Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
1
Effect of Removing Freeway Mainline Barrier Toll Plazas on Safety
Hong Yang, Ph.D. (Corresponding Author)
Post-Doctoral Associate
Rutgers Intelligent Transportation Systems (RITS) Laboratory
Department of Civil and Environmental Engineering
Rutgers, The State University of New Jersey
623 Bowser Road, Piscataway, NJ 08854, USA
Phone: (732) 445-0579 x127
Fax: (732) 445-0577
Email: yanghong@rci.rutgers.edu
Kaan Ozbay, Ph.D.
Professor & Director
Rutgers Intelligent Transportation Systems (RITS) Laboratory
Department of Civil and Environmental Engineering
Rutgers, The State University of New Jersey
623 Bowser Road, Piscataway, NJ 08854, USA
Phone: (732) 445-2792
Fax: (732) 445-0577
Email: kaan@rci.rutgers.edu
Bekir Bartin, Ph.D.
Research Associate
Rutgers Intelligent Transportation Systems (RITS) Laboratory
Department of Civil and Environmental Engineering
Rutgers, The State University of New Jersey
623 Bowser Road, Piscataway, NJ 08854, USA
Phone: (732) 445-5496
Fax: (732) 445-0577
Email: bbartin@rci.rutgers.edu
Ozgur Ozturk
Graduate Research Assistant
Rutgers Intelligent Transportation Systems (RITS) Laboratory
Department of Civil and Environmental Engineering
Rutgers, The State University of New Jersey
623 Bowser Road, Piscataway, NJ 08854, USA
Tel: (732) 445-5496, Fax: (732) 445-0577
Email: ozturko@eden.rutgers.edu
Word count: 4046 Texts + 5 Tables + 1 Figures = 5546
Abstract: 244
Initial Submission Date: August 1, 2012
Resubmission Date: November 15, 2012
Paper submitted for Publication in the
Transportation Research Record, Journal of Transportation Research Board after being presented
Transportation Research Board’s 92nd Annual Meeting, Washington, D.C., 2013
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2
ABSTRACT
Toll plaza safety is a critical issue. Toll plazas induce motor vehicle crashes and also put workers such as
toll collectors at risk. Therefore, enhancing safety at a toll plaza is crucial to improving safety on tolled
roadways. This study aims to evaluate the safety effect of removing mainline barrier toll plazas on
highways using Empirical Bayesian (EB) methodology. Recent removals of barrier toll plaza on the
Garden State Parkway in New Jersey were used as a case study. Multiple-year traffic and crash data
before and after the removals of the barrier toll plazas were used for analysis. Toll plaza crash frequency
models as a function of traffic flow and other factors were developed, with the modeling results
suggesting that there is a nonlinear relationship between toll plaza crash occurrences and both traffic flow
as well as toll booth configurations. The EB approach is also used to predict crash frequency assuming
that the barrier toll booths were not removed. These EB-based estimates were compared with the
observed number of crashes after the removals of the toll plazas. Individual comparisons show reductions
in crash frequency at almost all of the toll plazas and an estimated reduction of 47.2 percent overall at all
toll plazas due to the removal of the barrier toll booths. The estimated crash cost was reduced by 43.2
percent. These estimated reductions demonstrate that the removal of barrier toll plazas is a very beneficial
step towards improving safety of toll roads.
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
3
INTRODUCTION
Highway authorities have continued to use advanced technologies to improve efficiency at toll plazas. A
first step was the use of automatic coin machines (ACM), and later was the establishment of electronic
toll collection (ETC) systems. ETC enables non-stop toll charges via electronic transponders or other
forms of automatic vehicle identification (AVI). ETC is widely recognized as a successful Intelligent
Transportation Systems (ITS) application with numerous benefits such as lower transaction time,
improved throughput, and reduced air pollution and fuel consumption (e.g.,1-4). However, ETC systems
on traditional barrier toll plazas still require vehicles to slow down into channeled toll lanes, which itself
requires vehicles to make complex lane-choice decisions at relatively high speeds. Therefore, safety
issues at barrier toll plazas still exist despite increased throughput.
As stated in Brown et al. (5), even though toll plazas have been designed and constructed in the
United States for more than 50 years, there are no widely accepted design standards for toll plaza
uniformity or safety, with the only standards developed by individual toll operators based on their
experience (6-10). It is stated that there has been an increase in the number of toll roadways being built
with little information on how such barrier toll plazas affect crash rates (11). Similarly in Abdelwahab et
al. (12), the urgent need to study traffic safety issues of toll plazas is emphasized.
Removing existing toll plazas can be deemed safer as they eliminate complicate lane-choice
decisions as well as the barrier itself as an obstruction, which both increase the likelihood of crashes.
However, unlike easily measurable benefits such as reduction of transaction-related delays and pollution,
it is difficult to evaluate safety performance of removing the barrier toll plazas after a short period of
time. Few studies have briefly analyzed the safety problems created by the presence of toll plazas (13-15).
The main objective of this study is twofold:
(1) Model the crash occurrence at barrier toll plazas, and
(2) Evaluate the safety effects of removing the barrier plazas on toll roads.
The removals of mainline barrier toll plazas on the Garden State Parkway in New Jersey are used
as a case study. Crash models are developed to investigate the safety performance before and after the
removals of the mainline toll booths. The crash data available in this study cover the period from 2001 to
2010. Therefore, safety performance of removing the barrier toll plazas can be analyzed and compared
using the multi-year crash data (16).
REMOVAL OF MAINLINE TOLL PLAZAS ON FREEWAYS
The Gardens State Parkway (GSP) , a 172.4-mile limited-access tolled parkway with 359 exits and
entrances, is used as the case study to analyze the impacts of the removals of toll plazas on freeways. The
main reasons for selecting GSP as the case study are (1) GSP is a major and heavily congested limited
access highway; (2) 10 barrier toll plazas on the mainline have been removed since 2001; and (3) Before
and after crash data for GSP is available on-line (16) from the New Jersey Department of Transportation
(NJDOT). Over 380 million vehicles travel the GSP which stretches the length of New Jersey (NJ) from
the New York (NY) state line at Montvale to Cape May at the southern tip of the state. Tolls are collected
at 50 locations, including 11 mainline toll plazas and 39 entrance and exit ramps (17). It is among
America’s busiest highways, serving users from NJ and NY’s most marketable communities (18, 19).
A traditional low-speed ETC system was deployed on GSP in 1999. The entire system was
completed in August 2000 (19). The system has been widely adopted by travelers with an ETC
penetration rate beyond 70 percent (20).
In 2001, the state government issued an order to promote a ten-year congestion relief plan for
GSP (21). Under the plan, elimination of mainline barrier toll plazas in one direction and use of express
ETC lanes of open road tolling (ORT) in the other direction were recommended. By 2010, all mainline
barrier toll plazas but one were converted to one-way tolling (express ETC lanes were added to both
directions at the Toms River toll plaza). TABLE 1 summarizes the toll plaza removal date of each barrier
toll plaza. FIGURE 1 shows an example of the progression of removing toll collection at the Pascack
Valley northbound toll plaza. The original toll plaza [see FIGURE 1 (a)] was retrofitted to add two
express ETC lanes in each direction in 2004, shown in FIGURE 1 (b). Its remaining barrier toll booths
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
4
were removed in 2010 and the toll plaza becomes a standard roadway section without any toll collection
[see FIGURE 1 (c)]. Similar conversions were conducted at the other nine toll plazas in TABLE 1.
TABLE 1 Removal of the Mainline Barrier Toll Plazas on GSP
Toll Plaza
Milepost
Removal Date
Removed Toll booths
Current Lanes
Cape May SB
19.4
1/8/2006
5
2
Great Egg NB
28.8
1/8/2006
5
2
New Gretna SB
53.5
1/8/2006
6
2
Barnegat NB
68.9
3/10/2007
7
2
Asbury Park SB
104
9/12/2004
13
5
Raritan NB
125.4
9/12/2004
17
7
Union SB
142.7
3/6/2005
13
6
Essex NB
150.7
7/17/2005
8
4
Bergen SB
160.5
12/4/2005
8
4
Pascack Valley NB
166.1
2/20/2010
6
3
5
(a) Existing toll plaza configuration before 2004
SB
6
NB
(b) Adding Express ETC Lanes in 2004
SB
NB
7
(c) Removing Northbound Toll Plaza in 2010
SB
NB
8
9
10
11
12
13
14
15
FIGURE 1 Toll collection evolution at Pascack Valley Toll Plaza. (Source: Google Map & 22)
Removing those traditional barrier toll booths that have served more than fifty years on GSP
conceivably eliminates bottlenecks on the freeway. It can improve the highway capacity and reduce
emissions related to the heavy stop-and-go traffic at barrier toll booths. By realigning the plaza roadway
to allow for smooth traveling, it is deemed to avoid the safety pitfalls of conventional toll plazas.
However, the safety benefits of removing the barrier toll plazas have not yet been quantified.
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
5
SAFETY IMPACT ANALYSIS
Data Description
Several data were collected to investigate crash characteristics at the mainline toll plazas. The crash
records for toll plazas are obtained from the raw crash database of the New Jersey Department of
Transportation (NJDOT) (16). The records provide detailed information on each crash such as the time,
location, collision type, and severity. Generally, crash data are analyzed from sections of 0.5 miles or
longer (e.g., 23,24). Considering the signposting distances and physical lengths of the toll plazas on
GSP, toll plaza impact area is assumed to be one mile, which covers 0.5 mile before and after the toll
booths, respectively. Crashes that occurred within the impact area of each toll plaza were extracted from
the raw crash records. Data in the year when the toll plaza was removed were excluded for analysis. In
addition, the most recently removed toll plaza, Pascack Valley NB toll plaza listed in TABLE 1, was also
excluded for analysis. TABLE 2 presents the average number of crashes observed in each year before and
after the mainline barrier toll plazas were removed.
TABLE 2 Annual Crash Frequency Before-After Removing Toll Plazas
With Toll Plaza (Annual Observed Crashes)
Toll Plaza
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Without Toll Plaza (Annual Observed Crashes)
Period
All
PDO
Injury
Period
All
PDO
Injury
Cape May SB
2001-2005
6.8
5.4
1.4
2007-2010
3.0
2.0
1.0
Great Egg NB
2001-2005
8.2
7.4
0.8
2007-2010
6.0
5.0
1.0
New Gretna SB
2001-2005
15.2
12.2
3.0
2007-2010
5.3
3.5
1.8
Barnegat NB
2001-2006
22.3
19.0
3.3
2008-2010
23.3
15.3
8.0
Asbury Park SB
2001-2003
52.3
41.7
10.7
2005-2010
25.0
19.7
5.3
Raritan NB
2001-2003
129.0
101.0
28.0
2005-2010
65.8
51.7
14.2
Union SB
2001-2004
182.3
146.5
35.8
2006-2010
130.8
101.8
29.0
Essex NB
2001-2004
67.5
50.8
16.8
2006-2010
35.2
25.2
10.0
Bergen SB
2001-2004
46.8
34.5
12.3
2006-2010
48.2
39.8
8.4
The crash counts shown in TABLE 2 indicate that the crash counts at some toll plazas such as
Cape May SB were reduced whereas others such as Bergen SB were increased. However, the observed
crash counts before and after the removals of toll plazas should not be directly compared as many casual
factors have not been addressed. The GSP toll plaza configurations and annual average daily traffic
(AADT) data for the toll plazas between 2001 and 2010 were obtained from the New Jersey Turnpike
Authority (NJTA).
Empirical Bayesian Analysis
Safety assessment is a challenge task. Intuitively, safety effect/impact can be determined like previous
studies (25,26) by simply comparing the historical crash frequencies or rates before and after the
implementation of treatments at toll plazas:   A  B or   A B , where B is the reported crash
frequency or rate in the before period; A is the reported crash frequency or rate in the after period;  is
the change of crash frequency or rate; and  is the index of effectiveness. However, such simple analysis
is often unable to clarify the actual effectiveness of the treatment because of the natural change in safety
over time. In addition, it is undesirable to predict the index of effectiveness by merely using the observed
number of crashes as it does not account for the potential change in traffic volume. Besides, such a simple
before-after comparison cannot account for many issues such as regression-to-the-mean effect as well as
changes in other casual factors (27,28,29). Therefore better evaluation approaches are necessary to assess
the safety effect of the removing the mainline toll plazas.
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
6
The methodology adopted herein is the Empirical Bayesian (EB) method. The EB method
pioneered by Hauer (27) has been widely recognized as a very effective tool for various observational
before-after traffic safety studies (24,30-34). Instead of merely using observed crash counts or rates B in
the before period, the essential idea of EB procedure estimates the long-term safety of an entity using two
different sets of evidence, that is, the historical crash data of the entity and the predicted number of
crashes obtained from a safety performance function ( SPF ) for similar sites (24,35). Compared to other
simple before-after evaluation methods, the EB method has some distinct advantages. First and foremost,
the EB method offers improved estimates of the expected number of crashes  had the treatment not
been implemented in the after period by accounting for the regression-to-the-mean effect along with
changes in traffic volume and other factors (32,36). In addition, it is preferred as the need of a comparison
group is not compulsory (36,37,38). For before-after methods relying on a comparison group, it may be
difficult to obtain and appropriately select the comparison group, as discussed in (39,40).
The detailed EB procedure is not described in this study, but is well documented in literature
(e.g.,27,36,37,41). Only the overall index of effectiveness, its variance, and standard error for the
removals of the toll plazas are presented as follows:
Aa  a
1  Var  a   a2 
1 Aa  Var  a   a2 
2
Var     
2
1  Var  a   a2 

(1)
(2)
S .E.    Var  
(3)
where  is the overall index of effectiveness;  a is the expected toll number of crashes based on EB
approach had the toll plazas not been removed; Aa is the total number of crashes observed after the toll
plazas were removed; Var ( ) represents variance and S .E.( ) represents standard error of  .
The overall safety effectiveness as a percentage change of crash frequency across all removed toll
plazas is 100 1    . The 95 percent confidence interval (CI) for  is estimated as   1.96S.E.   . If   1
and its confidence interval does not contain 1, then the result implies that the removals of the toll plazas
have significantly improved the safety (in terms of crash reduction). In contrast, if   1 and the lower
limit of the confidence interval is greater than 1, then the removals of the toll plazas have a negative
impact on toll plaza safety (in terms of crash increase). Otherwise, the removals of the toll plazas have no
obvious impact on the safety performance of the toll plaza sections.
Data from thirteen other mainline toll plazas on GSP that have not been removed between 2001
and 2010 were used to develop the SPF . Specifically, negative binomial (NB) regression models for total
crashes, PDO crashes, and injury crashes (combined fatal crashes) were developed to link a set of
explanatory variables with crash counts at the toll plazas. The quantifiable variables considered in this
study are summarized in TABLE 3. It should be noted that some other variables such as roadway grade,
traffic composition and weather conditions may also affect the safety performance of the section.
Therefore, if data were available, these variables should be considered in the model. The final safety SPF
for predicting the number of crashes at toll plaza can be specified as follows:
ln(i )  0  ln( AADTi )1  ln( AZonei ) 2  3 ln( DZonei )  4 Boothi  5 ACM i  6CASHi  7 ExETCi  8Yeari
(4)
where i is the predicted number of crashes at ith toll plaza; 0 is a constant and  j ( j  1,2,...,8)
represents the coefficient of each explanatory variable.
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
TABLE 3 Variables Considered for the Safety Performance Functions
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
7
Variable
Symbol
Description
Number of crashes
Flow
Length of approach zone
Length of departure zone
Number of toll booths
Proportion of exact change toll booths
Proportion of cash toll booths
Proportion of express ETC lanes
Year
Y
AADT
Azone
Dzone
Booth
ACM
CASH
ExETC
Year
Annual crashes at a toll plaza (dependent variable)
Traffic flow (vehicles) per day per toll booth
Length of approach zone of a toll plaza (mile)
Length of departure zone of a toll plaza (mile)
Total number of toll booths at a toll plaza
Proportion of toll booths that primarily use exact change
Proportion of toll booths that primarily use cash
Proportion of toll booths that primarily use high-speed ETC
Order of year in study period to address time trend
RESULTS & DISCUSSION
The negative binomial (NB) regression models for toll plaza safety analysis were developed by crash
categories, with the estimated model parameters presented in TABLE 4.
The results show that toll plaza crash frequency is expected to increase with traffic volume. This
result is consistent with the common argument in traffic safety studies that more vehicles are likely to
cause more unsafe interactions and therefore more crashes (32,42). The length of the approach zone is
negatively associated with crash risk. In other words, the crashes can be reduced if a longer approach
zone is available. This may be attributable to the fact that the longer approach zone provides more time to
safely change lanes and select toll booths. Similarly, the length of the departure zone is also negatively
associated with crash occurrence. The more toll booths a toll plaza has, the more crashes are expected to
occur. The positive signs of the proportion of low speed toll booths (using cash and exact change
payment) indicate that these toll booths were all positively associated with the number of crashes at toll
plazas. This is expected as vehicles must slow down and pay tolls with cash at these toll booths. It causes
more stop-and-go traffic which in turn increases the crash risk. More importantly, the crash counts will
reduce if the proportion of express toll lanes is larger. This is consistent with our previous findings in (43)
that the use of high-speed toll lanes (or open road tolling) will significantly reduce the number of crashes.
In addition, there was no obvious time trend of crash frequency at the toll plazas. These SPFs are then
used to conduct EB estimation of crashes at the removed toll plazas listed in TABLE 2.
TABLE 4 Estimates of Coefficients for Safety Performance Functions
NB Model
Total Crashes
Variable
Coef.
Std.Err
P>|z|
Coef.
Std.Err.
P>|z|
Coef.
Std.Err
P>|z|
log(AADT)
1.768
0.132
0.000
1.789
0.141
0.000
1.965
0.206
0.000
log(Azone)
-0.728
0.157
0.000
-0.690
0.164
0.000
-1.088
0.224
0.000
log(Dzone)
-0.509
0.157
0.001
-0.488
0.162
0.003
-0.621
0.196
0.002
Booth
0.174
0.013
0.000
0.170
0.013
0.000
0.173
0.018
0.000
Cash
0.746
0.613
0.223
0.766
0.645
0.235
0.267
0.846
0.753
ACM
0.789
0.492
0.109
0.417
0.509
0.413
1.612
0.563
0.004
ExETC
-1.044
0.344
0.002
-1.072
0.359
0.003
-1.141
0.462
0.013
Year
0.017
0.015
0.245
0.020
0.015
0.195
0.005
0.018
0.793
Constant
-15.628
1.222
0.000
-15.786
1.300
0.000
-19.809
1.863
0.000
Dispersion
0.083
0.015
---
0.084
0.016
---
0.050
0.022
---
TRB 2013 Annual Meeting
PDO Crashes
Injury Crashes
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
8
TABLE 5 presents the results of the Empirical Bayesian (EB) before-after evaluation of removing
mainline barrier toll plazas on Garden State Parkway. The total number of observed crashes during the
period without the barrier toll plazas is compared with the number of crashes predicted had the toll plazas
not been removed. The results show that almost all toll plazas have a significant estimated reduction in
crash frequency after the toll plazas have been removed. The index of effectiveness  for Barnegat NB
toll plaza as well as Bergen SB toll plaza is relatively larger than the others. This indicates that the toll
plaza removals at these two sites were less effective than other sites. Detailed crash characteristics should
be examined to reveal the reasons why the projects at these two sites were less effective. The overall
index of effectiveness  for total crashes is 0.528 which implies a crash reduction factor of
100(1   )  47.2 percent. Similarly, the overall crash reduction factors for PDO crashes and injury crashes
are 49.0 percent and 40.3 percent, respectively.
TABLE 5 Impact of Toll Plaza Removal based on EB Prediction
Without Toll Plaza (observed)
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
With Toll Plaza (EB predicted)
Index of Effectiveness(  )
Toll Plaza
All
PDO
Injury
All
PDO
Injury
Overall
PDO
Injury
Cape May SB
12
8
4
28
22
5
0.424
0.350
0.761
Great Egg NB
24
20
4
42
38
10
0.559
0.514
0.390
New Gretna SB
21
14
7
57
46
11
0.361
0.299
0.610
Barnegat NB
70
46
24
78
72
14
0.892
0.632
1.718
Asbury Park SB
150
118
32
373
302
77
0.399
0.388
0.408
Raritan NB
395
310
85
933
742
173
0.422
0.417
0.486
Union SB
654
509
145
1093
891
205
0.598
0.570
0.704
Essex NB
176
126
50
386
296
90
0.455
0.423
0.547
Bergen SB
241
199
42
310
236
72
0.772
0.838
0.573
1743
1350
393
3300
2646
657
0.528 0.510
All Toll Plazas
Notes: (1) EB estimate is the expected number of crashes during after period had the toll plaza not been deployed;
(2)  is estimated index of effectiveness; and
(3) Statistically significant  is shown in blue number.
(4) Observation period for each toll plaza is listed in TABLE 2
0.597
To examine the economic impact of removing the barrier toll plazas in terms of safety cost,
benefit cost analysis is conducted. The average economic cost per crash ($8,900 for a PDO crash and
$70,200 for an injury crash) estimated by the National Safety Council is used for the analysis (44). The
benefit is estimated by production of crash cost with the number of reductions in PDO and injury crashes.
If the toll plazas were not removed, the estimated total crash cost is approximately $69.7 million during
the after period. In contrast, the estimated crash cost after the removals of the toll plazas is estimated at
only $39.6 million in total. Thus the overall crash cost reduction due to the removals of the barrier toll
plazas is estimated at 43.2 percent.
CONCLUSIONS
This study aimed to evaluate the safety effects of removing mainline barrier toll plazas on tolled
roadways. The removal of nine barrier toll plazas on the Garden State Parkway (GSP) in New Jersey
between 2001 and 2010 was used as the case study. Crash data, related traffic volume data, and lane
configurations for each toll plaza were collected to support the analysis. Crash modeling based on the
Empirical Bayesian (EB) method was conducted to determine the index of effectiveness of removing the
toll plazas. Results of the before-and-after comparisons show that the removals of these toll plazas
significantly improved the safety at these locations. The estimated results show that the overall crash
occurrence at these toll plazas was reduced by 47.2 percent after the removals of barrier toll plazas. The
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
9
estimated crash cost was reduced by 43.2 percent. The negative safety impacts of the barrier toll plazas
are identified as the mixed use of different toll booths at barrier toll plazas, which force drivers to make
more lane changes and travel at varying speeds at toll plazas. Frequent lane changes and large speed
variations in turn increase the likelihood of crash at these locations. It is thus concluded that the removals
of traditional barrier toll plazas can be a beneficial step towards significantly improving safety on the toll
roads.
Though the analyses focused on removing barrier toll plazas, the findings could be useful to
support some other treatments to enhance safety at toll plazas. For instance, if it is impossible to remove a
toll plaza, new tolling solutions like open road tolling (ORT) system can be used. Once toll booths are
converted to ORT lanes, the tolling function can be kept but the impact on traffic as well as safety can be
reduced because there would be no physical barrier toll booths to impede traffic.
DISCLAIMER
The contents of this paper reflect views of the authors who are responsible for the facts and accuracy of
the data presented herein. The contents of the paper do not necessarily reflect the official views or policies
of any agencies.
REFERENCES
1. Hensher, D.A. Electronic Toll Collection. Transportation Research Part A: General, Vol. 25, Issue
1, 1991, pp. 9–16.
2. Li, J., D. Gillen, and J. Dahlgren. Benefit-Cost Evaluation of the Electronic Toll Collection System:
A Comprehensive Framework and Application. Transportation Research Record: Journal of the
Transportation Research Board, No. 1659, Transportation Research Board of the National
Academies, Washington, D.C., 1999, pp. 31–38.
3. Coelho, M.C., T.L., Farias, and N.M. Rouphail. Measuring and Modeling Emission Effects for Toll
Facilities. In Transportation Research Record: Journal of the Transportation Research Board, No.
1941, Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 136–
144.
4. Venigalla, M. and M. Krimmer. Impact of Electronic Toll Collection and Electronic Screening on
Heavy-Duty Vehicle Emissions. In Transportation Research Record: Journal of the Transportation
Research Board, No. 1987, Transportation Research Board of the National Academies, Washington,
D.C., 2006, pp. 11–20.
5. Brown, L., D.R., McDonald, and E.J. Myers. Developing Traffic Control Strategies at Toll Plazas.
ITE Journal, November 2006, pp. 22–26.
6. Schaufler, A.E. NCHRP Synthesis of Highway Practice 240: Toll Plaza Design. Transportation
Research Board of the National Academies, Washington, D.C., 1997.
7. McDonald, D. and R. Stammer. Contribution to the Development of Guidelines for Toll Plaza
Design. Journal of Transportation Engineering, Vol. 127, No. 3, 2001, pp. 215–222.
8. Al-Deek, H., A. Mohamed, and A. Radwan. New Model for Evaluation of Traffic Operations at
Electronic Toll Collection Plazas. In Transportation Research Record: Journal of the
Transportation Research Board, No. 1710, Transportation Research Board of the National
Academies, Washington, D.C., 2000, pp. 1–10.
9. Stammer, R. and D. McDonald Tollway Sign and Pavement Markings-Recommended Design
Guidelines. Transportation Quarterly, Vol. 54, No. 3, 2000, pp. 67–84.
10. Dijk, N., M. Hermans, and M. Teunisse. Designing the Westerschelde Tunnel Toll Plaza Using a
Combination of Queuing and Simulation. In Proceeding of the Winter Simulation Conference,
Phoenix, Arizona, 1999.
11. Frisbie, T. American Highways Take a Toll. Traffic Safety, Vol. 91, 1991, pp. 22–25.
12. Abdelwahab, H.T. and M.A. Abdel-Aty. Artificial Neural Networks and Logit Models for Traffic
Safety Analysis of Toll Plazas. In Transportation Research Record: Journal of the Transportation
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
10
Research Board, No. 1784, Transportation Research Board of the National Academies, Washington,
D.C., 2002, pp. 115–125.
Ozbay, K., and A.M. Cochran. Safety Assessment of Barrier Toll Plazas. Advances in
Transportation Studies-An International Journal, Section A. Vol. 12, 2007, pp. 23-33.
Mohamed, A.A., M.A. Abdel-Aty, and J. Klodzinski. Safety Considerations in Designing Electronic
Toll Plazas: Case Study. ITE Journal, 2001, pp. 20–24.
Menta, V., H. Strate, D. Boss, and A. Saracena. Electronic Toll Collection and Safety at the Holland
Tunnel. In Conference of Traffic Congestion and Traffic Safety in the 21st Century: Challenges,
Innovations, and Opportunities. FHWA/NHTSA, U.S. Department of Transportation, Chicago, Ill.,
1997, pp. 236–242.
NJDOT. Crash Records. http://www.state.nj.us/transportation/refdata/accident/. Access April, 2012.
NJTA. http://www.state.nj.us/turnpike/who-we-are.html. Access April 2012.
Travelers’
Marke.
Garden
State
Parkway.
Available
online
at:
http://www.travelersmarketing.com/NewJersey.htm. Access July 2011.
Currie, J., and Walker, R. (2011). Traffic Congestion and Infant Health: Evidence from E-ZPass.
American Economic Journal: Applied Economics, Vol. 3, No. 1, pp. 65–90.
NJTA. Financial Summary for the Five Months Ended May 31, 2011.
http://www.state.nj.us/turnpike/documents/MonthlyFinancialMay2011.pdf. Access June 2011.
Weinstein, J. Garden State Parkway Congestion Relief Plan - A Proposed Framework. Department
of Transportation, New Jersey, 2001.
Raczynski, R. and J. Finn. Garden State Parkway Plaza Improvement Program. Presented at IBTTA
Facilities Management Workshop, Toronto, Ontario, May 14-18, 2005.
Resende, P. T. V. and Benekohal, R. F. (1997). Effects of Roadway Section Length on Accident
Modeling. Proceedings of the Conference-Traffic Congestion and Traffic Safety in the 21st
Century: Challenges, Innovations, and Opportunities, Chicago, IL, June 8-11, pp. 403-409.
Abdel-Aty, M., P.C. Devarasetty, and A. Pande. Safety Evaluation of Multilane Arterials in Florida.
Accident Analysis & Prevention, Vol. 41, Issue 4, 2009, pp. 777–788.
Klodzinski, J., E. Gordin, H.M. Al-Deek. Evaluation of Impacts of Open road Tolling on Main-Line
Toll Plaza. In Transportation Research Record: Journal of the Transportation Research Board No.
2012, Transportation Research Board of the National Academies, Washington, D.C., 2007, pp. 72–
83.
Gordin, E., J. Klodzinski, and C.D. Santos. Safety Benefits from Deployment of Open Road Tolling
for Main-Line Toll Plazas in Florida. In Transportation Research Record: Journal of the
Transportation Research Board No. 2229, Transportation Research Board of the National
Academies, Washington, D.C., 2011, pp. 85–92.
Hauer, E. Observational Before-After Studies in Road Safety: Estimating the Effect of Highway and
Traffic Engineering Measures on Road Safety. Pergamon Press, Elsevier Science, Ltd., Oxford,
United Kingdom, 1997.
Elvik, R. The Importance of Confounding in Observational Before-and-After Studies of Road
Safety Measures. Accident Analysis & Prevention, Vol. 34, Issue 5, 2002, pp. 631–635.
Shen, J. and A. Gan. Development of Crash Reduction Factors: Methods, Problems, and Research
Needs. In Transportation Research Record: Journal of the Transportation Research Board, No.
1840, Transportation Research Board of the National Academies, Washington, D.C., 2003, pp. 50–
56.
Persaud, B., E. Hauer, R. Retting, R. Vallurupalli, and K. Mucsi. Crash Reductions Related to
Rraffic Signal Removal in Philadelphia. Accident Analysis & Prevention, Vol. 29, Issue 6, 1997, pp.
803–810.
Elvik, R.,. The Predictive Validity of Empirical Bayes Estimates of Road Safety. Accident Analysis
& Prevention, Vol. 40, Issue 6, 2008, pp. 1964–1969.
Monsere, C.M. and E.L. Fischer. Safety Effects of Reducing Freeway Illumination for Energy
Conservation. Accident Analysis & Prevention, Vol. 40, Issue 5, 2008, pp. 1773–1780.
TRB 2013 Annual Meeting
Paper revised from original submittal.
Yang, Ozbay, Bartin, and Ozturk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
11
33. Shin, K., S.P. Washington, and I. Van Schalkwyk. Evaluation of the Scottsdale Loop 101
Automated Speed Enforcement Demonstration Program. Accident Analysis & Prevention, Vol. 41,
Issue 3, 2009, 393–403.
34. Persaud, B., B. Lan, C. Lyon, and R. Bhim. Comparison of Empirical Bayes and Full Bayes
Approaches for Before–After Road Safety Evaluations. Accident Analysis & Prevention, Vol. 42,
Issue 1, 2010, pp. 38–43.
35. Wen, C. and S.P. Washington. Experimental Evaluation of Hotspot Identification Methods. Accident
Analysis & Prevention, Vol. 37, Issue 5, 2005, pp. 870–881.
36. Hauer, E., D.W. Harwood, F.M. Council, and M.S. Griffith. Estimating Safety by the Empirical
Bayes Method: A Tutorial. In Transportation Research Record: Journal of the Transportation
Research Board, No. 1784, Transportation Research Board of the National Academies, Washington,
D.C., 2002, pp. 126–131.
37. Highway Safety Manual (HSM), 1st edition. American Association of State Highway and
Transportation Officials (AASHTO), Washington, D.C., 2010.
38. Elvik, R. Analytic Choices in Road Safety Evaluation: Exploring Second-Best Approaches.
Accident Analysis & Prevention, Vol. 45 (March), 2012, pp. 173–179.
39. Hauer, E. Comparison Groups in Road Safety Studies: An Analysis. Accident Analysis &
Prevention, Vol. 23, Issue 6, 1991, pp. 609–622.
40. Hauer, E., J.C.N. Ng, and P. Papioannou. Prediction in Road Sfety Studies: An Empirical Inquiry.
Accident Analysis & Prevention, Vol. 23, Issue 6, 1991, pp. 595–607.
41. Persaud, B. and C. Lyon. Empirical bayes before–after safety studies: Lessons learned from two
decades of experience and future directions. Accident Analysis & Prevention, Vol. 39, Issue 3, 2007,
pp. 546–555.
42. El-Basyouny, K. and T. Sayed. Accident Prediction Models with Random Corridor Parameters.
Accident Analysis & Prevention, Vol. 41, Issue 5, 2009, pp. 1118–1123.
43. Yang, H., K. Ozbay, and B. Bartin. Effects of the Open Road Tolling on Safety Performance of
Freeway Mainline Toll Plazas. In Transportation Research Record: Journal of Transportation
Research Board, Transportation Research Board of the National Academies, Washington, D.C.,
forthcoming, 2012.
44. National Safety Council. Estimating the Costs of Unintentional Injuries, 2010.
http://www.nsc.org/news_resources/injury_and_death_statistics/Pages/EstimatingtheCostsofUninte
ntionalInjuries.aspx. Accessed May 15, 2012.
TRB 2013 Annual Meeting
Paper revised from original submittal.