i-15 congestion pricing project monitoring and …monitoring and evaluation services task 1.3...
TRANSCRIPT
I-15 CONGESTION PRICING PROJECTMONITORING AND EVALUATION SERVICES
TASK 1.3BACKGROUND INFORMATION ON I-15
BASED ON HOV ASSESSMENT STUDY, 1988 - 1991
Prepared for
San Diego
ASSOCIATIONOF
GOVERNMENTS
June 10, 1997
SAN DIEGO STATE UNIVERSITY FOUNDATION
In association with
Jacqueline Golob Associ ates Department of Civil Bob M axwell & Associates Inc.
and Environmental Engineering Resource Decision Consultants Janusz Supernak, Project Director Eric Schreffler, ESTC
i
I-15 CONGESTION PRICING PROJECT
MONITORING AND EVALUATION SERVICES
TASK 1.3
I-15 CORRIDOR BACKGROUND INFORMATIONBASED ON HOV ASSESSMENT STUDY, 1988 - 1991
Janusz SupernakSan Diego State University
ii
TABLE OF CONTENTS
Executive Summary.........................................................................................................1
Introduction......................................................................................................................2
Results of the Volume/Occupancy Study ........................................................................5
Results of the Speed/Delay Study...................................................................................8
Results of the Land Use Study......................................................................................19
Results of the Park-And-Ride Study..............................................................................20
Results of the Bus Study ...............................................................................................21
Results of the Attitudinal Study .....................................................................................21
Summary and Conclusions............................................................................................25
Implications for the I-15 Congestion Pricing Monitoring and Evaluation Study.............26
References ....................................................................................................................29
iii
TABLES
Table 1 Changes in Vehicle Volumes on I-15 at Miramar Way, 76:00-9:00 A.M., 1988-1989-1990 ...............................................................................................7
Table 2 Extra Time Needed to Reach the 95th and 99th Percentiles of Travel Time,7:00-8:00 A.M., 1988-1989-1990...........................................................16
Table 3 Analysis of Significance in Speed Changes: Comparison of Means..............17
Table 4 Analysis of Significance in Speed Changes: Comparison of Variances.........17
Table 5 Delay Per Commuter, 1988-1989-1990..........................................................18
Table 6 User Cost Delay Reduction, 1988-1989-1990 Within the Length of the HOV Facility ...........................................................................................................18
Table 7 Daily Carbon Monoxide Emission Comparison, All Lanes, 6:00-9:00 A.M., 1988-1989-1990 ............................................................................................19
Table 8 Rideshare Choice by Wave............................................................................22
iv
FIGURES
Figure 1 Locations of the Speed Measurement Points on Main Lanes of I-15 along the Reversible HOV Facility...........................................................................4
Figure 2 Distribution by Time-of Peak and Occupancy/Classification..........................6
Figure 3 Volume Changes on I-15 at Miramar Way, 6:00 – 9:00 A.M., 1988 – 1989 – 1990.......................................................................................7
Figure 4 Speed Contours for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1988 ............................................................9
Figure 5 Speed Contours for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1989 ..........................................................10
Figure 6 Speed Contours for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1990 ..........................................................11
Figure 7 Level of Service Description for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1988......................................12
Figure 8 Level of Service Description for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1989......................................14
Figure 9 Level of Service Description for I-15 Main Lanes During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1990......................................15
Figure 10 Comparison of Photographs of Physical Models Illustrating SpeedDistributions on Main Lanes of I-15, 6:00 – 9:00 A.M., 1988 – 1989 – 1990 .....................................................................................16
Figure 11 Distribution of Attitudes Toward HOV Lanes (All Respondents) ..................23
Figure 12 Structural Relationship Among Endogenous Variables ...............................24
1
EXECUTIVE SUMMARY
The 1988-1991 study entitled "Assessment of the Effectiveness of the Reversible Roadwayfor the High Occupancy Vehicles on Interstate Route 15" provided a unique and veryuseful experience with the I-15 HOV facility and its evaluation. The multi-dimensionalassessment study demonstrated success in the implementation of the HOV reversibleroadway on I-15 in terms of congestion relief, increased HOV use and positive publicassessment of that project. It also created valuable infrastructure in terms of a database,methodology, logistics, and networking that is immediately available for the current I-15Congestion Pricing Project.
Both studies can be classified as high-visibility, important studies with a potential toinfluence transportation policy-making in metropolitan areas. Both studies combine trafficmeasurements with attitudinal panel surveys, thus offering rich, diversified data sets forstate-of-the-art analysis and interpretations.
Contrary to the 1988-1991 study, the current I-15 assessment study does not have acomplete database that dates back to the pre-project phase (before-study). Sensitivity tocontrol elements and statistical significance of analysis are particularly important in thecurrent study as the effects of the congestion pricing demonstration are likely to be moresubtle than those observed in the previous 1988-1991 I-15 HOV assessment study.
This report presents general examples of methodology and presentation ideas developedfor the 1988-1991 assessment study that are recommended for utilization in the currentcongestion pricing evaluation study. Compatibility of methods and presentations willbenefit the current project.
2
INTRODUCTION
This document summarizes the most significant findings of a comprehensive studyperformed by the Department of Civil and Environmental Engineering, San Diego StateUniversity during the 1988-1991 period. The Principal Investigator of this study was Dr.Janusz Supernak, Professor, Department of Civil and Environmental Engineering, SanDiego State University. The study was prepared for the California Department ofTransportation (Caltrans). The aim of this study was to provide a comprehensiveassessment of the effectiveness of the reversible HOV lanes on Interstate 15 (I-15) in SanDiego, in terms of their impact on mitigating the negative transportation consequences ofthe rapid development of bedroom communities along the Interstate 15 corridor. Theprimary focus of the study was the morning peak period, 6:00 - 9:00 am.
This study involved a series of technical measurements that recorded changes in volumes,vehicle occupancy, travel times, speed, delays, etc. These measurements werecomplemented by an attitudinal panel study that investigated changes in attitudes towardthe HOV facility, as well as changes in the travel behavior of both users and non-users ofthe reversible roadway. The study was performed as a three-wave set of measurementsand surveys that included the Before Study, After Study I, and After Study II, conductedduring the consecutive Springs of 1988, 1989, and 1990. The express lanes on I-15 wereopened to traffic in October 1988.
The $31.5 million eight-mile-long reversible carpool lanes were constructed in the freewaymedian between the junction of I-15 and State Route (SR) 163, and the Ted WilliamsParkway ramp (see Figure 1).
No intermediate exits or access points exist along the HOV facility's length. The lanes areopen between 6:00 and 9:00 a.m. and between 3:00 and 6:30 p.m. for use by carpools,vanpools, buses and motorcycles. A carpool is defined as a vehicle with two or moreoccupants.
The I-15 express lanes are equipped with a reliable system of safety devices, a criticallyimportant element of a reversible roadway. Changeable message signs indicate when thelanes are open to southbound traffic in the morning or northbound traffic in the afternoon. One of three standard messages is always displayed: either "Express Lanes Closed,""Express Lanes Carpool Only," or "Express Lanes All Traffic OK." The last message isdisplayed if an accident blocks two or more of the main lanes for more than two hours.
The two primary objectives of the HOV facility were:
1) to increase supply by adding two lanes to the critical segment of I-15; and
2) to modify demand by encouraging carpooling and thus using fewer vehicles to transportmore people. The volume of traffic along the route has tripled in the past decade,growing at an 11 percent rate during the last few years. It was expected that the HOVlanes would not only offer free-flow travel conditions to carpoolers, but also provide
3
needed relief to the traffic on main lanes. Potential improvement in traffic conditionsalong the corridor could also help reduce the emission of air pollutants and improve airquality.
The Final Report on the study, entitled "Assessment of the Effectiveness of the ReversibleRoadway for High Occupancy Vehicles on Interstate Route 15," is documented in sevenparts, each constituting a separate volume.
These volumes are:
Part 1.Executive Summary
Part 2.Volume/Occupancy Study
Part 3.Speed/Delay Study
Part 4.Land Use Study
Part 5.Park and Ride Study
Part 6.Bus Study
Part 7.Attitudinal Study
4
Figure 1Locations of the Speed Measurement Points
on Main Lanes of I-15 along the Reversible HOV Facility
5
RESULTS OF THE VOLUME/OCCUPANCY STUDY
Volumes and vehicle occupancies were recorded for main lanes, relevant ramps, andparallel streets during the Springs of 1988, 1989, and 1990. Figure 2 shows an exampleof combined volume/occupancy/modal split/time-of-peak recording, a format that will beparticularly useful for the I-15 Congestion Pricing Project.
Volumes and occupancies on the HOV facility were monitored monthly. Table 1 andFigure 3 summarize volumes on I-15 at the Miramar Way location during the morningpeak.
The vehicular volume also increased sharply between the Springs of 1988 and 1989, butin 1990 it actually slightly decreased, in spite of a moderate increase in commuter volume.A significant (by 53 percent) increase in the carpool population during the 1988-1990period (see Table 1) was an important finding. An important incentive to use the HOVlanes was that traffic conditions on the facility were free-flowing. Level of Service A wasmaintained during virtually the entire morning and afternoon peak periods.
6
Figure 2Distribution by Time-of Peak and Occupancy/Classification
SB Rt163 Main Lane at Kearny Villa Rd 6-6-90 0600-09001 2 3+ Bus MC Truck Total
6:15
6:30
6:45
7:00
7:15
7:30
7:45
8:00
8:15
8:30
8:45
9:00
1293
1359
1869
1779
2046
1791
1591
1686
1537
1391
1164
1085
119
181
206
210
272
210
161
149
130
127
94
92
17
16
7
17
31
23
16
13
13
9
10
12
1
4
5
6
10
13
3
3
4
4
2
6
19
15
18
17
15
13
13
11
3
6
5
3
36
33
44
36
34
52
43
62
41
61
38
52
1485
1608
2149
2065
2408
2102
1827
1924
1728
1598
1313
1250
Total 18591 1951 184 61 138 532 21457
TIME
0
500
1000
1500
2000
2500
6:15 6:30 6:45 7:00 7:15 7:30 7:45 8:00 8:15 8:30 8:45 9:00
1 2 3+ Bus MC Truck
7
Table 1Changes in Vehicle Volumes on I-15 at Miramar Way,
6:00-9:00 A.M., 1988-1989-1990
Volume Before Study1988
Volume After Study I1989
Volume After Study II1990
SOV, Main Lanes 24,244 27,457 26,537
HOV, Main Lanes 3,688 1,771 2,015
HOV, Express Lanes N/A 3,015 3,627
Total Volume(Veh/A.M. Peak)
27,932 32,243 32,179
Total Volume(Persons/A.M. Peak)
34,130 40,146 41,307
Figure 3Volume Changes on I-15 at Miramar Way,
6:00 – 9:00 A.M., 1988 – 1989 – 1990
BEFORE AFTER AFTERSTUDY STUDY I STUDY II
201517713688
4786 5642
28877
32243 32179
0
5000
10000
15000
20000
25000
30000
35000
1988 1989 1990
SOV Main LanesHOV Express LanesHOV Main Lanes
8
The volumes carried by the HOV facility in San Diego were much higher than those carriedby a similar facility when opened in September 1989, on Interstate 279 in Pittsburgh (2reversible lanes, 4.1 miles long). The San Diego morning peak-hour volumes were 1,375vehicles and 3,138 passengers, whereas the analogous Pittsburgh volumes were 160vehicles and 983 passengers. The difference in carpool occupancy requirement (SanDiego: 2+, Pittsburgh: 3+) was part of the reason for the performance difference. Thedecision to keep the carpool definition at the 2+ occupancy level for the early stage of theI-15 express lanes was proper: only approximately 12 percent of carpools had more thantwo persons in the car.
RESULTS OF THE SPEED/DELAY STUDY
Since the HOV facility was opened in October 1988, travel conditions on the main lanesimproved dramatically. Areas of deficient speed were substantially reduced between 1988and 1989, and further reduced between 1989 and 1990. The Rancho Penasquitos/Powayarea remained the most deficient area in the entire system, although a substantialimprovement could be observed there as well (Figures 4, 5, and 6 use a convenientmethod of speed contours in the distance-time domain to demonstrate I-15 travel conditionimprovements). The level of service improvements can be demonstrated by Figures 7, 8,and 9. Improvements of the speed distribution along the HOV facility during the morningpeak are best observed when comparing photographs of physical models of speeddistribution: "canyons" of deficient speeds were greatly reduced after the HOV facility wasopened in 1989 (see Figure 10).
Cumulative speed frequency distributions for the critical hour during the peak, 7:00-8:00a.m., revealed not only the substantial improvement in average speed after the HOVfacility opened, but also the significant reduction in speed variance, which allowed a highervolume through the corridor.
9
Figure 4Speed Contours for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1988
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
9: KEARNY VILLA ROAD OFF
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
10.486(not shown)
10
Figure 5Speed Contours for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1989
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
9: KEARNY VILLA ROAD OFF
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
10.486(not shown)
11
Figure 6Speed Contours for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1990
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
9: KEARNY VILLA ROAD OFF
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
10.486(not shown)
12
Figure 7Level of Service Description for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1988(LOS Definition Based on Speed Variable Alone)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
13
9: KEARNY VILLA ROAD OFF 10.486(not shown)
14
Figure 8Level of Service Description for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1989(LOS Definition Based on Speed Variable Alone)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
9: KEARNY VILLA ROAD OFF
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
10.486(not shown)
15
Figure 9Level of Service Description for I-15 Main Lanes
During the Morning Peak Period 6:00– 9:00 A.M., Late Spring of 1990(LOS Definition Based on Speed Variable Alone)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
Location
(Southbound on I-15)
Mile Marker
(from Location 1)
1: CAMINO DEL NORTE ON
1A – midpoint
2: CARMEL MOUNTAIN OFF
2A – midpoint
3: POWAY ROAD OFF
3A – midpoint
4: RANCHO PENASQUITOS ON
4A - midpoint
0.000(not shown)
0.538
1.076
2.237
3.398
3.577
3.775
4.685
5: MIRA MESA OFF
5A – midpoint
6: CARROLL CANYON OFF
6A – midpoint
7: MIRAMAR ROAD OFF
7A – midpoint
8: MIRAMAR WAY OFF
8A – midpoint
9: KEARNY VILLA ROAD OFF
5.615
6.148
6.680
6.996
7.312
7.737
8.168
9.327
10.486(not shown)
16
Figure 10Comparison of Photographs of Physical Models
Illustrating Speed Distributions on Main Lanes of I-15,6:00 – 9:00 A.M., 1988 – 1989 – 1990
The reduction in travel time over the 8.79-mile-long segment of the freeway was found tobe substantial. Not less significant was the reduction in travel time dispersion, animportant element in commuters' planning when to leave home to arrive at work on time. For example, in order for a traveler to arrive at work on time in 99 percent of workdays,he/she would have to leave home almost 22 minutes earlier than on an ideal day wherefree-flow travel conditions were guaranteed. This amount of extra time went down to 6minutes in 1990 (see Table 2).
Table 2Extra Time Needed to Reach the 95th and 99th Percentiles of Travel Time,
7:00-8:00 A.M., 1988-1989-1990
Extra Travel Time for D=8.79 miles
Percentile 1988 1989 1990
95th 17.96 min. 7.66 min. 4.56 min.
99th 21.56 min. 9.06 min. 6.26 min.
Speed changes that led into changes in the travel time were found to be statisticallysignificant (see Tables 3 and 4). The average delay decrease is summarized in Table 5.Delay is formulated as a difference between the actual and free-flow travel time neededto traverse the 8.79-mile-long segment of the freeway corresponding with the length of theHOV facility on I-15.
17
Table 3Analysis of Significance in Speed Changes: Comparison of Means
Change from 1988 Change from 1989
Year Time(a.m.)
Number ofobser.
Mean est.(M.P.H.)
Accuracy at5% level(Z=1.96)
Std.Dev.
(M.P.H.)
ZCALC Signif.Change?
ZCALC Signif.Change?
6 - 7 900 54.8 0.68 MPH 10.4 - - - -
7 - 8 900 37.9 1.02 MPH 15.6 - - - -1988
8 - 9 900 53.0 0.84 MPH 12.8 - - - -
6 - 7 450 63.1 0.64 MPH 6.9 17.460 YES** - -
7 - 8 450 54.1 1.26 MPH 13.7 19.538 YES** - -1989
8 - 9 450 62.7 0.81 MPH 8.8 16.300 YES** - -
6 - 7 450 61.3 0.50 MPH 5.4 15.113 YES** -4.358 YES**
7 - 8 450 55.5 1.05 MPH 11.4 23.536 YES** 1.666 YES*1990
8 - 9 450 62.7 0.45 MPH 4.9 19.993 YES** 0.000 NO
YES** = Significant at 1% levelYES* = Significant at 5% levelNO = Not Significant at 5% level
Table 4Analysis of Significance in Speed Changes: Comparison of Variances
Change from 1988 Change from 1989
Year Time(a.m.)
Number ofobser.
Std. Dev.(M.P.H.)
ZCALC Signif. Change? ZCALC Signif. Change?
6 - 7 900 10.4 - - - -
7 - 8 900 15.6 - - - -1988
8 - 9 900 12.8 - - - -
6 - 7 450 6.9 9.140 YES** - -
7 - 8 450 13.7 3.104 YES** - -1989
8 - 9 450 8.8 8.431 YES** - -
6 - 7 450 5.4 13.539 YES** 5.136 YES**
7 - 8 450 11.4 7.176 YES** 3.871 YES**1990
8 - 9 450 4.9 17.872 YES** 11.616 YES**
YES** = Significant at 1% levelYES* = Significant at 5% levelNO = Not Significant at 5% level
18
Table 5Delay Per Commuter, 1988-1989-1990
Average Delay / Commuter
Period 1988 1989 1990
6:00 - 9:00 a.m. 5.02 min. 1.10 min. 0.72 min.
7:00 - 8:00 a.m. 10.24 min. 3.20 min. 2.01 min.
Changes in volume and in speed distributions brought dramatic changes in the monetaryvalue of delay. A conservative estimate (a rate of $4.25 per hour, commonly used byCALTRANS in 1991, was used) showed that during the first two years of the HOV laneoperation, approximately $30 million was saved through decrease in delay (see Table 6).The 95th or 99th percentile of travel time was used for calculating delay cost for themorning peak (the commuter has to have a sufficiently high probability of arriving at workon time); afternoon delay cost was based on average travel time. ΑIf no HOV≅ columnwas based on the estimation of travel conditions on I-15 when the observed increase involume had to be handled by the main lanes only.
Table 6User Cost Delay Reduction, 1988-1989-1990 Within the Length of the HOV Facility
Annual User Cost (Millions of Dollars)
Travel TimeCalculation Base
1988 1989Actual
1989If no HOV
1990Actual
1990If no HOV
A.M.: 95th PercentileP.M.: Mean
12.39 3.81 17.36 2.35 18.67
A.M.: 99th PercentileP.M.: Mean
15.06 5.24 20.47 3.48 21.87
Substantial reductions in travel at low speeds and in stop-and-go conditions led tosignificantly reduced emissions of air pollutants. Application of the Raus−version of theEPA model based on VTM and speed data showed that the CO emission rate per userwent down by 25 percent between 1988 and 1990. The 1990 level would have been 65percent higher than it was, if the HOV lanes had not been built (see Table 7).
19
Table 7Daily Carbon Monoxide Emission Comparison, All Lanes,
6:00-9:00 A.M., 1988-1989-1990
CO Emission Levels (in thousands)
Variable 1988 1989Actual
1989If no HOV
1990Actual
1990If no HOV
CO Emission Rate(g/mile)
1,025 939 1,390 931 1,534
Vehicle CO EmissionRate (g/vehicle-mile)
36.7 29.1 39.9 28.9 43.5
User CO Emission Rate(g/person-mile)
30.0 23.4 34.6 22.5 37.1
The improvement in the travel conditions on the main lanes of I-15 was accomplisheddespite the closure of an alternate route, Pomerado Road, which was reopened aftercompletion of the study (in May of 1991). It is worth mentioning that I-15 corridor users donot have any viable alternative route, and that any potential shift from another mode oftravel (e.g. rail) does not apply to the I-15 corridor. This seems to imply that the observedchanges in volumes on I-15; increase in the size of the carpool population andimprovement in the level of service on the main lanes should be primarily attributed to theintroduction of the HOV lanes in the corridor.
RESULTS OF THE LAND USE STUDY
Northern San Diego County experienced tremendous population growth from 1980 to1990. The combined subregional areas of Del Mar-Mira Mesa, North San Diego, Poway,Escondido, and San Marcos grew by 70.1 percent in population and 72.1 percent innumber of occupied housing units between 1980 and 1990. In only a two-year period(1988-1990), residential land use acreage grew by 10.1 percent.
The North City and North County East major statistical areas (MSA) grew by 65.1 percentin employment between 1980 and 1988, and traffic volume across the entire corridor morethan doubled over the same period. Between 1979 and 1989 the increase in traffic volumeranged from 158.7 percent to 199.4 percent north of the HOV facility and from 91.9 percentto 137.8 percent parallel to the reversible roadway.
Similar growth in the region is projected to continue over the next 20 years. Population,number of occupied housing units, and civilian employment are expected to rise by 62.1percent, 83.5 percent, and 68.7 percent respectively between 1988 and 2010 in the NorthCity and North County East MSAs.
20
Population and employment increases will heavily affect traffic along the I-15 corridor. Bythe year 2010, traffic volume is expected to increase by 50 percent between SR 163 andSR 56 and by 107.1 percent between SR 56 and SR 78.
RESULTS OF THE PARK-AND-RIDE STUDY
This study analyzed changes in park-and-ride facilities along the relevant segments of theI-15 corridor and plans for their future development. Park and ride facilities constitute animportant, integral part of the transportation system on I-15 that promotes the use of highoccupancy vehicles: carpools, vanpools, and buses.
There were eight formal park-and-ride sites within the segment of I-15 between Route 78and Mira Mesa Boulevard, offering a total of 818 spaces for potential carpoolers,vanpoolers or bus users. Out of this total, only two sites with a total of 168 parking spacesare located north of Ted Williams Parkway and were therefore suitable for using the HOVlanes. These numbers could hardly be seen as adequate for the approximately 7,000persons using the HOV facility. There was a clear disproportion in park-and-ride sitedistribution between the 8-mile segment of the corridor south of Ted Williams Parkway(470 spaces) and the segment directly north of Ted Williams Parkway (168 spaces).
The two lots that were conveniently located for use with the HOV express lanes werepractically filled, one (lot 53) over its capacity. Most of the other park-and-ride locationswere under-used, sometimes drastically. As all park and ride sites do not differ in regardto security, lighting, etc., this finding may indicate that HOV riders preferred informal sitesto awkwardly located formal sites that require substantial extra distance to travel. The foursites located south of the entry to the HOV facility were used on average at 22 percent ofcapacity.
The distribution of the park-and-ride facilities in the vicinity of I-15 north of Mira Mesa didnot seem to be adequate for current and, particularly, future needs. New park-and-ridelocations in the direct vicinity of the HOV lanes entry were needed to correct thedeficiencies in park and ride facilities for HOV facility users. Expansion of park and rideLots 53 and 30 was recommended if feasible. Extra park and ride sites were needed toserve commuters from Southern Escondido and San Pasqual.
21
RESULTS OF THE BUS STUDY
Bus use on the HOV facility was relatively low if compared to other similar facilities,particularly on the East Coast. For example, the Shirley Expressway facility in Virginiacarried 5.9 passengers per vehicle due to heavy use of buses. The analogous figure forthe HOV facility in San Diego was 2.3 passengers per vehicle.
The rising trend in bus use was clearly visible in the I-15 corridor as a whole and on theHOV facility in particular. Express Routes 810 and 820 from Escondido and Powayexperienced increases in ridership from 1988 to 1990. The patronage figures on theseroutes doubled since the opening of the HOV facility. If the level of comfort offered by theexpress buses increased and appropriate park and ride facilities were added at the mostdesirable locations, a further rapid increase in bus use on the HOV facility could beexpected.
As of March 1991, there were 40 bus trips daily using the HOV facility and carrying almost1,100 passengers. The overall number of passengers served rose from 702 in the fourthquarter of 1988 to 1,074 in the first quarter of 1991, an increase of 53 percent.
RESULTS OF THE ATTITUDINAL STUDY
The Attitudinal Study consisted of two parts. The first part was done as a three-wavepanel survey. Randomly selected residents in the I-15 service area were interviewed threetimes in early Spring 1988, 1989, and 1990. The second part was a separate survey inearly Spring 1991, of new residents in the areas adjacent to the HOV facility entry.
The analysis of sample-wide averages, incorporating all available observations, indicatedthat HOV lanes positively contributed to a reduction in perceived travel time. Sample-wideaverages suggested that travel times were perceived to decrease by over 10 percent forI-15 main lane users and over 15 percent for ridesharers shortly after the opening of theHOV facility. Part of this improvement was perceived to erode one year later, which mighthave indicated some disappointment that the HOV facility did not offer an increasinglyeffective remedy to congestion.
The fraction of respondents who rideshared increased gradually from 17.4 percent in thefirst wave to 22.1 percent in the second wave, and to 24.3 percent in the third wave. Thisrepresented a significant (almost 40 percent) increase in the fraction of carpoolers. Thelong-term frequency of ridesharing estimated from the transition frequencies was 28percent, and if corrected for sample attrition, 31 percent (see Table 8).
22
Table 8Rideshare Choice by Wave
OBSERVED FREQUENCY
Drive Alone Rideshare Total
Wave 1 928 (82.6%) 196 (17.4%) 1124 (100.0%)
Wave 2 659 (77.9%) 187 (22.1%) 846 (100.0%)
Wave 3 645 (75.7%) 207 (24.3%) 852 (100.0%)
WEIGHTED FREQUENCY
Drive Alone Rideshare Total
Wave 1 928 (82.6%) 196 (17.4%) 1124 (100.0%)
Wave 2 664 (78.5%) 182 (21.5%) 846 (100.0%)
Wave 3 543 (75.0%) 181 (25.0%) 724 (100.0%)
In spite of a consistently perceived improvement in speed and travel time, the respondents'rating of I-15 main lane traffic gradually worsened between 1988 and 1990. Ride sharersperceived travel conditions on main lanes as worse than did solo-drivers. Both groupsalike showed declining traffic conditions ratings. The commuters seemed to perceive thetravel conditions continuously declining over time, with HOV lanes able to only partiallyand temporarily suspend this decline.
Attitudes toward the HOV lanes remained predominantly "strongly positive" or "positive"over all three waves, but had slightly declined from 77 percent in 1988 to 70 percent in1990 (see Figure 11). These changes in attitudes were not consistent with the perceivedreduced average trip duration for the same respondents. This result could indicate a) theeffect of a gradually decreasing interest in a "new toy," or b) the positive impact of the HOVfacility on stabilizing the expected duration of commuting time (reducing variations in traveltime caused by congestion). A significant reduction in speed variance was indeedconfirmed by speed measurements. Thus commuters may have correctly perceived thesystem as more crowded now than before. Yet, objectively, the system was more efficient,as it carried more traffic at a higher speed. It mostly operated at the upper portion of thevolume/speed distribution curve, going into the forced-flow area only occasionally. Theattitudinal study found that HOV users liked the facility more and more, whereas non-usersless and less. The probability that a commuter with a negative attitude would switch to apositive attitude was higher than vice versa.
23
Figure 11Distribution of Attitudes Toward HOV Lanes (All Respondents)
Although the overwhelming majority of the population expressed positive assessment ofthe HOV facility, it is worthwhile to summarize the sporadic complaints about the facilityand its operation. The most common complaints were: "Lanes should be open to alltraffic," "It was not worth it to spend the money on a facility that is used for only six hoursa day," "Why don't people use it more often?" (complaint popular among non-users), and "I would like to use the lanes but I cannot access them."
The increase in ridesharing does not imply that a fixed group of commuters rideshared andthat the size of the group was gradually increasing. The panel survey suggested arelatively high turnover between driving alone and ridesharing. The probability that aridesharer remained a ridesharer one year later was 75 percent; the probability that a solodriver remained a solo driver one year later was 90 percent.
Using the structural equation modeling approach, explained in detail in Part 7 of theAssessment of the Effectiveness of the Reversible Roadway for High Occupancy Vehicleson Interstate 15, Final Report, the analysis of this section unveiled a complex and dynamic
TRAVELTIME
TRAVELTIME
TRAVELTIME
RIDESHARECHOICE
RIDESHARECHOICE
RIDESHARECHOICE
HOV LANEATTITUDE
HOV LANEATTITUDE
HOV LANEATTITUDE
TRAFFICPERCEPTION
TRAFFICPERCEPTION
TRAFFICPERCEPTION
HOV LANESOPEN
T1 T2 T3
24
relationship among travel time, perception of the main lane traffic conditions, attitudetoward HOV lanes, and rideshare behavior. Substantial changes took place in therelationships among these four factors immediately after the opening of the HOV lanes.In particular, main lane traffic perception affected attitudes toward the HOV lanes, whichin turn affected rideshare choice. The period shortly after the opening of the HOV laneswas the time of major adjustment for I-15 commuters (see Figure 12).
Figure 12Structural Relationship Among Endogenous Variables
In other periods, rideshare behavior was affected by attitude, but after HOV lane opening,rideshare behavior was affected only by the travel time that commuters experienced. Attitudes and perceptions were formed by the commuter's rideshare choice and travel
25
experience, but did not affect rideshare choice.Findings that link perception and behavior in a causal way are useful in the policy context.If behavior (experience) affects attitude, then drivers could be encouraged to carpool ifgiven some reward for trying first. If the opposite is true, an effective marketing effort couldbe seen as a more promising tool to increase carpooling.
A separate study of new homeowners located near the HOV facility revealed that thedefinition of "a carpooler" should be broadened. Only 23 percent of all respondents neverused the HOV facility. Twenty-two percent of them used it regularly (at least 3 times aweek), 24 percent occasionally (average 5 times a month), 30 percent sporadically(average 5 times a year). Even a community of retired persons (Seven Oaks) used theHOV facility occasionally (44 percent) or sporadically (28 percent). Only 9 percent of newresidents had a "negative" or "strongly negative" opinion about the HOV facility. Over 22percent indicated that their decision to purchase a home in the area served by expresslanes was related to the availability of this facility for commuting.
Respondents indicated that the most popular form of carpooling was with work associatestraveling to the same destination. The average carpooler believed that using the HOVfacility saved 22 minutes a day and almost $3.00 a day. Travel time saving perceptiongoes beyond the actually experienced travel time saved. It indicates that commutersperceive a reduction in actual travel time together with decreasing uncertainty about theactual travel time on a given day, both changes confirmed by the technical study.
SUMMARY AND CONCLUSIONS
The HOV facility may be seen as a balancing element of the I-15 corridor that offeredfree-flow travel for those who choose to carpool. Main lane traffic conditions were normallyworse and more unpredictable than those on HOV lanes were.
Technical measurements confirmed the success of the HOV facility. Users saved about$30 million during the first two years in delay costs alone. Carbon monoxide (CO)emissions were down by 25 percent per user mile from the 1988 level; if no HOV lane hadbeen built, the 1990 CO emission would be 65 percent higher than it was in 1990. Average speed on the main lanes during the critical hour, 7:00-8:00 a.m., was up from 38M.P.H. in 1988 to 55 M.P.H. in 1989, and to 56 M.P.H. in 1990. The carpool populationhad increased by 53 percent between 1988 and 1990.
The attitudinal panel survey paralleled technical findings. About 75 percent ofrespondents had a positive opinion about the facility and reported increasing timesavings. Among survey respondents, the carpool population had increased by 40 percent between1988 and 1990. Commuter attitudes and behavior were interrelated. An increasingnumber of commuters were finding the HOV facility a viable alternative for regular oroccasional use.
26
The success of the San Diego HOV facility-although thoroughly documented in sevenvolumes of the Final Report of the assessment study-may still not necessarily be fullyunderstood or be considered a model solution to be copied elsewhere. Yet, it seems thatthe factors that primarily contributed to the success of the HOV express lanes on I-15 canbe identified as follows:
• The HOV lanes were added, not taken away from the main lanes.
• The relaxed carpool definition (2+ persons per vehicle) was very helpful, as the fractionof vehicles with 3+ persons was still rather small.
• Contributions from buses and cars with 3+ passengers were growing fast, but remainedlow in 1990.
• The technical performance of the HOV facility had been excellent, and the system wasenjoying a positive public image.
• As Level of Service A was virtually always offered, the reward for using the facility waswell defined and reliable.
• The HOV facility was long enough (8 miles) for commuters to notice the advantages ofusing it.
• This solution to increasing commuter traffic was compatible with transportation andenvironmental policies that were considered right for the region by the vast majority ofthe population.
• The lanes received mostly positive media attention.
IMPLICATIONS FOR THE I-15 CONGESTION PRICING MONITORING ANDEVALUATION STUDY
The 1988-1991 study entitled "Assessment of Effectiveness of the Reversible Roadwayfor High Occupancy Vehicles on Interstate 15" provides an invaluable source ofinformation, data, and experience all elements relevant for the I-15 Congestion PricingProject Monitoring and Evaluation Study.
There are several important similarities between both studies:
1) Both studies deal with the same facility: the reversible roadway in the median of I-15.Organization of traffic operations and safety devices, as well as hours of operationremained the same during the period between 1991 and 1997. Geometry of the HOVfacility also remained unchanged during that period and will be only slightly modifiedto accommodate the necessary ETC system to be introduced for the "per transaction"
27
phase of the congestion pricing project.
2) Both studies combined both technical measurements (primarily traffic data such asvolumes, occupancies, modal split, time-of-day traffic distribution, speed-delay, etc.)with series of attitudinal studies intended to monitor attitudes, perceptions and travelbehavior of I-15 travelers.
3) Both studies utilized longitudinal panel survey as the main tool to understand changesin travelers' perceptions and resulting behavior. The 1988-1991 study involved threewaves; the 1997-1999 study is designed to involve five waves of the panel.
4) Both studies were to monitor public perception of the respective ideas that could beperceived as somewhat controversial to some segments of the society. The original1988-1991 study was to test public reaction to the very idea of the HOV roadway in I-15corridor and to its performance. The current study is to test public reaction to theparticular version of congestion pricing in a large metropolitan area, and to measureeffectiveness of such a policy.
5) Both projects were designed as pioneering ones nationwide. Both are considered veryimportant studies that could offer new unconventional ways to combat increasingcongestion and air pollution in major metropolitan areas in the U.S.
6) Both projects received federal attention and assistance.
7) Both projects received extensive media attention and coverage as they affected vitalpublic interest.
There are several important differences between both studies:
1) The earlier 1988-1991 study was designed as one before study and two after studies.The first wave of both technical and attitudinal measurements was gathered before theHOV facility was opened for traffic. This gave a comprehensive base for comparisonsaimed at understanding the influence of the project. The I-15 Congestion PricingProject Monitoring and Evaluation Study started after the project was initiated inDecember 1996. Although some data can be tracked retroactively, this situationcreates a serious void that reduces opportunities to study completely changes causeddirectly by the Congestion Pricing Project.
2) The new project has a very large, diversified scope, exceeding the scope of theprevious study.
3) The Congestion Pricing Project has a much stronger federal interest in it. Methodologies are scrutinized to a larger degree to assure both technical soundness,as well as compatibility of the I-15 study with other U.S. projects designed as pilotstudies for congestion pricing.
28
4) Surveys and analytical methods have to be much more sensitive in the new I-15 studyto the fact that the expected effects and changes should be much more subtle thanthese observed in the 1988-1991 study. Opening of the HOV facility created in 1989a radical change in the way I-15 commuters could be served and predictably dramaticchanges in many characteristics. This should not be the case now: care has to be givento separate the project-related effects from those that are not project-related. Statisticalsignificance of the effects measured becomes even more important than in the previousstudy. For example, the 1988-1991 speed-delay study previously revealed significantday-to-day variability. Multi-day speed-delay study becomes imperative as it servesas an important database for cost of delay and emission analysis.
5) The Congestion Pricing Project requires even more coordination and closecollaboration with many partners involved with this project. Scheduled interactions inthe form of meetings and progress reports are essential for the success of this study.
Implications of the previous I-15 study from 1988-91 for the new I-15 Congestion Pricingstudy are limited as both studies are different. The main benefits coming from the 1988- 1991 study are: established infrastructure in terms of the database, some commonmethodologies, logistics, and networking. The new attitudinal panel will be much moreelaborate that the previous one, thus giving opportunity to more diversified and morecomplete analysis to understand dynamics and causality of the congestion pricing impacts.Careful use of control corridor (I-8) for the congestion pricing evaluation project will giveSDSU team a better tool to draw conclusions than was the case with the previous studyintended as before-after study with no extra corridor designated as control corridor.
A very important lesson learned during the previous study is the need to assure statisticalvalidity of the results obtained by designating an appropriate number of days to gatherdata on speed distribution, modal split, vehicle classification, etc.
29
REFERENCES
Assessment of the Effectiveness of the Reversible Roadway for High Occupancy Vehicleson Interstate Route 15, Final Report 1991.
Part 1. Supernak, J.: Executive Summary
Part 2. Supernak, J.: Volume/ Occupancy Study
Part 3. Supernak, J.: Speed/ Delay Study
Part 4. Chandler, R. and J. Supernak: Land Use Study
Part 5. Supernak, J. and L. Honda: Park and Ride Study
Part 6. Supernak, J., Manela D., and S. Sexton: Bus Study
Part 7. Golob T., Kitamura R. and J. Supernak: Attitudinal Study
Appendix 2.1 Supernak, J: Results of Volume/ Occupancy Counts, Before Study, Spring1988.
Appendix 2.2 Supernak, J: Results of Volume/ Occupancy Counts, After Study I, Spring1989.
Appendix 2.3 Supernak, J: Results of Volume/ Occupancy Counts, After Study II, Spring1990.