Copyedit A

Formatted: Underline
Variable Speed Limits Signs: Effects oOn
Speed aAnd Speed Variation iIn Work Zones
Commented [SO1]: Note to the author: Tables and spacing
changes were made silently so as not to overwhelm the document
with correction marks.
Abstracts
Variable Speed Limit (VSL) signs are used across the country to lower posted speed
limits in certain areas. They allow for operators to adjust the posted speed limit without
changing the physical sign. They are used in conjunction with Intelligent Transportation
Systems to lower speed limits for several reasons including: congestion, construction,
accidents, and weather fog, snow, and ice. As technology advances, the ease of using
VSL signs is also increasing, as since the speeds can now be changed remotely via email
or telephone, at pre-set times of day, or manually.
The Utah Department of Transportation desired to evaluate the effectiveness of this

technology on the driver behavior before it uses the technology widely in work zones on
state highways. The primary focus of this research was geared towards driver response to
VSL sign usage. A single test site consisting of a relatively long duration and long
distance work zone on I-80 north of Wanship, Utah, was used to test the response of
drivers to speed limits presented by VSL signs.
The speed data analyses showed that the driver response was very positive. Both the
average speed and the variation in speeds were reduced by providing drivers variable
speed limit information which reflects the construction conditions in the field. The net
Formatted: Indent: First line: 0"
result of lower speed variations, as well as lower speeds where conditions require, should
result in improved safety to highway users and construction personnel. Long term use of
VSL signs is recommended for further application in Utah.
Introduction
The Variable Speed Limit (VSL) signs are used across the country to lower posted speed
limits in certain areas (see Figure 1 for the VSL sign used in the study). They allow for
operators to adjust the posted speed limit without changing the sign. They are used in
conjunction with Intelligent Transportation Systems to lower speed limits for several
reasons including congestion, construction, accidents, fog, snow, and ice.
Utah Department of Transportation (UDOT) wanted to test the use of VSL signs in
conjunction with a construction project. Long construction zones often have actual
construction occurring in only short segments of the zone. In addition, construction
zones are signed using static speed limit signs , as if construction activity occurs
throughout the entire day on each and every day. In reality,even though the intensity,
duration, and location of construction activity varies throughout the construction zone
depending on the extent of work being performed. Providing more accurate and variable
speed restrictions based on the level of construction activity provides more accurate
information to drivers and can help drivers comply with the posted regulatory speed
while improving safety in the construction zone.
Reducing vehicle speeds in construction zones improves safety for drivers and
construction workers. andIt can also reduce the severity of crashes which occur in the
Commented [SO2]: Delete. This is word for word from the

abstract.
zone. It is often said that crashes involving vehicles traveling at high speed are more
severe than those at low speed. (1) R. Reducing the speed of individual drivers is
Commented [SO3]: What is this referring to?

Formatted: Font: Not Italic
important; however, reducing the variation of speed of all drivers in a section of roadway
has an even greater safety enhancing effect. (1). Reducing speed deviation amongst
vehicles also contributes to increased safety. Also, iIt is reported that regardless of the
average speed on the highway, the more a vehicle deviates from the average speed, the
greater are its chances of becoming involved in a crash. (1).
The purpose of this study is to investigate the use of VSL signs for application in the
State of Utah. This study attempts` to answer the question, Would the use of VSL signs
better control the variability of vehicle speeds by providing drivers with more accurate
information, particularly during construction periods?
This paper presents a brief literature review, study methodology, study findings, and
conclusions and recommendations.
Literature Review
In the, tTwo VSL signs were set with speed limits based on a pre-determined schedule
varying by week and/or by day/night time. In other studies, VSL signs utilize traffic
speed and volume detection, weather information, and road surface condition technology
to determine appropriate speeds for drivers, and display them on overhead or roadside
variable message signs.
Washington DOT has a VSL system on a mountain pass on I-90 which responds to
weather conditions to advise motorists of safe traveling speed limits. And, they have
instituted a VSL system on a 23 mile section of US-2 which alters speed due to road
conditions. Washington DOT also uses VSL for work zones, for example on I-5 at the
SR 502 Interchange construction zone speed limits were changed from 70 mph to 60 mph
Commented [SO4]: Delete. All of the uses of VSL are
mentioned in the following paragraph.
using VSL.
Many other transportation entities have utilized this technology to reduce speeds on
roads. VSL has been used for accidents, congestion, construction, and weather. VSL
systems have also been used to determine safe speed limits for day versus night. VSL
technology can further be used to weigh downhill travelling vehicles and advise trucks to
alter their speed. This has successfully reduced truck-related accidents even when truck
traffic increases. On the New Jersey Turnpike 120 signs collect speed and volume data
and display variable speed limits based on accidents, congestion, construction or weather
(Reference #). New Mexico had a VSL system on an urban section of I-40 from 1989 to
1998 which used traffic speed, day versus night and precipitation to determine safe speed
limits (Reference #). In Colorado on I-70 at the Eisenhower Tunnel, VSL technology
weighs downhill traveling vehicles and advises trucks to alter their speed. It is reported
that since system deployment, truck-related accidents have declined on the steep downhill
grade sections while the volume of truck traffic has increased by an average of 5 percent
per year (Reference #). Oregon also uses a similar system for downhill trucks on I-84
(Reference #). Minnesota used VSL in work zones on I-494 in the Twin Cities area to
decrease upstream traffic speed from 65 mph to 45 mph when drivers approach the work
zone bottleneck. There are also numerous examples of non-US countries using this
Commented [SO5]: Summarize this paragraph? Revision okay?
technology (Reference #).
Commented [SO6]: Delete this part if the summary is okay.
Some of the documented VSL usage appears to be aimed at increasing vehicles
Formatted: No underline
speed by giving drivers confidence that faster speeds are safe in conditions where they
expect multiple slow-downs due to construction and congestion. A VSL test case was
undertaken on I-96 in Michigan. In this study, VSL was signs were used to improve
traffic flow through a work zone by providing more realistic speed limits which would
enable vehicles to travel at increased speeds and minimize delay. The regulatory speed
limit in the work zone was 50 mph versus 70 mph normally. Their study found that the
average speed of motorists did increase, and travel time decreased when the VSL was
operating.
VSL signs have been successfully tested in the past in various applications as
discussed in this brief literature review. Hence, UDOT desired to test the effectiveness of
VSL signs in helping drivers be more compliant with reduced speed limits in work zones,
thus protecting both the motorists and the construction workers in work zones.
Formatted: Font: Bold, Underline
Study Methodology
Formatted: UDOT body text, Space Before: 6 pt

Hypothesis Tested
Study Site Location and Construction Project Description

The study site was located on I-80 in Summit County near the Utah-Wyoming Border.
This study site was chosen because the section became an active construction area just
prior to the data collection. This study site had a pavement crack and seal work, from
mile post 185 to 191, with lane closures in both the eastbound and westbound directions.
Travel in both directions was restricted to one lane through the work zone at all hours
through the completion of the project, from summer to fall 2007. For this work zone, only
westbound traffic was used due to steep continual upgrades on the eastbound lanes,
slowing the traffic significantly.
The size of the Speed Limit sign was of 30 inches by 36 inches following the MUTCD
standard and the size of each digit (two digits in the display) was 18 inches high by 12
inches wide.
Commented [SO7]: Moved from section below to fit better with

project description.
Figure 2-1 Variable Sspeed Eequipment Uused in the Sstudy.
Formatted: Left
(Photo by Thomas McMurtry, July 2007)
Formatted: Left, Indent: First line: 0"
Data Collection
Vehicle speed data were collected at five locations on westbound I-80 for all hours of the
day. The data were collected from July 10 , 2007 through September 29, in 2007. Data
were collected by using Jamar tube counters. This study involved an extensive amount of
data collection. The five locations were chosen to give a representative spectrum of the
traveling speeds under multiple conditions of the construction zone. Figure 2-2 shows a
map of the study section with the speed counter locations and VLS sign locations.
The first counter was placed upstream of the construction zone to capture vehicles
under normal conditions. The second counter was placed at the beginning of the
construction zone just after the roadway was tapered down to one lane. The third and
fourth counters were placed through the construction zone as shown in Figure 2-2, and
were placed about a mile and a half apart. The fifth counter was placed near the end of
the construction zone where there was still only one lane open, but not in active
construction.
Figure 2-2 Location of the VSL Ssigns and Ssensors locations
Formatted: Font: 12 pt, Bold, Underline

Commented [SO8]: Chart made smaller to fit better.

Source: GIS data from Utah Automated Geographic Reference Center (AGRC)
Traffic counting tubes were used due to accuracy of defined speed bins, although
maintenance requirements were high. There were several occasions when the tubes were
9
removed for paving or were torn out by traffic. The data missed during those times were
anticipated and the data collection period was set at 12 weeks to account for losses and to
accurately count all roadway scenarios.
For the purpose of this study, two VSL sign conditions were tested against the existing
static sign condition. Thus, data were collected under three different signing conditions
listed below.
Standard 65 MPH speed limit signs
VSL sign posted at 65 MPH 24 hours per day, 7 days per week
VSL sign varying between 55 MPH during the day and 65 MPH at night
Data were collected for the standard 65 MPH sign for the first three weeks of the testing
period. Following that, the two VSL conditions were alternated in two week blocks. The
schedule of the data collection under the different signing conditions is displayed in the
calendar in Figure 2-3.
Figure 2-3 Data Ccollection Sschedule.
10
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Commented [SO9]: Instead of listing the description ON the

calendar with its color, it might be better to make a key. That will
keep all of the boxes the same height.
Saturday
JULY
CONSTRUCTION
ON-GOING
STANDARD
SIGN 65 MPH
24/7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
AUGUST
29
30
VSL 65 MPH
24/7
10
11
12
VSL
65 MPH
NIGHT
55 MPH DAY
14
15
16
17
18
19
20
21
22
23
24
25
VSL 65 MPH
24/7
27
28
30
31
SEPTEMBER
29
VSL
65 MPH
NIGHT
55 MPH DAY
11
12
13
14
15
16
17
18
19
20
21
VSL 65 MPH
24/7
23
24
25
26
27
28
29
The signing conditions are color coded in the schedule. Days where standard 65 MPH
speed limit signs were used are shown in light blue. Days where VSL signs posted at 65
MPH were used are shown in yellow. Days where VSL signs varied between 55 MPH
during the construction periods of the day and 65 MPH at night are shown in pink.
11
Commented [SO10]: This is pretty clear in the calendar already.

Turn this into a key and delete it from the calendar.
OFive tube counters collecting speed data for each vehicle that traveled through the
construction zone for 12 weeks, or approximately 80 days, generated a great deal of data.
Commented [SO11]: Delete this sentence.
Overall, almost three million data points were collected and analyzed.
The data were collected in five mile per hour5 mph intervals for each hour of the day
throughout the duration of the study. Analysis could be performed, therefore, to answer
the specific question, for example, How many vehicles traveled between 55 to 59 MPH
on Tuesday from 9 AM to 10 AM at counter number 3? Data indicating the type of
vehicle, truck, trailer, RV, motorcycle, etc. were not collected.
Periods where data were not collected, generally due to equipment mal-function when
tubes were stripped off the pavement, occur randomly by date, by time of day and by
Commented [SO12]: Moved from Data points because it talks
about data collection.
counter.
Commented [SO13]: Is this relevant? Delete?
Study Findings

Formatted: UDOT HEADING 1, Space Before: 6 pt
Data Points
Speed Distribution
Formatted: UDOT HEADING 1, Space Before: 6 pt
The distribution of speeds observed at various counters in general followed an expected

bell-shaped curve; however, speed distribution was more coherent with VSL speed limit
signs than with regular static speed limit signs. Figure 4 shows the difference between the
two methods. It should be noted that these show speed distributions of typical one- day
data. Figure 2-4 (a) shows speed distributions at the five counters on Thursday, July 12,
2007, when the speed limit was presented by regular static speed limit signs; Figure 2-4
(b) shows speed distributions at the five counters on Thursday, August 30, 2007.
12
Figure 2-4 Comparison of Sspeed Ddata Ddistribution Uunder Ddifferent Sspeed
Llimit Ppresentation.
4500
4000
Number of Vehicles
3500
3000
Counter 1
Counter 2
Counter 3
Counter 4
Counter 5
Formatted: Left
2500
2000
1500
1000
500
0
2529
3034
3539
4044
4549
5054
5559
6064
6569
7074
7579
8084
8589
90999
MPH Bins
(a) Thursday, July 12, 2007: Static 65 MPH sign all day
Formatted: Justified, Indent: First line: 0"
4000
Counter 1
3500
Counter 2
Counter 3
Number of Vehicles
3000
Counter 4
2500
Counter 5
2000
1500
1000
500
0
2529
3034
3539
4044
4549
5054
5559
6064
MPH Bins
13
6569
7074
7579
8084
8589
90999
(b) Thursday, August 30, 2007: VSL sign 65 MPH all day
As seen in these two plots, with VSL speed limit signs, there were two distinct speed
groups: one bell-shaped speed distribution for the area upstream of the work zone at
counter 1 and the other bell-shaped distributions for the counters within the work zone.
However, the speed distributions at the five counters were not so distinct with regular
static speed limit signs. VSL signs seem to encourage drivers to maintain their speed
close to the speed limit.
Standard Deviation of the Data
Examination of the daily standard deviation of speed showed that for the most part,
Counter 1 has greater standard deviation than the other counters as shown in Figure 2-5
(a). On July 31, the variable speed signs were installed. After installation of the variable
speed signs, for the 50 days with complete data, Counter 1 has the highest standard
deviation: 31 days, or 62% of the time. During the period between August 13 and 20 and
during the period of September 10 through 20, Counter 1s standard deviation is lower
than the other counters. This may be because construction on those days was such that
traffic was frequently interrupted, leading to greater deviation in speed.
Figure 2-5 Standard Ddeviation of Sspeed.
14
(a) Standard deviation of speed during the entire data collection period
The pattern of construction slow-downs during the day could not be separated from the
data in any methodical way. The existence of wide speed fluctuations during heavy
construction created greater standard deviation in the data than expected. This study
sought to compare the variability of speeds under different speed limit sign conditions.
Therefore, to examine speed variation in the absence of construction interruptions, the
nighttime, from 7pm to 6am, standard deviation was analyzed.
Nighttime standard deviation was highest at Counter 1 when VSL signs were present as
shown in Figure 2-5 (b). Of the 22 days with complete data from August 13 through
September 9, Counter 1 has the highest standard deviation for 16 days, or 70%. This is a
slightly higher percentage of occurrences than found was found when measuring the full
day periods.
15
(b) Night time standard deviation of speed

Before the installation of VSL signs, Counters 2 through 5 had speeds whose daily
standard deviation ranged from 4.3 to 14. In these locations, after variable signs are were
installed, the highest daily speed standard deviation was 7.9. The graph also
demonstrates that during the VSL sign period set to 65 mph, standard deviation for
Counters 2 through 5 was higher than 6 mph on only 3 occurrences. This low standard
deviation of average speeds indicates that few vehicles are moving either much faster or
slower than other vehicles. The majority of vehicles are traveling at the speeds very
similar to each other.
Average Speed
Daily average speeds at Counters 2 through 5 are less than Counter 1 for the entire study
period except a short period in September as shown in Figure 2-6. Slow speeds at
16
Counter 1 may be due to congestion just upstream of the work zone. Counter 5
experienced very high average speeds during that period, and other counters exhibit
average speeds lower than other periods. Increased speeds at Counter 5 may reflect
vehicles speeding up significantly to make up for time lost in congestion.
Figure 2-6 Daytime Aaverage sSpeed July 10 to September 28.
Nighttime Average Speed
Formatted: Font: Bold, Not Italic, Underline
In order to evaluate the differences in the effect of VSL signs without the effects of
construction work, average speeds at night hours from 7:00 PM to 8:00 AM of the next
day at the five counter locations were plotted. With static 65 mph reduced speed work
zone signs, average speeds vary as shown in Figure 2-7 (a). The standard deviation of the
mean speeds at each counter location varied from about 1.5 mph to 5.0 mph. Ten samples
were obtained to evaluate the effectiveness of static speed limit signs. The static limit at
17
Counter 1 was 75 mph, while at Counters 2 through 5 it was posted as 65 mph. As you
seen in Figure 2-7 (a), Counter 1 recorded average speeds between 70 to 75 mph.
Between Counter 2 and 4, there is a general trend of decreasing average speeds in the
work zone; however, at Counter 5, speeds actually seemed to increase. Figure 2-7 (a)
shows that there were wide variations of average speeds during the nighttime with static
signage as compared to the VSL signs.
Figure 2-7 Night time Aaverage sSpeed Ccomparison.
July
July
July
July
July
July
July
July
July
July
75.0
70.0
65.0
60.0
10
11
13
14
15
23
24
25
26
28
Counter 5
Counter 4
Counter 3
50.0
Counter 2
55.0
Counter 1
Average Speed, mph
80.0
(a) Nighttime average speed July 10 to July 28
Formatted: Font: Not Bold

With VSL signs displaying 65 mph, reductions in the average speeds are consistent at
each location as shown in Figure 2-7 (b). The standard deviation of the means at each
counter location varied from about 0.5 mph to 1.0 mph. This means that at night, drivers
are more compliant to the reduced speed limit with VSL signs than with static work zone
speed limit signs.
18
80.0
Average Speed, mph
75.0
July 31
70.0
August 1
August 2
65.0
August 3
60.0
August 4
August 5
55.0
Counter 5
Counter 4
Counter 3
Counter 2
Counter 1
50.0
(b) Nighttime average speed July 31 to August 5

Daytime Average Speed
During the day from 9:00 AM to 7:00 PM, it seems that average speeds are affected by
the magnitude of work underway in the work zone as shown in Figures 2-8 (a) and 2-8
(b). Some of the days show that average speeds were significantly lower than the other
samples. Figure 2-8 (a) shows the trend in average speeds at each counter location with
the static work zone speed limit sign and Figure 2-8 (b) shows the trend in average speeds
at each location with VSL signs. In general, speed changes are more consistent with VSL
signs than with the static work zone speed limit signs.
19
Figure 2-8 Daytime Aaverage Sspeed Comparison
70.0
80.0
comparison.
July 12
60.0
July 16
July 20
July 21
50.0
July 24
July 26
July 28
July 29
40.0
Average Speed, mph
July 14
30.0
July 30
Counter 1
Counter 2
Counter 3
Counter 4
Counter 5
Formatted: Caption
(a) Daytime average speed July 12 to 30, 65 MPH regular, static speed limit sign
80.0
August 2
August 3
Average Speed, mph
75.0
August 4
August 5
70.0
August 27
65.0
August 29
August 31
60.0
September 1
September 2
55.0
September 3
50.0
September 5
September 8
45.0
September 9
September 27
Counter 5
Counter 4
Counter 3
Counter 2
Counter 1
40.0
(b) Daytime average speed during 65 MPH VSL speed limit sign
Figure 2-8 (c) shows the trend in speed reduction with VLS signs displaying a 55 mph
speed limit. Because the static work zone speed limit signs showing 55 mph speed limit
was never shown to drivers, it was infeasible to compare the benefit of VSLS signs over
20
static speed limit signs at the 55 mph speed limit level. Whats obvious is that with a 55
mph speed limit, drivers tend to speed up at the downstream end of the work zone as
demonstrated by the average speeds at Counter 5. Four samples started with 75 mph at
the upstream of the work zone and their speeds decreased to 50 to 60 mph and as shown
in Figure 8 (c);, drivers maintained that speed level until they pass the active work zone.
One day (September 12) started with a lower speed (53 mph) upstream of the work zone.
It is likely a slow moving queue was forming upstream, because drivers seemed to
become impatient to the slow traffic and they began to speed up at Counter 3 and near
70.0
August 13
August 15
60.0
August 21
August 22
August 25
September 11
September 12
50.0
Counter 5
Counter 4
Counter 3
Counter 2
September 19
Counter 1
40.0
Average Speed, mph
80.0
Counter 5 they sped up to 77 mph (see Figure 2-8 (c)).
(c) Daytime average speed during 55 MPH speed limit sign

Selected Period Comparison
Due to missing data in early August and late September, three periods with complete data
are examined. The first, from July 10 through July 30, has static signs; the second,
August 13 through 25, has variable signs set to 55 mph during the day and 65 mph at
21
night,; and the third, August 27 through September 9, has variable signs set to 65 mph.
Figure 2-9 shows the plots daily average speeds for these three periods.
Figure 2-9 Selected Pperiods, Ddaily aAverage Sspeed.
85
8/13/07-8/25/07
7/10/07-7/30/07
8/17/07-9/9/07
80
Daily Average Speed
75
70
65
60
55
Counter
Counter
Counter
Counter
Counter
50
45
1
2
3
4
5
40
Daily July 10 thru Septem ber 9, 2007
The daily average speed , by day for each counter location indicates that the location at
Counter 1 experienced the highest average speeds. During the first period, when only the
static speed limit signs were used, daily average speeds ranged from 45 to 73 mph at
Counters 2 through 5 which were marked for 65 mph speed limit. During the second
period, when VSL signs limited speed to 55 mph daytime and 65 nighttime, average
speeds at these counters ranged from 46 to 62 mph. . During the third period, when
Counters 2 through 5 followed VSL signs of 65 mph, average speeds ranged from 55 to
69 mph.
22
Daily average speeds at Counter 1, located before the construction zone and signed
throughout the test period by a static 75 mph sign, did not vary by more than 8 mph.
Commented [SO14]: Which one? All threetest periods?
Their lowest daily average occurred in July at 70 mph and the highest occurred in August
at 78 mph.
Due to construction activity in the work zone, at some times, hourly speed slowed to
congested traffic levels in the 10 to 30 mph range. The construction log kept at the test
site, shows paving, equipment movement, and other interruptions to the single lane open
for travel. For example, on August 16, paving in the vicinity of Counter 3 occurred and
is reflected in the data as shown in Figure 2-10. On some heavy travel days, particularly
on Sunday afternoons, speed reduction can be observed in the data at all counters, even at
Counter 1 upstream of the construction zone.
Figure 2-10 Average Sspeed for Ccounter 3, August 16.
23
Average Speed by Hour (Counter 3 August 16th)

70
65
65 MPH Speed
Limit
65 MPH Speed
Limit
60
55
55 MPH Speed
Limit
MPH
50
45
40
35
30
25
12
:0
0
AM
01
:0
0
AM
02
:0
0
AM
03
:0
0
AM
04
:0
0
AM
05
:0
0
AM
06
:0
0
AM
07
:0
0
AM
08
:0
0
AM
09
:0
0
AM
10
:0
0
AM
11
:0
0
AM
12
:0
0
PM
01
:0
0
PM
02
:0
0
PM
03
:0
0
PM
04
:0
0
PM
05
:0
0
PM
06
:0
0
PM
07
:0
0
PM
08
:0
0
PM
09
:0
0
PM
10
:0
0
PM
11
:0
0
PM
20
Statistical Analysis on Speed Change from Nighttime 65 MPH to Daytime 55 MPH
The previous discussions on speed reduction from the 75 mph speed limit toto the 65
mph speed limit on the approach to the work zone indicated that both speed reduction
signs (static 65 mph vs. VSL 65 mph) could achieve necessary speed reductions although
VLS reduces speed variances, thus contributing to stabilizing speed variation. Though not
presented in this paper due to space limitation, these speed reductions were statistically
significant at a 95% confidence level. In this section, some of the statistical analyses on
speed changes are presented.
When speed reduction is large, it may be possible that drivers may comply; hence, z-tests
were performed to see if the drivers are complying with daytime 55 mph speed limit in
the work zone. Several days had complete data sets to analyze the conditions of speed
transition from nighttime 65 MPH speed limit to daytime 55 MPH speed limit, indicated
24
by VSL signs. Table 2-1 (a) presents the results of z-tests on speed differences between
the two speed limits. As shown in the table, mean speed differences between the
nighttime 65 mph and the daytime 55 mph speed limit were significant at these seven
sample dates, indicating that , indicating drivers do respond to large speed reduction.
Table 2-1 Statistics on Speed Reduction by VSL Signs, from Nighttime 65 MPH to
Daytime 55 MPH Speed Limit
(a) Z-Tests on Mean Speed Differences at Counter 2 (Critical Z-score = 1.65)
Mean Speed
Calculated
Significance
Difference
Z-Value
Test
7.2
36.3
Significant
11.8
61.00
Significant
6.1
34.29
Significant
3.1
19.33
Significant
7.2
46.01
Significant
9/12/2007 Daytime 55 MPH
10.8
54.63
Significant
9/12/2007 Nighttime 65 MPH to
7.1
47.80

8/14/2007 Z Daytime 55 MPH
25
Significant
In order to find out whether drivers comply with the daytime 55 mph speed limit
throughout the work zone, a data set for the 55 mph speed limit with VSL signs was
extracted and analyzed. Table 2-1 (b) shows the results of z-tests. As shown in the table,
in the upstream of the work zone at Counter 1, the average speed was 76.2 MPHmph,
statistically higher than 75 mph speed limit. Right after the first VSL sign the average
speed lowered to 58 MPHmph. Although the difference between this value and the 55
mph speed limit was significant, the sign did help reduce the average speed. At the
remaining counters, the mean speed was close to 55 mph, although the difference
between those speeds and the 55 mph speed limit was statistically significant.
Nevertheless, drivers seemed to comply with the 55 mph speed limit throughout the work
zone.
(b) Z-Tests on Mean Speeds with Daytime 55 MPH Speed Limit for Work Zone by VSL
Sign, Tuesday, 8/12/2007, between 9:00 AM and 7:00 PM (Critical Z-Score = 1.96)
Mean
Standard
95%
Counter
Number of
Speed
Deviation
Confidence
Number
Samples
(MPH)
(MPH)
Interval
Remarks
Statistically
4,689
6.6
76.0 76.4
higher than 75
76.2
mph
58.0
Statistically
26
4,921
5.6
57.9 58.2
higher than 55
mph
Statistically
4,916
3
5.8
56.2 56.5
56.4
higher than 55
mph
Statistically lower
4,919
54.6
8.0
54.3 54.8
than 55 mph
Statistically
4,974
5
4.9
56.5 56.7
56.6
higher than 55
mph
Note: Counter 1 is in the 75 MPH speed limit zone and the other counters are in the 55
mph speed limit zone. When the confidence interval does not contain 55 mph, the actual
mean speed is either statistically higher or lower than 55 mph.
Speed Limit Compliance
If the average speed of travel is taken, for each hour, at each data collection point, for the
entire length of the study, you can compare the observed speed can be compared to the
posted speed limit. Figure 2-11 shows, by counter and by the type of posted speed limit
(S for Static or V for Variable), the percentage of vehicles obeying the speed limit. At
Counters 2, 3, and 5, the VSL at 65 MPH mph has greater compliance than the Static 65
MPH mph signs. You can see this when comparing C2 S65 to C2 V65, C3 S65 to C3
V65, and C5 S65 to C5 V65. At Counter 4, the static sign had 100% compliance for
hourly average speed of travel. At all the other counters, however, the VSL signs posted
27
at the 55 mph speed limit, has the worst compliance rate (C2 V55, C3 V55, C4 V55, and
Commented [SO15]: Moved down to keep the night and day
discussion together.
C5 V55).
Figure 2-11 Speed Ccompliance by Hhourly Aaverage Sspeed.
100%
90%
80%
70%
Over
Speed
Limit
60%
50%
40%
Obey
Speed
Limit
30%
20%
10%
0%
C1
S75
C2
S65
C2
V65
C2
V55
C3
S65
C3
V65
C3
V55
C4
S65
C4
V65
C4
V55
C5
S65
C5
V65
C5
V55
S=Static Speed Limit Sign V=Variable Speed Limit Sign
Conclusions aAnd Recomendations
VSL signs resulted in average speeds which were lower than the regular static signs. In
evening conditions, where when VSL were set to 65 MPH and no construction activity
was in place, VSL signs resulted in average speeds which were statistically lower than
static speed limit signs through the construction zone. Daytime conditions showed
greater variation due to the level of congestion which occurred related to the construction
activity and lane closures. However, even dismissing the data impacts of low speeds
during daytime congestion, VSL signs showed greater compliance over static speed limit
signs.
28
A The potential concern of lower speed limits resulting in a greater variation in driver
speeds due to some drivers obeying the speed limit and other drivers traveling at
comfortable speeds did not occur. VSL signs reduced both the average speed and the
variation in speed. During evening conditions with the VSL sign set to 65 MPHmph, the
standard deviation of speeds reduced to 0.5 MPH mph to 1.0 MPHmph, down from 1.5
MPH mph to 5.0 MPH mph with the regular static signs.
Although crash data were not directly collected, it can be inferred from the reduction in
average speeds and speed variations that the VSL signs could contribute to reduction in
work zone crashes. It is considered that providing drivers with speed restrictions that
reflect the real time construction conditions contributes to building trust in the posted
speed by drivers. Drivers accept greater speed reductions while construction works are
underway during the day and lesser speed reductions mainly for mainly driver safety
considerations while no construction work is underway during the night.
VSL signs appear to be a promising technology which can result in safer work zones and
more effective speed regulations. The application of VSL signs in should notother states
is not be limited to work zones. Applications in Utah could include permanent signs
which vary the speed based on prevalent weather conditions, traffic volume, or truck
traffic volume. No immediate application has been researched, but UDOT is advised to
not limit the scope of VSL signs to work zones and should continue to watch the state-ofthe-practice usage of VSL signs. VSL signs equipped with traffic counting capability
29
would allow the use of their data for research as additional VSL signs are purchased and
put into operation by UDOT.
Improved safety to motorists and construction personnel occurs when drivers follow
predictable patterns and speeds. Speed limit signs which vary the limit based on prevalent
conditions will have greater compliance than static signs. This compliance will result in
more drivers traveling within the posted speed limit and fewer drivers traveling either
significantly faster or slower than the posted speed.
Commented [SO16]: Moved down from Study Methodology

because it sounds more like results and creates a nice conclusion.
Acknowledgements
Formatted: Left, Space Before: 6 pt
The authors would like to thank Daniel Hsiao of the Research & Development of UDOT
for his support as the project manager of this study and Deryl Mayhew of UDOT Region
2 for his help with coordinating among InterPlan, L2 Data Collection, UDOT, and WW
Clyde Construction. They are also appreciative of the assistance received from Kris
Peterson and Lisa Wilson (UDOT Region 2 traffic engineers), Rob Clayton (UDOT
Traffic Operations Center), and the personnel of L2 Data Collection, who collected speed
data in the field.
References
1. American Association of State Highway and Transportation Officials (ASSHTO),
2004. A Policy on Geometric Design of Highways and Streets.
30
2. Minnesota Department of Transportation, 2007. Field Evaluation of a Variable

Advisory Speed Limit System for Reducing Traffic Conflicts at Work Zones.
Transportation Research Board Annual Meeting 2007 Paper #07-2551.
3. Robinson, Mark. Examples of Variable Speed Limit Applications. Presentation given at
the Speed Management Workshop held in conjunction with the 79th Annual Meeting
of the Transportation Research Board. January 2000.
http://safety.fhwa.dot.gov/fourthlevel/ppt/vslexamples.ppt. Accessed December 2007.
4. TRB January 2000. Speed Management Workshop,
http://safety.fhwa.dot.gov/tools/docs/vslexamples.ppt. Accessed December 2007.
5. U. S. Department of Transportation Federal Highway Administration, December
2004. A Field Test and Evaluation of Variable Speed Limits in Work Zones.
6. Utah Department of Transportation. Policy 06C-5, 06C-25 and 06C-61.
7. Utah Automated Geographic Reference Center (AGRC). http://agrc.its.state.ut.us/.
Accessed December 2007.
LIST OF FIGURES
Figure 1 Variable speed equipment used in the study.

Figure 2 Location of the VSL signs and sensor locations.
Figure 3 Data collection schedule.
31
Figure 4 August 30, 2007 24 hour speed data counters 2 thru 5 65 mph speed limit 24/7.
Figure 5 Standard deviation of Speed.
Figure 6 Daytime average speed July 10 to September 28.
Figure 7 Night time average speed comparison.
Figure 8 Daytime average speed comparison.
Figure 9 Selected periods, daily average speed.
Figure 10 Average speed for counter 3, August 16.
Figure 11 Speed compliance by hourly average speed.
Commented [SO17]: Delete
32
LIST OF TABLES
Table 1 Statistics on Speed Reduction by VSL Signs, from Nighttime 65 MPH to

Daytime 55 MPH Speed Limit
Formatted: Normal, Indent: Left: 0", Hanging: 0.25", Tab

stops: 0", Left + 0.25", Left + 1.25", Left + 1.75", Left +
2.25", Left + 2.75", Left + 3.25", Left + 3.75", Left +
4.25", Left + 4.75", Left + 5.25", Left + Not at 6.49"
Commented [SO18]: Delete.
33

Copyedit A

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Copyedit A

Uploaded by

Copyright:

Available Formats

Formatted: Underline

Variable Speed Limits Signs: Effects oOn

Speed aAnd Speed Variation iIn Work Zones

The Utah Department of Transportation desired to evaluate the effectiveness of this

Formatted: Indent: First line: 0"

Formatted: Indent: First line: 0"

Commented [SO2]: Delete. This is word for word from the

Commented [SO3]: What is this referring to?

Formatted: Font: Not Italic

Formatted: Font: Not Italic

Formatted: Font: Not Italic

Formatted: Indent: First line: 0"

conclusions and recommendations.

Formatted: Indent: First line: 0"

technology (Reference #).

Commented [SO6]: Delete this part if the summary is okay.

Some of the documented VSL usage appears to be aimed at increasing vehicles

Formatted: Font: Bold, Underline

Formatted: UDOT body text, Space Before: 6 pt

Study Site Location and Construction Project Description

Formatted: Font: Bold, Underline

Formatted: Indent: First line: 0"

Commented [SO7]: Moved from section below to fit better with

Figure 2-1 Variable Sspeed Eequipment Uused in the Sstudy.

Formatted: UDOT body text, Space Before: 0 pt

(Photo by Thomas McMurtry, July 2007)

Formatted: Left, Indent: First line: 0"

Formatted: Font: Bold, Underline

Formatted: UDOT body text, Space Before: 0 pt

Figure 2-2 Location of the VSL Ssigns and Ssensors locations

Formatted: Font: 12 pt, Bold, Underline

Commented [SO8]: Chart made smaller to fit better.

Standard 65 MPH speed limit signs

Figure 2-3 Data Ccollection Sschedule.

Commented [SO9]: Instead of listing the description ON the

Formatted: Indent: First line: 0"

Commented [SO10]: This is pretty clear in the calendar already.

Commented [SO11]: Delete this sentence.

Commented [SO13]: Is this relevant? Delete?

Formatted: Font: 12 pt, Bold, Underline

Formatted: UDOT HEADING 1, Space Before: 6 pt

The distribution of speeds observed at various counters in general followed an expected

Formatted: Indent: First line: 0"

Figure 2-4 Comparison of Sspeed Ddata Ddistribution Uunder Ddifferent Sspeed

Formatted: Justified, Indent: First line: 0"

Formatted: Indent: First line: 0"

Figure 2-5 Standard Ddeviation of Sspeed.

(b) Night time standard deviation of speed

Formatted: Indent: First line: 0"

Figure 2-6 Daytime Aaverage sSpeed July 10 to September 28.

Nighttime Average Speed

Formatted: Font: Bold, Not Italic, Underline

Formatted: Indent: First line: 0"

Figure 2-7 Night time Aaverage sSpeed Ccomparison.

Average Speed, mph

(a) Nighttime average speed July 10 to July 28

Formatted: Font: Not Bold

Average Speed, mph

(b) Nighttime average speed July 31 to August 5

Formatted: Font: Bold, Not Italic, Underline

Formatted: Indent: First line: 0"

Figure 2-8 Daytime Aaverage Sspeed Comparison