scenario Results
( Daily Vehicle Hours of Delay)
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
Base Year ( 2006) Scenario 1A Scenario 3AA
Analysis Scenario
Delay
AM Peak Period
PM Peak Period
Exhibit III- 3 shows the percent reductions in delay for the two peak periods and overall.
Note that Scenario 1A reduces delay by more than nine ( 9) percent in the AM Peak
period and by less than four ( 4) percent in the PM Peak period. However, Scenario
3AA ( i. e., adding ALINEA) leads to almost equal delay reductions in both peak periods.
Exhibit III- 4 presents the percent reductions in delay by direction. Note that Scenario
1A reduces delay more significantly in the southbound direction. Again, Scenario 3AA
( i. e., adding ALINEA) leads to almost equal delay reductions in both directions.
1 The Cal- B/ C model is a PC- based spreadsheet model developed by the Office of Transportation
Economics at Caltrans. It can be used to analyze many types of highway construction and operational
improvement projects, as well as some Intelligent Transportation System ( ITS) and transit projects. It can
be accessed and downloaded via the web at: http:// www. dot. ca. gov/ hq/ tpp/ offices/ ote/ benefit. html
I- 880 Corridor Management Plan Demonstration
Page 19 of 88
Corridor Management Plan Page - 19 System Metrics Group, Inc.
Exhibit III- 3: Peak Periods Percent Delay Reductions Compared to 2006 Base Year
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
11%
12%
13%
14%
15%
Scenario 1A Scenario 3AA
Analysis Scenario
Percent Delay Reduction
AM Peak Period
PM Peak Period
Total
Exhibit III- 4: Percent Delay Reductions by Direction Compared to 2006 Base Year
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
11%
12%
13%
14%
15%
16%
17%
18%
19%
20%
Scenario 1A Scenario 3AA
Analysis Scenario
Percent Delay Reduction
Northbound
Southbound
Total
The following summarizes the results of the simulation using the Base Year Model:
1. The three programmed projects to be delivered in the short term representing
Scenario 1A reduce overall delay on the corridor by almost seven ( 7) percent,
which is significant for a congested urban corridor like I- 880.
2. Adding advanced ramp metering such as ALINEA in the short term reduces
delay further. At a minimum, the combination of the three programmed projects
and ALINEA reduce delay by 10 percent.
I- 880 Corridor Management Plan Demonstration
Page 20 of 88
Corridor Management Plan Page - 20 System Metrics Group, Inc.
In other words, this advanced ramp metering contributes more than three ( 3) percent of
delay reductions over and beyond the three programmed projects. Note however that
ALINEA significantly reduces northbound delays and actually increases southbound
delays. Investigating this further, we found that the increase in delay southbound is
primarily on the ramps and, to a lesser extent, arterials. Hypothetically, we could
eliminate the ALINEA simulation in the southern direction and gain even more benefits.
However, ALINEA, like other advanced metering systems ( e. g., SWARM) requires
multiple ( perhaps) tens of simulations to optimize its settings. For instance, our first
simulation using ALINEA led to increased delay overall on the corridor. We then
changed parameters ( e. g., the density threshold at which ALINEA gets activated) and
the results improved. We could have gone back and forth several times to get the best
results for each direction. However, due to resource constraints, this was not possible.
We therefore believe that the results can be improved further with additional parameter
optimization.
Moving on to the 2020 Horizon Year, Exhibit III- 5 compares the 2006 Base Year with
the 2020 Do Minimum Scenario ( Scenario 3AA). Note that 2020 “ Do Minimum” delays
are projected to be double 2006 Base Year delays ( southbound delays increase more).
In total, corridor delay increases from about 15,500 hours to almost 31,000 hours during
the two peak periods.
Exhibit III- 5: Base Year 2006 and Do Minimum 2020 Horizon Year Corridor Delays
( Daily Vehicle Hours of Delay)
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
Northbound Southbound
Delay
2006 Base Year
2020 Do Minimum
( Scenario 3AA)
Exhibit III- 6 presents the delay results for the scenarios tested on the 2020 horizon year
model. The scenarios are compared against the “ Do Minimum” Horizon Year.
Scenario 4A includes the three programmed projects, Scenario 5A adds ALINEA to
scenario 4A, Scenario 6A adds traveler information to scenario 5A ( which is then
dropped), Scenario 7A adds the TCIF project to Scenario 5A, Scenario 8A adds the
multiple interchange improvements and auxiliary lanes to Scenario 7A, and finally,
Scenario 9A adds the HOV extension to Scenario 8A. Exhibit III- 7 presents the percent
delay reductions for each of the scenarios when compared against the Do Minimum
Scenario 3AA results.
I- 880 Corridor Management Plan Demonstration
Page 21 of 88
Corridor Management Plan Page - 21 System Metrics Group, Inc.
Exhibit III- 6: 2020 Horizon Model Scenario Results
( Daily Vehicle Hours of Delay)
-
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
20,000
Scenario 3AA
Do Minimum
Scenario 4A Scenario 5A Scenario 6A Scenario 7A Scenario 8A Scenario 9A
Analysis Scenario
Delay
AM Peak Period
PM Peak Period
Exhibit III- 7: Percent Delay Reductions Compared to 2020 Do Minimum Scenario
( Daily Vehicle Hours of Delay)
0%
5%
10%
15%
20%
25%
30%
35%
40%
Scenario 3AA
Do Minimum
Scenario 4A Scenario 5A Scenario 6A Scenario 7A Scenario 8A Scenario 9A
Analysis Scenario
Delay
AM Peak Period
PM Peak Period
Total
The following summarizes the results of the 2020 Horizon Year model results:
1. The three programmed projects ( Scenario 4A) to be delivered by 2012 reduce
delay in 2020 by 18 percent, much more than the 7 percent projected using the
2006 Base Year model. This means that the effectiveness of these projects
increases as demand increases in the future.
2. Adding ALINEA to these three projects ( Scenario 5A) reduces overall corridor
delay by 24 percent. In other words, advanced ramp metering adds another 6
percent in delay reductions. Moreover, the delay reductions are projected in both
I- 880 Corridor Management Plan Demonstration
Page 22 of 88
Corridor Management Plan Page - 22 System Metrics Group, Inc.
directions ( as opposed to the 2006 Base Year). Again, increasing demand
improves the effectiveness of advanced ramp metering.
3. The attempt at simulating traveler information by increasing driver familiarity ( i. e.,
Scenario 6A) actually increases delay compared to Scenario 5A. Delay
reductions are estimated to be 16 percent. Investigating this further, we found
that many drivers diverted to bypass the I- 238 freeway metering and other
bottlenecks, and created new bottlenecks in both directions. We do not believe
this result represents what would really happen. The reason is that the
simulation network includes only limited arterials and therefore only permits
limited diversion. As this diversion gets exaggerated due to these limits, corridor
delay at arterials increases significantly. The driver familiarity increase was
therefore dropped from subsequent scenarios.
4. Adding the TCIF project to the combination of the three programmed projects
and ALINEA ( Scenario 7A) reduces total corridor delay by 26 percent with the
majority of the delay reductions in the northbound direction. This represents an
additional two ( 2) percent reduction compared to Scenario 5A.
5. Scenario 8A, which added a number of interchange improvements and auxiliary
lanes to Scenario 5A reduced delay by 29 percent, a further three ( 3) percent
reduction from Scenario 7A. We suspect that many of the interchange
improvements would improve delay on arterials not included in the model and
were therefore not captured.
6. Finally, Scenario 9A, which adds the HOV extension to Scenario 8A only reduces
delay by another one ( 1) percent. We believe this result to understate actual
HOV benefits. However, microsimulation models do not have a mode shift
component to estimate the additional carpooling that would take place as a result
of the HOV extension. In other words, it assumed a constant number of carpools
with and without the extension.
These results show that most of the congestion relief in the modeled network would be
captured by the three short term programmed projects ( Scenario 1A) and advanced
ramp metering. The other projects do reduce delay further, and in many cases, the
model probably understates these impacts. Hence, these projections should be
considered to be conservative.
In summary, near term total delays are projected to be reduced by 10 percent ( from
15,500 to around 14,000 daily peak period hours of delay), a significant achievement for
a highly congested urban corridor. These near term results reflect current vehicular
demand ( based on the 2006 model). As demand increases over time, longer term
( based on the 2020 model) delays are projected to be reduced by 30 percent ( from
31,000 to 21,000 daily peak period hours of delay).
I- 880 Corridor Management Plan Demonstration
Page 23 of 88
Corridor Management Plan Page - 23 System Metrics Group, Inc.
IV. POST MODEL ANALYSIS
The detailed results from the model as shown previously in Exhibit II- 16 were further
analyzed using the Caltrans Benefit Cost Model ( Cal- B/ C), which has recently been
enhanced to allow for link by link analysis and to estimate green GHG emission
reductions2.
Note that the benefit cost computations take all the costs into account even though the
benefits of several projects extend beyond the modeled I- 880 corridor. Examples
include, but are not limited to:
Alameda I- 238 Widening – The project cost was provided at almost $ 86 million.
This project should improve mobility and reduce delay on I- 580, I- 238, and I- 880.
However, the microsimulation model does not include I- 580 and only the I- 238
Interchange. As such, benefit cost results would be significantly understated.
The SR- 92/ I- 880 and the SR- 262/ I- 880 Interchange improvements do not show
benefits or reductions on congestion for either SR- 92 or SR- 262. Again, the
benefit cost ratios would be understated.
The TCIF project ( for 23th and 29th overcrossings and arterial improvements) will
help congestion on I- 880 and arterials. However, the model likely understates the
arterial benefits. Nevertheless, the model estimates that the mobility benefits on
the freeway will be relatively modest compared with other scenarios. This does
not mean that it is not a good project as it is designed to provide additional
benefits over and beyond the mobility benefits captured by modeling, such as
safety improvements.
The interchange improvements in Scenario 8A presumably improve mobility on
several local arterials not included in the model. Again, the results of the benefit
cost may be understated.
The HOV lane extension modeling does not forecast additional mode shifts to
carpooling which means that the reductions in delay are also likely understated.
The above caveats may lead the reader to believe that the model should have been
extended to include other facilities ( e. g., I- 238, SR- 92, SR- 262, arterials). However,
extending the model beyond its current limits would have been too complex for a
microsimulation model and would have probably added hundreds of thousands of
dollars to the cost. The best we can do is to understand the results and the limitations.
Hypothetically, we could have only included a portion of the costs for each project or set
of projects and made some assumptions as the percentage of the project that is
applicable to the modeled corridor. However, after consulting with District project
management, it was decided to keep the full costs and explain the associated
limitations.
2 Only Carbon Dioxide ( CO2) Green House Gas Emission reductions are estimated by Cal- B/ C
I- 880 Corridor Management Plan Demonstration
Page 24 of 88
Corridor Management Plan Page - 24 System Metrics Group, Inc.
Scenario Costs
Scenario component costs are presented in Exhibit IV- 1 and were compiled from
Caltrans and the ACCFS. The study team estimated the ALINEA implementation to
cost $ 25 million which we believe to be more than adequate.
Exhibit IV- 1: Scenario Component Costs as Provided ( in mil. $)
Project Costs
Short Term Programmed
Projects $ 2 67.60
ALINEA $ 2 5.00
TCIF Projects $ 8 5.00
Interchange and Auxiliary
Lane Projects $ 9 2.50
HOV Extension $ 1 55.50
Exhibit IV- 2 shows these initial component costs were then added for each scenario in
constant $ 2007 dollars. Note that the short term projects $ 2007 costs are less than the
ones originally provided since 2007 was one of the rare years when the construction
costs index actually declined.
Exhibit IV- 2: Scenario Costs Summary ( in 2007 mil. $)
Benefit Category
Short Term
Programmed
Projects + ALINEA + TCIF Projects
+ Interchange
and Auxiliary
Lane Projects
+ HOV
Extension
Life- Cycle Costs $ 249.00 $ 274.00 $ 359.00 $ 451.50 $ 607.00
These costs were then used in the Cal- B/ C model together with the microsimulation
model results to derive monetized benefits which are discussed next.
Scenario Benefits
Benefits for the different scenarios can be divided into three categories:
Travel Time Reductions
Vehicle Operating Cost Savings
Emissions
For more information on how the Cal- B/ C computes these different benefits, please
refer to the Caltrans web site at http:// www. dot. ca. gov/ hq/ tpp/ offices/ ote/ benefit. html.
Note that in this case, actual model speeds were used instead of having Cal- B/ C
estimate them. Also note that for scenarios that were tested on both the 2006 Base
Year Model and the 2020 horizon year model, both model results were used to estimate
I- 880 Corridor Management Plan Demonstration
Page 25 of 88
Corridor Management Plan Page - 25 System Metrics Group, Inc.
life cycle cost benefits. Therefore, Scenario 1A and 4A were combined to estimate
benefits for the three short term programmed projects. The same applies to Scenarios
3A and 5A, both of which added ALINEA to these three programmed projects.
Exhibit IV- 3 presents the benefits for each scenario by category and in total in 2007.
Note that the negative vehicle operating costs for the short term programmed project
represents another microsimulation nuance where total VMT increases because the
model can process more vehicles.
More VMT means more fuel utilization, for instance, which increases operating costs.
However, this negative should be ignored to some extent since increased VMT means
that shoulder hours would have reduced VMT ( absent induced demand). By far the
biggest benefit category is time savings ( i. e., congestion reduction).
Exhibit IV- 3: Monetized Delay Reductions Compared to 2020 Do Minimum
Scenario
Benefit Category
Short Term
Programmed
Projects + ALINEA + TCIF Projects
+ Interchange
and Auxiliary
Lane Projects
+ HOV
Extension
Travel Time Savings $ 315 $ 440 $ 477 $ 535 $ 550
Veh. Op. Cost Savings ($ 20) $ 14 $ 6 $ 17 $ 17
Emission Cost Savings $ 5 $ 9 $ 8 $ 11 $ 12
TOTAL BENEFITS $ 299 $ 464 $ 491 $ 563 $ 579
Benefits are in $ 2007 millions
Given the increased focus on global warming, we have extracted GHG emission
reduction results for the different scenarios. Exhibit IV- 4 shows the additional benefits
related to GHG emissions in aggregate. The results represent the additions of 20 years
of reductions only.
Alternatively, Exhibit IV- 5 shows annual reductions in GHG emissions starting in 2012
and going up to 2040 for Scenario 9A. Note that the reductions in GHG emissions start
at 2012 for all near term projects, but extends these benefits through 2040. Reductions
from longer term scenarios start at 2020. The graph goes through 2040 to maintain the
overall trends. This is based on the results that show that the benefits of the projects
increases as congestion levels increase ( i. e., benefits for projects using the 2020 model
are higher than the benefits for the same projects using the 2006 model).
Exhibit IV- 4: Aggregated GHG Emission Benefits by Scenario
Short Term
Programmed
Projects + ALINEA + TCIF Projects
+ Interchange
and Auxiliary
Lane Projects
+ HOV
Extension
Additional CO2 Emissions ( tons) 1 86,911 3 17,112 2 93,577 3 98,916 4 19,163
Additional CO2 Benefits ( mil. $) $ 4.8 $ 8.4 $ 7.8 $ 10.3 $ 10.7
I- 880 Corridor Management Plan Demonstration
Page 26 of 88
Corridor Management Plan Page - 26 System Metrics Group, Inc.
Exhibit IV- 5: Aggregated GHG Emission Reductions by Scenario
0
10,000
20,000
30,000
40,000
50,000
60,000
Reductions in CO2 Emissions ( tons per year)
2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038
Year
Scenario Benefit Cost Ratios
The final step was to combine the benefits and the costs and compute the benefit cost
estimates for each scenario component, which are shown below. The useful life for all
projects was assumed to be 20 years. Note that these ratios are fairly low except for
the ALINEA component. This is partly due to the understatement of the benefits
previously discussed and partly due to costs being relatively higher for some
investments compared to projected benefits.
Exhibit IV- 6: Benefit Cost Ratios for Scenario Components
Short Term
Programmed
Projects ALINEA TCIF Projects
Interchange and
Auxiliary Lane
Projects HOV Extension
BENEFIT COST RATIO
( OVER 20 YEARS) 1.30 7.12 0.47 1.16 0.15
Note that the short term programmed projects would likely yield a cost benefit ratio of
over three ( 3) if the benefits of SR- 238 and I- 580 were included. However, the TCIF
projects would probably yield relatively low benefits regardless of model extent. The
HOV extension benefit cost ratio would increase depending on the projected mode shift.
I- 880 Corridor Management Plan Demonstration
Page 27 of 88
Corridor Management Plan Page - 27 System Metrics Group, Inc.
V. CONCLUSIONS
The projects evaluated in this study are summarized on Exhibit V- 1, with evaluated
projects limited to bottleneck- related problem areas. The three short- term programmed
projects ( i. e., Scenario 1A) should yield impressive benefits that will only grow in time.
In addition, the evaluation recommends that Caltrans and its partners focus on
implementing more advanced ramp metering algorithms on the I- 880 Corridor. If
implemented correctly, this improvement will provide the highest benefits relative to its
costs. The delay reductions projected for the ALINEA implementation are but a proxy of
what can be attained with more advanced algorithms. As discussed earlier, with more
testing and optimization, we believe these results can be improved.
The TCIF projects around 23rd and 29th provide geometric upgrades resulting primarily
in safety benefits and slight mobility gains. These safety benefits cannot be quantified
in a Paramics microsimulation model and as expected, the model results show only
small improvements in mobility.
The large list of interchange improvements and auxiliary lanes that were combined and
tested together provide for a reasonable return on investment along with delay
reductions. Additional interchange by interchange modeling may be useful to delineate
specific investment benefits. The HOV extension will provide a higher return on
investment when significant shift to carpooling and transit takes place.
Finally, GHG emission reductions on this one network could add up to an average of
20,000 tons per year. This demonstrates that operational improvements can and
should contribute to the attainment of GHG emission targets mandated by Assembly Bill
32 ( AB 32) and Senate Bill 375 ( SB 375).
I- 880 Corridor Management Plan Demonstration
Page 28 of 88
Corridor Management Plan Page - 28 System Metrics Group, Inc.
Exhibit V- 1: Summary of Planned and Recommended Projects Related to Corridor Bottlenecks
Project/ Strategy
Scenario
Package Comments
Programmed/ In Construction
Alameda 238 Widening 1A
Project mostly complete. Addresses major interchange,
requires implementation of freeway to freeway metering
for full effectiveness.
92/ 880 Interchange Reconstruction 1A
In construction. Addresses major interchange, includes
auxillary lane and freeway to freeway metering.
Alameda 880/ 262 Interchange Construction 1A In construction. Addresses major bottleneck, includes
HOV lane.
Short- Range Recommended ( 2012)
Advanced Ramp Metering ( e. g., ALINEA, SWARM) 5A
Not programmed. Very high benefits compared to cost.
Could potentially exceed projected benefits with more
calibration of metering variables.
Advanced Traveler Information 6A
Unable to analyze with microsimulation; generally
favorably viewed by other research.
Long- Range Planned ( 2013- 2020)
TCIF Project ( includes 23rd and 29th Overcrossings) 7A
Project provides upgrades to geometric standards with
benefits primarily related to safety and some mobility
benefits.
I- 880 Auxiliary Lanes, Paseo Grande to Winton Avenue 8A Increases merge limits, not tested separately.
I- 880 Auxiliary Lanes, Whipple Road to Industrial Parkway
West 8A Increases merge limits, not tested separately.
I- 880/ West A Street Interchange 8A Related to occasional bottleneck, not tested separately.
I- 880/ West Winton Avenue Interchange 8A Related to occasional bottleneck, not tested separately.
I- 880/ Whipple Road Interchange 8A
Not directly related to existing bottleneck ( Whipple used
to be an occasional bottleneck in 2003 and 2004), not
tested separately.
HOV extension from Hegenberger Street to Marina
Boulevard 9A Benefits will increase with expected increase in transit
and ridesharing.
I- 880 Corridor Management Plan Demonstration
Page 29 of 88
Corridor Management Plan Page - 29 System Metrics Group, Inc.
APPENDIX SECTION
I- 880 Corridor Management Plan Demonstration
Page 30 of 88
Corridor Management Plan Page - 30 System Metrics Group, Inc.
A1. CORRIDOR DESCRIPTION
Within the Caltrans District 4 area and the Northern California Bay Area region, I- 880
East Shore South corridor from the Grant Avenue to SR- 237 was selected for this study
based on a number of criteria:
Serve significant inter- regional travel
Multi- modal in nature
Congestion is high and projected to grow
High potential for benefits and B/ C ratios
Good detection infrastructure and data
Serve the goods movement industry
Note that other corridors, especially the I- 580 and the I- 80 were originally preferred by
the stakeholders. However, due to the lack of detection on these corridors and the
need for the detection for a comprehensive performance assessment, I- 880 was
selected instead.
Freeway
The I- 880 corridor selected for this study begins from the SR- 237 interchange in
Fremont to the Oakland to Grant Avenue. SR- 237 runs in an east- west direction with
connectors to the northbound and southbound segments of I- 880. The eastbound SR-
237 to northbound I- 880 connector has three travel lanes with two metered single
occupancy vehicle ( SOV) lanes and one high occupancy vehicle ( HOV). The right- most
lane drops approximately 560 feet north of the merge providing a total of five travel
lanes for northbound traffic.
Southbound traffic at this interchange provides three through lanes and two exit lanes to
eastbound SR- 237. Outside shoulders are approximately 8- feet wide while inside
shoulders range from approximately 18- to 25- feet wide. A concrete median divides
the freeway. North of the California Circle interchange, both the northbound and
southbound directions are reduced to three through travel lanes.
A peak period HOV lane begins at the SR- 262 interchange with an auxiliary lane that
extends from SR- 262 to the Fremont Boulevard interchange. In this segment, the
freeway has three through travel lanes, one peak- period HOV lane in each direction,
and intermittent auxiliary lanes to facilitate merging and diverging traffic.
From the SR- 92 to the I- 238 interchange, both northbound and southbound lanes have
four through travel lanes and one peak period HOV lane. From the I- 238 interchange to
the I- 980 interchange, the number of total travel lanes varies from four to five in each
direction. Major interchanges in this study corridor include the SR- 237, SR- 238, SR- 84
( Dumbarton Bridge), and SR- 92 ( San Mateo- Hayward Bridge).
I- 880 Corridor Management Plan Demonstration
Page 31 of 88
Corridor Management Plan Page - 31 System Metrics Group, Inc.
The I- 880 corridor is a Surface Transportation Assistance Act ( STAA) route and
therefore large trucks are allowed to operate on it. The segment just south of the I- 980
interchange to Alameda is a California Legal Advisory route. According to the 2004
Annual Average Daily Truck Traffic on the California State Highway System published
by Caltrans in August 2005, this segment of the study corridor’s 2004 daily truck traffic
ranges from 4.4% to 10.7% of the total daily traffic.
Transit
Major transit operators within this regional corridor are the Bay Area Rapid Transit
( BART) and the Alameda- Contra Costa Transit ( AC Transit). Intercity rail service from
Amtrak also offers service from Sacramento to the Bay Area region. The Fremont line,
shown as part of the BART map on Exhibit A1- 1 below, serves an almost parallel route
to the I- 880 corridor under study.
Exhibit A1- 1: Bay Area Rapid Transit Map
I- 880 Corridor Management Plan Demonstration
Page 32 of 88
Corridor Management Plan Page - 32 System Metrics Group, Inc.
BART service operates on Mondays through Fridays from 4 a. m. to midnight, on
Saturdays from 6 a. m. to midnight, and on Sundays from 8 a. m. to midnight. In many
cases, service extends beyond midnight depending upon the station coordination of the
last running train. Based on the BART Station Profile Study conducted by San
Francisco Bay Area Rapid Transit District in 1999, BART surveys show an average
daily ridership of more than 300,000. Several million Bay Area residents take BART
each year, often for occasional travel to events, shopping, or visiting friends and family.
On a typical weekday however, most of BART’s customers are regular riders who use
BART to commute to work. With regional population growth expected to grow to 7.8
million in 2020, a 22% increase from 1995, and the elderly population expected to also
nearly double during this period, forecasts show that the BART ridership could
potentially be affected by the growing population and changes in more flexible work
schedules.
The BART Fremont service lines serve the I- 880 corridor by providing connectivity to
three end points:
The Fremont to Richmond line provides connectivity between Fremont to
Oakland, Berkeley, and Richmond.
The Fremontline also allows transfers at the Oakland City Center/ 12 Street and
MacArthur to provide a connection to the Pittsburg/ Bay Point terminus
The Fremont to Daly City line starts from Fremont with transfer stations at Bay
Fair and Balboa Park where the connection provides access to San Francisco
International Airport and Millbrae.
The BART Strategic Plan was adopted in 1999 and updated in 2003. The BART’s
system capacity goal is to create capacity for the BART core system to carry 500,000
average weekday riders by 2025. Subject to funding, the BART may be extended south
of the Fremont terminal to Warm Springs, providing additional access to handle future
increased ridership along the I- 880 corridor.
The Alameda- Contra Costa Transit District ( AC Transit) serves more than 100 local
lines within the East Bay and more than 27 Transbay to San Francisco and the
Peninsula. As the third largest all- bus system in California, AC Transit provides
connection to 21 BART stations. AC Transit’s strategic vision anticipates that ridership
will be increased to approximately 100 million per year by 2010. A new Transbay
Terminal in downtown San Francisco is expected to begin construction in 2008 and
completed within five years. This rebuilt structure will be a modern and multimodal
facility that would serve more than 100,000 passengers a day on Transbay buses, Muni,
intercity buses and Caltrain and ultimately California High Speed Rail services.
AC Transit Line S West Hayward runs parallel to I- 880 from Oliver Eden Shores Park to
the San Francisco Transbay Terminal via I- 880 with no stops. This line operates during
I- 880 Corridor Management Plan Demonstration
Page 33 of 88
Corridor Management Plan Page - 33 System Metrics Group, Inc.
the weekday directional commuting peak hours from 5: 16 a. m. to 8: 45 a. m. in the
westbound direction and from 4: 10 p. m. to 7: 10 p. m. in the eastbound direction with
frequencies ranging from 30 to 45 minutes. The Line SA Washington Manor runs south
of I- 880 starting from the San Lorenzo Village station to the San Francisco Transbay
Terminal via I- 880 with no stops. This line also operates during the weekday directly
commuting peak hours from 5: 20 a. m. to 8: 45 a. m. in the westbound direction and from
4: 00 p. m. to 7: 45 p. m. with frequencies ranging from 20 to 45 minutes. The Line SB
Newark runs south of I- 880 from the Cedar Boulevard & Stevenson Boulevard
intersection to the San Francisco Transbay Terminal also via I- 880 with no stops. This
line also operates during the weekday directional commuting peak hours from 5: 17 a. m.
to 8: 40 a. m. in the westbound direction and from 4 p. m. to 9: 15 p. m. in the eastbound
direction with frequencies from 20 to 45 minutes.
The Line OX runs along the freeway from Park Street in Oakland to the San Francisco
Transbay Terminal. This line operates during the weekday directional commuting peak
hours from 5: 30 a. m. to 9: 00 a. m. in the westbound direction and from 4: 10 p. m. to 8: 39
p. m. in the eastbound direction with frequencies from 10 to 20 minutes. The Line O
also runs along the freeway from the Posey and Webster Tube in Oakland to the San
Francisco Transbay Terminal. Line O operates daily with frequencies from 10 to 45
minutes. In the westbound direction during the weekdays, it operates from 5: 26 a. m. to
12: 10 a. m. In the eastbound direction during the weekdays, it operates from 6: 22 a. m.
to 12: 41 a. m. During the Saturdays, Sundays, and holidays, the westbound line
operates from 6: 01 a. m. to 11: 29 a. m. and the eastbound line operates from 6: 25 a. m.
to 12: 51 a. m.
The Line W runs along the freeway from the Oakland Posey and Webster Tube along
the freeway to the San Francisco Transbay Terminal. This line operates during the
weekday directional commuting peak hours from 5: 46 a. m. to 9: 00 a. m. in the
westbound direction and from 4 p. m. to 8: 49 p. m. in the eastbound direction with
frequencies of 20 minutes.
The Amtrak Capitol Corridor ( Auburn- Sacramento- Emeryville[ San Francisco]- Oakland-
San Jose) provides service between the Sacramento region and the Bay Area with
many stops in between. It starts at Auburn, runs southwest to Emeryville and
terminates south at the San Jose transfer station to Caltrain and motorcoach service
lines south of San Jose. A station is available just south of the SR- 237 and I- 880
interchange in Santa Clara and the line runs adjacent to I- 880 to north of Oakland and
Sacramento.
Intermodal Facilities
The Port of Oakland is a major seaport facility that is growing and planning to capture a
larger share of west coast maritime activities. The Port currently processes almost 1.7
Twenty- foot Equivalent Units ( TEUs) annually. As such, the Port is a major origin and
destination of significant truck trips. An aerial of the Port is shown in Exhibit A1- 2.
I- 880 Corridor Management Plan Demonstration
Page 34 of 88
Corridor Management Plan Page - 34 System Metrics Group, Inc.
Exhibit A1- 2: Port of Oakland Aerial
Ten Container terminals and two intermodal rail facilities serve the Oakland waterfront.
The Union Pacific and the Burlington Northern and Santa Fe Railway Company ( BNSF)
railroad facilities are located adjacent to the marine terminal area to provide a reliable
and efficient movement of cargo between the marine terminals or transload facilities and
the intermodal rail facilities.
Through its Vision 2000 Maritime Development Program, the BNSF and Port of Oakland
reached an agreement in 2002 for BNSF to operate the Port's Joint Intermodal
Terminal, known as Oakland International Gateway. BNSF will also be able to provide
service to other third parties for this facility, which will also benefit the community by
taking more than 20,000 truck moves a year off Interstate 80. Oakland International
Gateway ties into BNSF's rail network by way of trackage rights and specific access
conditions approved by the Surface Transportation Board ( STB) to BNSF as part of the
1995 Union Pacific/ Southern Pacific Merger Settlement Agreement.
I- 880 also serves the Oakland Airport, which grew even after the 9/ 11. Exhibit A1- 3
below presents the overall trend for passenger volumes over time. Note that in 2008,
the airport passenger volumes exceeded dropped significantly back to 2001 levels.
Exhibit A1- 4 shows the overall trend for cargo volumes over time. Cargo volume also
dropped in 2008, albeit by a smaller percentage than passenger volume.
I- 880 Corridor Management Plan Demonstration
Page 35 of 88
Corridor Management Plan Page - 35 System Metrics Group, Inc.
Exhibit A1- 3: Oakland Airport Passenger Volume Trends
Exhibit A1- 4: Oakland Airport Cargo Volume Trends
I- 880 Corridor Management Plan Demonstration
Page 36 of 88
Corridor Management Plan Page - 36 System Metrics Group, Inc.
Special Event Facilities
I- 880 corridor also serves the McAfee Coliseum, which is the home stadium for Major
League’s Baseball’s Oakland Athletics and the National Football League’s Oakland
Raiders. Right next to it is the enclosed Sports Arena which is the home of the National
Basketball Association’s Golden State Warriors. Between these three professional
franchises, there are more than 130 events that impact the mobility on I- 880. Other
events such as concerts also contribute to the transportation demand on I- 880 corridor.
An aerial of the McAfee Coliseum and the Sports Arena is shown on Exhibit A1- 5 below.
Exhibit A1- 5: McAfee Coliseum and Adjacent Sports Arena Aerial
Land Use
The Association of Bay Area Governments ( ABAG) is a regional land use planning
agency responsible for describing existing conditions, forecasting changes to the
population and economy, and assisting local governments to identify policies that
address a changing environment.
The traditional focus of ABAG's research and analysis has been its biennial long run
forecast of the region known as Projections. The next forecast, Projections 2007, is
I- 880 Corridor Management Plan Demonstration
Page 37 of 88
Corridor Management Plan Page - 37 System Metrics Group, Inc.
expected to be issued at the end of 2006. Projections 2007 will describe the changes in
population, housing and employment within the region over the next 25 years.
ABAG also produces a short term forecast which identifies economic changes for the
coming two years. This short term forecast is released each January at a special
conference. The conference information includes a state- level forecast, regional retail
sales forecast information and information on the regional housing situation.
The ABAG Projections 2005 forecasted population, housing, jobs, and income for the
nine- county San Francisco Bay Region to year 2030. Comparing Projections 2005 to
Projections 2002 extended to 2030, the newer forecast predicts 121,970 more housing
units.
The additional housing would mean that almost 330,000 additional residents will live in
the region by 2030. The additional housing is also expected to provide a home for
approximately 180,000 more employed residents than forecasted by the Projections
2002 base- case forecast. The increase in employed residents is significant, when
compared to the number of jobs in the region, because it gives a rough estimate of the
net interregional commute.
Projections 2005 forecasts over 46,000 fewer jobs than Projections 2002. This is a
result of the slow pace of job growth in the Bay Area during the early part of the
forecast. With the forecasted increase in residents by 2030, the Construction and
goods and services sector jobs are expected to increase while jobs in other economic
sectors are expected to slow due to the slower economy of the last few years.
Projections 2005 also included forecasts based on implementation of Smart Growth
policies. It assumes that state, local, or regional policy makers would change land use
policies or other types of funding decisions in a way that would change regional
development. This in effect results in a higher number of housing units produced than
under previous forecasting assumptions. Although ABAG did not adopt the numerical
values of the Smart Growth Vision, the Projections 2005 analysis included information
from the Smart Growth Vision.
Government Lands
As shown in Exhibit A1- 6, I- 880 corridor includes an array of government- owned lands.
Within three miles of the corridor nearly 50 square miles are owned by federal, state, or
local governments. Most of the land consists of recreational areas and abandoned
military bases, including:
Oakland Army Terminal, which was closed as part of a 1993 government base
closure program. The Oakland Base Reuse Authority approved plans for
conversion to civilian use, involving creation of an industrial park and job- training
center, with much of the waterfront being placed under control of the Port of
Oakland
I- 880 Corridor Management Plan Demonstration
Page 38 of 88
Corridor Management Plan Page - 38 System Metrics Group, Inc.
Alameda Naval Station, which was the headquarters for Fleet Air Pacific, but
closed as part of a 1993 government base closure program
Coast Guard Island, which is the Northern California Headquarters for the US
Coast Guard
Knowland Park, which includes the Oakland Zoo
Garin Regional Park, a recreation area built on former ranch lands containing a
blacksmith shop and exhibits about ranching history
Ardenwood Regional Preserve, which is part of the East Bay Regional Park
District and includes the Patterson Ranch mansion and gardens as well as a
working demonstration farm
San Francisco Bay National Wildlife Refuge, which encompasses over 18,000
acres of estuarine habitat, including uplands, open water, mudflats, salt ponds,
and salt marshes
Exhibit A1- 6: Government- Owned Land
Parks and Recreational Areas
I- 880 Corridor Management Plan Demonstration
Page 39 of 88
Corridor Management Plan Page - 39 System Metrics Group, Inc.
Exhibit A1- 7 shows the location of 18 recreational areas along the I- 880 corridor.
Extensive recreational opportunities are available and a number of these are part of the
East Bay Regional Park District. There are no public or private golf courses within three
miles of the corridor.
Exhibit A1- 7: Recreational Areas
Schools
I- 880 corridor includes over 275 public and private elementary schools, middle schools,
high schools, and public academies. Exhibit A1- 8 shows the location of these
educational facilities. These schools impact the traffic on the I- 880 corridor both in the
morning and afternoon.
I- 880 Corridor Management Plan Demonstration
Page 40 of 88
Corridor Management Plan Page - 40 System Metrics Group, Inc.
Exhibit A1- 8: Educational Facilities
Hazardous Material Sites
According to the hazardous waste sites database provided by Caltrans District 4, there
are more than 44,000 sites within three miles of the corridor. These include sites
identified by the State and the United States Environmental Protection Agency ( US
EPA) under the Comprehensive Environmental Response, Compensation, and Liability
Act ( CERCLA) and the Resource Conservation and Recovery Act ( RCRA). There are
sites on the state priority list and two sites are on the national priority list ( NPL).
Exhibit A1- 9 shows the location of the seven largest sites:
Site 1 the former Alameda Naval Air Station) includes underground linking
storage tanks and is on the RCRA large generators lists and the state priority list.
Site 2 is on the Department of Toxic Substances Control Hazardous Waste and
Substances Sites ( Cortese) List
Site 3 is on the state list of underground storage facilities.
I- 880 Corridor Management Plan Demonstration
Page 41 of 88
Corridor Management Plan Page - 41 System Metrics Group, Inc.
Site 4 is on the state CERCLIS list.
Site 5 is an RCRA small generator.
Sites 6 and 7 are on the NPL, the state priority list, the regional priority list, and
RCRA CERCLIS.
Exhibit A1- 9: Primary Hazardous Waste Sites
Wetlands
According to the US Fish and Wildlife Service’s National Wetland Inventory ( NWI),
nearly 74,000 acres of wetlands are within three miles of I- 880 corridor. The NWI
groups wetlands into five major classes under the Cowardin system:
Marine – open ocean overlying the continental shelf and coastline exposed to the
open ocean
I- 880 Corridor Management Plan Demonstration
Page 42 of 88
Corridor Management Plan Page - 42 System Metrics Group, Inc.
Estuarine – deepwater tidal habitats and adjacent tidal wetlands semi- enclosed
by land, but with access to the ocean
Riverine – wetlands and deepwater habitats contained with a channel
Lacustrine – wetland and deepwater habitats situated in a topographical
depression or dammed river
Palustrine – Nontidal wetlands dominated by trees, shrubs, persistent emergents,
emergent mosses, or lichens.
Since I- 880 corridor parallels the San Francisco Bay coastline, much of the adjacent
land west of the corridor consists of estuarine wetlands ( approximately 47,000 acres),
as shown in Exhibit A1- 10. Most of the other wetlands are lacustrine ( approximately
20,000 acres) or palustrine ( less than 7,000 acres).
Exhibit A1- 10: Wetland Locations
I- 880 Corridor Management Plan Demonstration
Page 43 of 88
Corridor Management Plan Page - 43 System Metrics Group, Inc.
A2. COMPREHENSIVE PERFORMANCE ASSESSMENT
Comprehensive performance measurement and evaluation is the foundation for
implementing the system management philosophy. Without understanding how any
corridor performs and why it performs the way it does, it is impossible to truly practice
system management. For the I- 880 corridor, the performance assessment efforts
included three critical steps as follows:
Compute and evaluate corridor- wide performance and trends thereof
Identify key bottlenecks
Understand the relative contributions of each bottleneck to overall corridor
performance
For this project, freeway performance was measured using the Performance
Measurement System ( PeMS), a software tool designed at the University of California,
Berkeley to host, process, retrieve and analyze road traffic conditions information. The
PeMS database logs data from California freeway traffic detectors, as well as incident-related
data from the California Highway Patrol ( CHP) and weather data. PeMS
features a web- based Graphical User Interface ( GUI) that provides the ability to extract
various representations of the data. PeMS is a joint effort by Caltrans, the University of
California, Berkeley, and the Partners for Advanced Transit and Highways ( PATH) - a
joint venture between Caltrans, the University of California, other public and private
academic institutions, and private industry. PeMS is a traffic data collection, processing
and analysis tool to assist traffic engineers in assessing the performance of the freeway
system. PeMS extracts information from real- time and historical data and presents this
information in various forms to assist managers, traffic engineers, planners, freeway
users, researchers, and traveler information service providers. PeMS obtains 30-
second loop detector data in real- time from each Caltrans District Transportation
Management Center ( TMC). The data are transferred through the Caltrans wide area
network ( WAN) to which all districts are connected. The 30- second data received by
PeMS consist of counts ( number of vehicles crossing the loop), and occupancy ( the
average fraction of time a vehicle is present over the loop). Exhibit A2- 1 presents
PeMS connectivity with the TMCs and two of its GUI screens.
PeMS processes the data in real- time and performs the following steps:
Performs diagnostics on the data to determine if the loop detector is faulty;
Aggregates 30- second values of counts and occupancy to lane- by- lane, 5- minute
values
Calculates the speed for each lane based on individual g- factors ( which
represent the average vehicle length) for each loop detector in the system
Aggregates the lane- by- lane value of flow, occupancy, and speed across all
lanes at each detector station
Computes performance measures
Aggregates across geographical boundaries.
I- 880 Corridor Management Plan Demonstration
Page 44 of 88
Corridor Management Plan Page - 44 System Metrics Group, Inc.
Exhibit A2- 1: PeMS Connectivity to TMCs and Example Screens
Corridor- wide Performance Measures and Trends
Corridor- wide performance measures were computed for five years ( 2003 through
2007) where data was available. A notable exception is safety performance results,
which were computed using the Caltrans TASAS database from January 1999 through
December 2006. The measures computed include:
Mobility Measures – Delay, travel time
Reliability Measures – Variability of travel time
Safety Measures – Number of collisions, number of incidents
Productivity – Lost Lane miles
I- 880 Corridor Management Plan Demonstration
Page 45 of 88
Corridor Management Plan Page - 45 System Metrics Group, Inc.
Corridor- Wide Mobility Results - Delay
Delay was computed for four time periods: AM peak ( from 6 am to 9 am), mid day ( 9 am
to 3 pm), PM peak ( 3 pm to 7 pm), and evening/ early am ( from 7pm to 6 am). Delay is
computed as the difference in travel time between actual congested conditions and
freeway conditions ( assumed to reflect speeds of 60 miles per hour). Exhibits A2- 3 and
A2- 4 on the next page show the three- year trend in overall weekday delay ( i. e.,
excluding weekends and holidays) for the three years analyzed for the northbound and
southbound directions respectively. Note that the PM peak period generally has the
highest delays, followed by the AM peak period.
It is evident from the two exhibits that the southbound travel experiences higher delays
overall than the northbound direction. Finally, it is evident that delay varies significantly
from day to day, week to week, and month to month. All the spikes on both exhibits
show that using one or two days of data can lead to less than defensible conclusions.
In 2006 for instance, some days experienced less than 5,000 hours of total delay, and
others 10,000 hours or more. Clearly, to truly compute “ average delays”, the sample
size of days must be quite large.
To compare, we averaged daily delay for each year using the same data that was used
to develop the charts. The results are shown below in Exhibit A2- 2. In the northbound
direction, after a decline in average delay in 2004, delay in the PM peak period has
been growing steadily since that time. Morning ( AM Peak) and midday delays grew
until the year 2006, and declined again in 2007. In the southbound direction, AM peak
period delays have remained somewhat constant over the five year period, but midday
and PM peak period average delays grew sharply until 2006. In the year 2007, these
delays in the southbound direction declined slightly.
Exhibit A2- 2: I- 880 Study Area Average Daily Delay by Time Period
Year AM Peak Mid Day Evening and Early
AM PM Peak Total Daily
2003 1,499 1,237 552 2,547 5,835
2004 1,124 1,067 360 2,317 4,867
2005 1,331 1,434 285 2,351 5,402
2006 1,436 1,716 308 2,644 6,103
2007 1,251 1,533 335 2,804 5,922
Year AM Peak Mid Day Evening and Early
AM PM Peak Total Daily
2003 1,924 1,397 276 2,249 5,846
2004 1,728 1,427 291 2,375 5,821
2005 1,678 1,848 232 2,444 6,202
2006 1,988 2,766 277 3,367 8,398
2007 1,976 2,426 159 2,477 7,039
Year AM Peak Mid Day Evening and Early
AM PM Peak Total Daily
2003 3,423 2,634 828 4,796 11,682
2004 2,852 2,494 651 4,691 10,688
2005 3,009 3,282 517 4,795 11,604
2006 3,425 4,482 584 6,010 14,501
2007 3,227 3,959 494 5,281 12,961
Northbound Direction
Southbound Direction
Total Corridor
I- 880 Corridor Management Plan Demonstration
Page 46 of 88
Corridor Management Plan Page - 46 System Metrics Group, Inc.
Exhibit A2- 3: Northbound Average Daily Delay by Time Period
0
5,000
10,000
15,000
20,000
25,000
Jan- 03
Apr- 03
Jul- 03
Oct- 03
Jan- 04
Apr- 04
Jul- 04
Oct- 04
Jan- 05
Apr- 05
Jul- 05
Oct- 05
Jan- 06
Apr- 06
Jul- 06
Oct- 06
Jan- 07
Apr- 07
Jul- 07
Oct- 07
Vehicle- Hours of Delay ( based on 60mph)
Evening and Early AM
PM Peak
Midday
AM Peak
2003 2004 2005 2006 2007
I- 880 Corridor Management Plan Demonstration
Page 47 of 88
Corridor Management Plan Page - 47 System Metrics Group, Inc.
Exhibit A2- 4: Southbound Average Daily Delay by Time Period
0
5,000
10,000
15,000
20,000
25,000
Jan- 03
Apr- 03
Jul- 03
Oct- 03
Jan- 04
Apr- 04
Jul- 04
Oct- 04
Jan- 05
Apr- 05
Jul- 05
Oct- 05
Jan- 06
Apr- 06
Jul- 06
Oct- 06
Jan- 07
Apr- 07
Jul- 07
Oct- 07
Vehicle- Hours of Delay ( based on 60mph)
Evening and Early AM
PM Peak
Midday
AM Peak
2003 2004 2005 2006 2007
I- 880 Corridor Management Plan Demonstration
Page 48 of 88
Corridor Management Plan Page - 48 System Metrics Group, Inc.
The next set of exhibits enables further understanding of delay characteristics and
trends. Exhibits A2- 5 and A2- 6 below show the average daily delay by month for the
northbound and southbound by time period respectively.
Exhibit A2- 5: Northbound Average Monthly Daily Delay by Time Period
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
11,000
12,000
Jan- 03
Apr- 03
Jul- 03
Oct- 03
Jan- 04
Apr- 04
Jul- 04
Oct- 04
Jan- 05
Apr- 05
Jul- 05
Oct- 05
Jan- 06
Apr- 06
Jul- 06
Oct- 06
Jan- 07
Apr- 07
Jul- 07
Oct- 07
Hours of Delay ( based on 60 miles per hour)
Evening and Early AM
PM Peak
MidDay
AM Peak
2003 2004 2005 2006 2007
Exhibit A2- 6: Southbound Average Monthly Daily Delay by Time Period
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
11,000
12,000
Jan- 03
Apr- 03
Jul- 03
Oct- 03
Jan- 04
Apr- 04
Jul- 04
Oct- 04
Jan- 05
Apr- 05
Jul- 05
Oct- 05
Jan- 06
Apr- 06
Jul- 06
Oct- 06
Jan- 07
Apr- 07
Jul- 07
Oct- 07
Hours of Delay ( based on 60 miles per hour)
Evening and Early AM
PM Peak
MidDay
AM Peak
2003 2004 2005 2006 2007
I- 880 Corridor Management Plan Demonstration
Page 49 of 88
Corridor Management Plan Page - 49 System Metrics Group, Inc.
These two exhibits better reflect the higher PM peak delays, the trend of increasing
delays in 2006, and the fact that southbound delays are higher than northbound delays.
They also show the seasonality of delay during the year. At this point, the very high
evening/ early morning delays in the northbound direction in December, 2003 have not
been explained. It may be related to construction and/ or maintenance activities that
were undertaken at night.
As mentioned earlier, delays presented to this point represent the different in travel time
between actual conditions and free flow conditions at 60 miles per hour. This delay can
be segmented into two components:
Severe delay – delay that occurs when speeds are below 35 miles per hour
Other delay – delay that occurs when speeds are between 35 miles per hour and
60 miles per hour
Severe delay represents breakdown conditions and is generally the focus of congestion
mitigation strategies. On the other hand, “ other” delay represents conditions
approaching the breakdown congestion, leaving the breakdown conditions, or areas that
do not cause wide- spread breakdowns, but cause at least temporary slowdowns.
Although combating congestion requires the focus on severe congestion, it is important
to review “ other” congestion and understand its trends. This could allow for pro- active
intervention before the “ other” congestion turns into severe congestion.
Exhibit A2- 7 shows the severe congestion related delay averages by year for both the
northbound and southbound directions. Exhibit 24 presents the information for the non-severe
or “ other” congestion related delay.
With the exception of year 2003, Fridays have tended to be the most congested
weekdays in both directions, although the difference between Friday and other
weekdays is more noticeable in the northbound direction within a given year. In
contrast, Mondays have tended to be the least congested weekdays in the northbound
direction along with Tuesdays.
The significant spike in congestion in the year 2006 identified earlier is also noticeable
in these exhibits. Most of this spike has been in the southbound direction. In 2007,
most of this congestion appeared to have been alleviated, which could have been
caused by construction activities on the southern end of the corridor.
Exhibit A2- 8 shows non- severe congestion, and illustrates that slowing below between
35mph and 60mph only contributes around 30% to 35% to total delay, with severe delay
contributing the remaining two- thirds.
Another way to understand the characteristics of congestion and related delays is
shown on Exhibits A2- 9 and A2- 10, which summarize average weekday hourly delay for
the three years analyzed.
I- 880 Corridor Management Plan Demonstration
Page 50 of 88
Corridor Management Plan Page - 50 System Metrics Group, Inc.
Exhibit A2- 7: Average Severe Congestion by Day of Week
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
Mon
Tue
Wed
Thu
Fri
Sat
Sun/
Hol
Mon
Tue
Wed
Thu
Fri
Sat
Sun/
Hol
Northbound Southbound
Day of Week
Average Daily Vehicle- Hours of Delay (@ 60mph)
2003 Severe Delay
2004
2005
2006
2007
Exhibit A2- 8: Average Non- Severe, Other Congestion by Day of Week
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
Mon
Tue
Wed
Thu
Fri
Sat
Sun/
Hol
Mon
Tue
Wed
Thu
Fri
Sat
Sun/
Hol
Northbound Southbound
Day of Week
Average Daily Vehicle- Hours of Delay (@ 60mph)
2003 Severe Delay
2004
2005
2006
2007
I- 880 Corridor Management Plan Demonstration
Page 51 of 88
Corridor Management Plan Page - 51 System Metrics Group, Inc.
Exhibit A2- 9: Average Northbound Weekday Hourly Delay
0
100
200
300
400
500
600
700
800
900
1,000
1,100
1,200
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Hour of the Day
Average Daily Vehicle Hours of Delay ( based on 60mph)
2007 Hourly Delay
2006
2005
2004
2003
Exhibit A2- 10: Average Southbound Weekday Hourly Delay
0
100
200
300
400
500
600
700
800
900
1,000
1,100
1,200
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Hour of the Day
Average Daily Vehicle Hours of Delay ( based on 60mph)
2007 Hourly Delay
2006
2005
2004
2003
I- 880 Corridor Management Plan Demonstration
Page 52 of 88
Corridor Management Plan Page - 52 System Metrics Group, Inc.
These two exhibits help identify the peak hour for congestion both in the AM and PM
peak periods ( 8 to 9 am in the morning, and 5 to 6 pm in the afternoon). They also help
review “ peak spreading” trends, which reflect the extent to which congestion spreads ( or
compresses) during the peak commute periods.
The significant trend from these two charts is that the PM peak period in the northbound
direction is spreading into the midday period. In the years 2004 and 2005 the PM peak
actually got shorter and started nearly 45 minutes later in the afternoon. However, by
the year 2006, the PM peak period started around 2: 00 or 2: 30 PM in the afternoon.
The northbound AM peak period, though not spreading is becoming more intensely
congested.
In the southbound direction, the spike in the year 2006 is very apparent in Exhibit A2- 10
as that trend line stands out, particularly in the midday and PM peak periods. However,
of note in the southbound direction is that the AM peak period intensity of congestion
has grown in both 2006 and in 2007. Congestion in the southbound direction is more
intense than in the northbound direction as illustrated by both the height of the peak
periods in Exhibit A2- 10 compared to Exhibit A2- 9 and the widths of the peak periods.
Corridor- Wide Mobility Results – Travel Time and Reliability of Travel Time
In addition to understanding delay characteristics and trends, it is useful to understand
the impacts of congestion on the traveler. The best mobility result the traveler relates to
is travel time.
For the purposes of the I- 880 corridor study area, the entire corridor delineates points A
and B. Travel time statistics provided represent either an entire southbound trip ( from
Grand Avenue to SR- 237) or an entire northbound trip ( SR- 237 to Grand Avenue).
Obviously, these travel times differ by time of day. Exhibits A2- 11 and A2- 12 show the
average weekday travel times by time of day for the five years analyzed for the
northbound and southbound directions respectively. Note that the hourly travel time
trends are consistent with the hourly delay trends in Exhibits A2- 9 and A2- 10 ( as should
be expected).
I- 880 Corridor Management Plan Demonstration
Page 53 of 88
Corridor Management Plan Page - 53 System Metrics Group, Inc.
Exhibit A2- 11: Average Northbound Travel Times by Hour
30
35
40
45
50
55
60
65
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Hour of the Day
Average Weekday Travel Time ( minutes)
2007 Avg Travel Time ( min)
2006
2005
2004
2003
Travel Time @ 60mph
Travel Time @ 35mph
Night AM Midday PM Night
Travel Time @ 35 mph ( Congested Travel Time)
Travel Time
@ 60 mph
( Free- Flow
Travel Time)
Exhibit A2- 12: Average Southbound Travel Times by Hour
30
35
40
45
50
55
60
65
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Hour of the Day
Average Weekday Travel Time ( minutes)
2007 Avg Travel Time ( min)
2006
2005
2004
2003
Travel Time @ 60mph
Travel Time @ 35mph
Night AM Midday PM Night
Travel Time @ 35 mph ( Congested Travel Time)
Travel Time
@ 60 mph
( Free- Flow
Travel Time)
I- 880 Corridor Management Plan Demonstration
Page 54 of 88
Corridor Management Plan Page - 54 System Metrics Group, Inc.
Despite the consistencies between delay and travel times, it is critical to understand that
significant delay changes often mean small travel time changes. For instance, using the
data from Exhibit A2- 2, the most dramatic change in delay occurred in the AM peak
period in the northbound direction from 2003 to 2004. Delay was reduced from 1,499
hours to 1,124 hours representing a decline of almost 25 percent. During that same
time, the maximum reduction in average travel time was less than 3 minutes.
Decision makers are sometimes surprised when improvement strategies are predicted
to reduce travel time by only two or three minutes. However, as was just shown, two to
three minutes can mean a reduction of 25 percent in congestion related delay. Without
these reduced few minutes, as demand increases, travelers face the compounded
increase in congestion. Stated differently, small reductions in overall average travel
times often relate to significant reductions in delay.
Another factor that is important to understand is the variation of travel time. Perhaps
the single most frustrating aspect for the traveling public is the “ not knowing” about how
long a particular commute is going to take. Even though average travel times at 8 AM
in the northbound direction are shown to be around 46 minutes in 2007 ( as shown in
Exhibit A2- 11), few travelers experience the exact average travel time on a day to day
basis. In fact, commuters experience a large variation in travel times due to
seasonality, accidents, special events, road closures, and small changes in demand
( among others). Understanding these variations are important to address the
customers’ frustrations and evaluate strategies meant to reduce these variations and
thereby increasing the overall reliability of the trip.
Exhibits A2- 12 through A2- 17 illustrate this point. Exhibits A2- 12 through A2- 14 along
the top row represent the northbound direction between 2005 and 2007 while Exhibits
A2- 15 through A2- 17 on the bottom row show the southbound direction for the same
years. The axes are the same as in Exhibits A2- 12 and A2- 13 with the x- axis
representing the hour of the day and the y- axis showing the travel time. For each year
and direction, the average travel time is shown ( as in Exhibits A2- 12 and A2- 13), but in
addition the travel time is shown for the following percentiles for the given year: 70th,
85th, 95th, and 99th. For example, the 70th percentile travel time is was the travel time for
that hour of the day that a traveler would arrive within 70% of the days traveled on along
the corridor in that year as measured by PeMS.
The key finding from these exhibits is that even though the average travel time did not
vary much from one year to the next, the variability – particularly during the midday and
PM peak periods – deteriorated dramatically between 2004 and 2007. As an example
for the northbound direction. In 2006, a traveler in the 17: 00 hour ( 5: 00 PM) would have
to add nearly 18 minutes to the 46- minute average travel time for a total of 65 minutes
to ensure arrival with confidence at the 99th percentile. By 2007, this same traveler
would have to add more than 24 minutes to the average travel time of 47 minutes, and
would have to leave more than 71 minutes early to ensure arrival with 99% confidence.
I- 880 Corridor Management Plan Demonstration
Page 55 of 88
Corridor Management Plan Page - 55 System Metrics Group, Inc.
Exhibit A2- 13, A2- 14, A2- 15: Northbound Travel Time Variability
A2- 12 – 2005 NB Travel Time Variability A2- 13 – 2006 NB Travel Time Variability A2- 14 – 2007 NB Travel Time Variability
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
Exhibit A2- 15, A2- 16, A2- 17: Southbound Travel Time Variability
A2- 15 – 2005 SB Travel Time Variability A2- 16 – 2006 SB Travel Time Variability A2- 17 – 2007 SB Travel Time Variability
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
30
35
40
45
50
55
60
65
70
75
80
85
90
0: 00
1: 00
2: 00
3: 00
4: 00
5: 00
6: 00
7: 00
8: 00
9: 00
10: 00
11: 00
12: 00
13: 00
14: 00
15: 00
16: 00
17: 00
18: 00
19: 00
20: 00
21: 00
22: 00
23: 00
Time of Day
Travel Time ( in minutes)
99th Percentile Travel Time
95th %
85th %
70th %
Average Travel Time
I- 880 Corridor Management Plan Demonstration
Page 56 of 88
Corridor Management Plan Page - 56 System Metrics Group, Inc.
Corridor- Wide Productivity Results
Productivity, defined as throughput during peak congestion conditions, can be
represented by the “ lost lane miles” measure discussed in the introduction section. As
congestion occurs, flow rates on the freeway diminish due to merging, weaving, and
queuing. Exhibits A2- 18 and A2- 19 summarize the productivity losses on I- 880 for the
five years analyzed for the northbound and southbound travel directions respectively.
Similar to the delay results, productivity worsened steadily from 2003 to 2007 in the
northbound direction. Southbound, productivity declined steadily for all years, with the
exception of the year 2007, which showed an improvement over 2006.
Exhibit A2- 18: Average Northbound Lost Lane Miles
3.6
1.4
4.5
0.7
2.8
1.0
3.8
0.2
3.6
1.6
4.5
0.2
3.9
1.8
4.7
0.2
3.6
1.8
4.9
0.2
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
AM Midday PM Night
Time Period
Estimated Equivalent Lost Lane Miles
2003 Lost Lane Miles
2004
2005
2006
2007
Exhibit A2- 19: Average Southbound Lost Lane Miles
3.6
1.4
3.3
0.7
4.0
1.8
3.3
0.2
4.9
2.4
5.4
0.2
6.1
3.3
6.6
0.3
5.6
2.8
4.7
0.2
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
AM Midday PM Night
Time Period
Estimated Equivalent Lost Lane Miles
2003 Lost Lane Miles
2004
2005
2006
2007
I- 880 Corridor Management Plan Demonstration
Page 57 of 88
Corridor Management Plan Page - 57 System Metrics Group, Inc.
Strategies to combat such productivity losses are primarily related to operations and
include building new or extending auxiliary lanes, developing more aggressive ramp
metering strategies without negatively impacting the arterial network, improvements in
incident clearance times. These types of improvements will be tested using the micro
simulation models to identify the most promising and cost effective strategies.
Corridor- Wide Safety Results
Safety results are based on the TASAS database, which Caltrans maintains. It contains
all collisions on the State Highway System. In addition, incident data ( which includes
collisions and other incidents) was collected from the California Highway Patrol ( CHP)
for a week to understand the relationship between incidents and collisions ( e. g., how
many non- collision incidents occur compared to collisions).
Exhibit A2- 20 shows the results of synthesizing the incident data from CHP. The graph
depicts daily number of incidents reported. Surprisingly, the number of incidents
exceeded 100 every day analyzed. Next, incidents were discarded if their descriptions
did not suggest a likely impact on congestion ( e. g., changeable message sign
malfunction). The remaining incidents that were likely to impact congestion were then
divided by time period for the entire week. Exhibit A2- 21 shows the results of this
second step. These show that at least for the week analyzed, the highest number of
incidents occurs between 3 pm and 6 pm ( over 100 incidents during the week or over
15 per day) and between 6 pm and 9 pm ( more than 90 incidents or 13 per day).
Exhibit A2- 20: Daily CHP Incidents Reported
0
20
40
60
80
100
120
140
160
180
Number of Incidents
7/ 1/ 2004 7/ 2/ 2004 7/ 3/ 2004 7/ 4/ 2004 7/ 5/ 2004 7/ 6/ 2004
Date
Number of Incidents by Day
I- 880 Corridor Management Plan Demonstration
Page 58 of 88
Corridor Management Plan Page - 58 System Metrics Group, Inc.
Exhibit A2- 21: Week Total CHP Incidents Likely to Impact Congestion
By Time Period
0
20
40
60
80
100
120
Number of Incidents
Midnight to 3
AM
3 AM to 6
AM
6 AM to 9
AM
9 AM to
Noon
Noon to 3
PM
3 PM to 6
PM
6 PM to 9
PM
9 PM to
Midnight
Time of Day
Number of Traffic Impacting Incidents by Time of Day
The results of the collision analysis are summarized in Exhibits A2- 22 and A2- 23. The
first shows a daily count of collisions for the more than five years analyzed. The second
shows average number of daily collisions by month to better understand the overall
trend. Note that on a daily basis, the number of collisions generally ranges between 5
and 15. Obviously, these collisions add to the daily congestion, especially when they
occur during peak commute periods.
Around the beginning of the year 2002, a downward trend in average number of
collisions was established. Around that same time, Caltrans started metering the
corridor after working with the local stakeholders to agree on the ramp metering
approach. Although the data does not conclusively prove that metering was the direct
cause of the reduction in the number of collisions, it is consistent with federal and state
studies such as the Minnesota Ramp Metering Study that imply that such a correlation
in fact exists.
Comparing the number of incidents to the number of collisions, one can deduce a rule
of thumb that we have approximately five to six incidents for each collision. Of course,
incidents in general do not contribute to congestion as much as collisions do. In
subsequent discussions regarding bottlenecks, collisions will be discussed again to
better understand where and when collisions do occur and how they relate to the major
bottlenecks on the study corridor.
I- 880 Corridor Management Plan Demonstration
Page 59 of 88
Corridor Management Plan Page - 59 System Metrics Group, Inc.
To properly understand congestion and resulting delay, it is imperative to understand its
causes. Until recently, even with the detection data, it was impossible to divide
congestion into components, each relating to a specific cause.
Exhibit A2- 22: Daily CHP Collisions Reported
0
5
10
15
20
25
30
35
40
Jan- 99
Jul- 99
Jan- 00
Jul- 00
Jan- 01
Jul- 01
Jan- 02
Jul- 02
Jan- 03
Jul- 03
Jan- 04
Jul- 04
Jan- 05
Jul- 05
Jan- 06
Jul- 06
Daily Number of Collisions
1999 2000 2001 2002 2003 2004 2005 2006
Exhibit A2- 23: Average Daily CHP Collisions Reported by Month
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Jan- 99
Jul- 99
Jan- 00
Jul- 00
Jan- 01
Jul- 01
Jan- 02
Jul- 02
Jan- 03
Jul- 03
Jan- 04
Jul- 04
Jan- 05
Jul- 05
Jan- 06
Jul- 06
Month
Average Daily Collisions
1999 2000 2001 2002 2003 2004 2005 2006
I- 880 Corridor Management Plan Demonstration
Page 60 of 88
Corridor Management Plan Page - 60 System Metrics Group, Inc.
Recently, Caltrans and UC Berkeley developed an algorithm that estimates the
congestion by cause. Even though these algorithms are new and do not identify each
cause, they present an approach to estimate the contributions of major causes of
congestion. Exhibits A2- 24 and A2- 25 illustrate the results of these algorithms. They
divide overall congestion into three components: collisions, excess demand, and
potential reduction. The first two categories are self- explanatory. The third, “ potential
reduction”, reflects the potential reduction in delay if it were possible to optimize
operational strategies. Of course, it is almost impossible to fully optimize operational
strategies. However, focusing on these strategies in conjunction with reducing
collisions and/ or removing them faster will have significant congestion- relief benefits.
Exhibit A2- 24: AM Percent Delay Estimates by Cause in 2006
Accidents
24%
Miscellaneous
21%
Potential Reduction
52%
Excess Demand
3%
Exhibit A2- 25: PM Percent Delay Estimates by Cause in 2006
Accidents
23%
Miscellaneous
22%
Potential Reduction
50%
Excess Demand
5%
I- 880 Corridor Management Plan Demonstration
Page 61 of 88
Corridor Management Plan Page - 61 System Metrics Group, Inc.
Bottle