URBAN FREEWAY TRAFFIC CHARACTERISTICS IN KUWAIT
A PRELIMINARY ANALYSIS
Dr. Ahmad H. Aljassar
Assistant Professor
Department of Civil Engineering,
Kuwait University, Kuwait
Dr. Aljassar obtained his Ph.D. from the University of
Waterloo, Canada in 1994. He is also an advisor to the
Maintenance Division of MPW.
Dr. Mohammed A. Ali
Transportation Engineer
Department of Civil Engineering,
Kuwait University, Kuwait
Dr. Ali obtained his Ph.D from Regional Engineering
College, Warangal, India. He is associated with Kuwait
University since 1993.
Dr. Omar I. Al-Saleh
Assistant Professor
Department of Civil Engineering,
Kuwait University, Kuwait
Dr. Al-Saleh earned his Ph.D. from the University of
Pittsburgh, USA. He has been Chairman, Dept. of Civil
Engg., Vice-Dean, College of Engg., and Asst. Vice
President of Kuwait University.
Major Eng. Sadon A. Al-Khaldi
Head of Engineering Section
General Traffic Department
Ministry of Interior, Kuwait
Major Al-Khaldi holds a B.Sc.in Civil Engineering and
an M.Sc. in Transportation Engineering from Kuwait
University.
Summary
Traffic volume on a road network is an important factor in the planning and design of roadway
facilities. Geometric and structural designs of roadways depend among other factors on traffic
volumes, their mix characteristics, loads, and distribution with time of the day, day of the week, and
month of the year. Kuwait University has commissioned a research project to establish Permanent
Traffic Counting Stations in Kuwait in cooperation with the General Traffic Department of the
Ministry of Interior. A total of fifteen permanent traffic counting stations, consisting of
representative locations for all the functional classes of roads and a screen line into the Kuwait
urban area are established. Loops and piezo sensor combinations are employed to collect
information on traffic volume, and vehicle speed. The paper presents a preliminary analysis of the
results from freeway locations. The results include, Peak Hour Volume, Peak hour variation,
Directional distribution, and Speed analysis. Hourly variation and Daily variation are also
presented in the paper.
Keywords: Traffic Volume, AADT, ESAL, Automatic Traffic Count, Continuous Count.
1.
Introduction
1.1.
General
The traffic count information is extremely important to traffic planning, design, and operation. This
data is regularly requested and used by developers, consultants, real estate agents, homeowner
associations, government agencies, and other citizens. Geometric and structural designs of roads
depend among other factors on traffic volumes, their mix characteristics, loads, and distribution
with time of the day, day of the week, and month of the year. Traffic volumes are measured using
the average annual daily traffic (AADT). This terminology implies that the count is a representative
of the average traffic conditions for the whole year. During some parts of year, traffic may be
higher than the AADT, and in others it may be lower than this value. However, AADT provides a
typical daily traffic volume at any location, usable for most situations where traffic counts are
needed as an input to a planning and/or engineering analysis.
The conversion of raw traffic count data collected during some part of the year to AADT requires
the use of AADT Conversion Factors. These factors are normally developed from data collected at
permanent traffic count stations. Typically, these factors vary by area type and functional
classification of the roadway where count is taken. Generally, the following three types of factors
are used:
1.
2.
3.
Hour-of-Day (HOD) Factors: To convert less than 24-hour data to 24-hour value.
Day-of-Week (DOW) Factors: To convert traffic in a specific day of the week to an average
weekday traffic.
Month-of-Year (MOY) Factors: To convert a typical weekday traffic for a specific month, to
an AADT.
This paper presents preliminary results of a research project sponsored by the Research
Administration of Kuwait University.
1.2.
The need for the study
Continuous monitoring of traffic on all sections of Kuwait’s road network is impractical and almost
impossible. This research is designed to monitor traffic on representative road sections on different
road classes. Traffic distribution with time on these road classes will be used to calculate expansion
factors that can be used to expand short-term traffic counts collected on any road at any time of the
day, any day of the week, or any month of the year.
Further, the data acquisition equipment collects data on axle loads and vehicle types. This
information is required in the structural design of pavement structures. Estimation of this
information may lead to under-design, which could result in pre-mature failure of pavements; or
over-design which means unnecessary expenditure. Traffic trends by time shall also be established
by this project, which is useful in the development of traffic forecasting procedures in Kuwait. The
results of this research are expected to be of extreme benefit to many agencies including the
Ministry of Public Works, Kuwait Municipality, Ministry of Interior’s (MOI) General Traffic
Department, and Planning and Engineering Consultants.
1.3.
Study objectives
The objectives of the research study are:
1. Establish permanent traffic count stations (PTCS) at selected locations in Kuwait.
2. Collect traffic volume, classification of vehicles as per FHWA (Federal highway
Administration) and speed data from the PTCS.
3. Calculate ESAL values for different functional classes of roads.
4. Determine daily, weekly, and monthly variation of traffic volumes.
5. Develop road class-specific traffic expansion factors to calculate AADT from short counts.
6. Establish volume trends and traffic forecasting procedures.
7. Compare and evaluate different traffic counting technologies.
2.
Review of Literature and Traffic Count Studies
Transport and highway authorities worldwide have established permanent traffic count stations.
Vermont State in the US has established a Continuous Traffic Counting (CTC) Program throughout
the state [1]. Monthly Average Daily Traffic (MADT), AADT, and Average Annual Weekday
Traffic (AAWDT) were calculated at each CTC station. The monthly factors for each CTC station
are grouped by similar seasonal traffic patterns established by FHWA guidelines. The results reveal
six generally "definable" groups for Vermont. Nevada Department of Transportation (DOT) [2]
administers about 5400 miles of roads in the state of Nevada. During 2001, hourly traffic volumes
were monitored continuously at 71 locations statewide. In addition, traffic volumes were collected
in short periods (7 days) and factored to Annual Average Daily Traffic (AADTs). Data is collected
in an hourly increment at various count locations state-wide. In the small city of Lloydminster in
Canada [3], since 1990 a counting program is in place utilizing both the automatic tube counters
and manual intersection traffic volumes. Automatic traffic counts are obtained using mechanical
counters with air tubes to sense the number of axles crossing a point. The machine counts each axle
and then divides by two. This type of count has a certain amount of error in that a vehicle with more
than two axles may get counted twice or three times as every time two axles cross the tubing, a
count is registered. This roadway volume is calculated by averaging the Tuesday, Wednesday and
Thursday count information. Manual counts are obtained by using employees (summer staff)
stationed at an intersection to record all pedestrian and vehicular movements within that intersection
and classify vehicles into different vehicle classes. Telemetry Traffic Monitoring Sites (TTMSs)
continuously record the distribution and variation of traffic flow by hours of the day, days of the
week, and months of the year from year to year and transmit the data to office of TranStat via
telephone lines [4]. Florida’s continuous count program has been expanded from the original 10
sites in 1936, to 285 sites at present. Florida DOT is working with local jurisdictions to obtain the
data from their continuous counters and thus Florida will have over 300 permanent counters in
operation.
In Hampshire County in the UK, traffic data is collected by three methods [5]: Manual classified
counts, temporary and permanent count sites. A team of enumerators, who use hand held capture
devices to classify traffic, usually over 12-hour period, undertakes manual classified counts.
Automatic temporary traffic counts are undertaken by means of pneumatic tubes. The data is also
collected from 100 permanent count sites using inductive loops cut into the carriageway. Four
PTCS were established in the state of Andhra Pradesh in India by the Ministry of Surface Transport
using loop and dynamic axle sensors (DYNAX) [6]. The purpose of this study was to get the
characteristics of the National and State highways and to develop a procedure for traffic forecasting
on these highways. Intersection traffic characteristics along with link volume characteristics were
also studied in Abu Dhabi, UAE [7].
The use of expansion factors has been reported in a number of studies. Erhunmwunsee [8] has
reported the use of two-stage process using expansion factors to estimate AADT. The short period
is first expanded to the daily total and then expanded to the annual total. This study attempted to
determine the effect of duration between 4 and 16 hours on accuracy, and the best time period to
begin each count. The study also determined the expansion factors for each length of count, and the
month of the year that would be most appropriate for estimating AADT. Effects of various
expansion factors on estimation errors were explored in another study by Sharma et al [9]. The
appropriateness of volume adjustment factors is expressed in terms of assignment effectiveness,
which mirrors the degree of correctness with which a sample site is assigned to an automatic traffic
recorder (ATR) group. Investigators found that AADT estimation errors are very sensitive to the
assignment effectiveness. Granato [10] presented an analysis of how much day of week/month of
year factors can reduce the error of predicting AADT from a short-term traffic count, utilizing data
from an ATR station maintained by the Iowa DOT in Cedar Rapids, Iowa. The benefits of factoring
are shown to be a one-quarter reduction in error of AADT prediction for a 24-hour count at this
station, with minimal added benefit of a (consecutive) multiple-day count. A combination of
approaches - using statistical measures such as the coefficient of variation, statistical procedures
such as cluster analysis, plots of monthly traffic factors, and geographical mapping of continuous
count sites - can produce seasonal factor groups and seasonal adjustment factors to substantially
account for seasonal variation and thus produce more accurate AADT estimates for end users [11].
3.
Methodology
3.1.
Study Area Characteristics
Kuwait is a rapidly developing country with a population of more than 2.5 million. It has an
excellent road infrastructure comparable to that of any industrialized country. The roads in urban
areas are classified into the following functional categories [12]:
Special Road Network (SRN)
Primary Road Network (PRN)
Secondary Roads (SR) and
Local Roads (LR)
Special Road Network includes motorways and expressways that are major through-routes for
traffic with grade-separated junctions and full acceleration and deceleration lanes. Roads in this
category have at least one hard shoulder per carriageway and do not have u-turn facilities. The
accessibility to SRN is restricted to motor vehicles only. Speed limit on such roads is 120 km/hr.
Traffic directions are divided by raised concrete barriers. SRN roads are mainly ring roads or radial
roads. A total length of about 900 km or 16.3 % of total road network consists of SRN.
Primary Road Network includes through traffic routes that are usually of a lower design standard
than SRN routes. Roads under this category have more frequent at-grade junctions. Speed limit on
these roads is 80 km/hr. Traffic directions are divided by paved or landscaped islands. A total length
of about 1654 km or 30 % of total road network consists of PRN.
Secondary Roads are used to distribute local traffic through a district and perhaps to serve a place
of importance within a local community. Such roads usually have U-turn facilities. Speed limit on
these roads is 60 km/hr and islands generally divide traffic directions. Secondary roads usually run
between blocks in a district to collect traffic from local roads and distribute them on SRN or PRN
roads. About 1113 km or 20.2 % of total road network consists of SR.
Local Roads include those, which provide access to individual commercial or local residential units.
They run within blocks in a district and distribute traffic on secondary roads. Speed limit on local
roads is 45 km/hr, and islands do not divide traffic directions. A Total length of about 1845 km or
33.5 % of total road network consists of LR.
Nine representative sites from among the four roadway types were selected for continuous
monitoring of traffic for this project in accordance with the guide lines of the Federal Highway
Administration's (FHWA) guide lines for site selection of the permanent traffic counting stations
[13]. Additional six locations established by the MOI’s General Traffic Department would also be
included in the study. The project’s nine sites comprise of three locations on SRN, two on PRN,
two on SR and two on LR roadways. The selected locations are listed in Table 1.
3.2.
PTCS Instrumentation
Inductive Loop type detector is selected for the study, keeping in view the advantages it offers on
other technologies. ADR 3000 automatic data recorder from PEEK TRAFFIC was selected for
monitoring of traffic on the selected locations.
Table 1. Selected Traffic Monitoring Sites
PTCS No.
1*
2
3
4
5
6
7*
8*
9*
10
11
12*
13*
14
15
Location
First Ring Road
Second Ring Road
Nuzha
Damuscus Street
Rawda
Hawally
Shaab
Fourth Ring Road
Fifth Ring Road
Ghazali Expressway
Sixth Ring Motorway
King Faisal Motorway
King Fahad Motorway
Fahaheel Motorway
King Fahad Motorway
Roadway
First Ring Road
Second Ring Road
Quraish Street
Damuscus Street
Rawda Street
Tunis Street
Cairo Street
Fourth Ring Road
Fifth Ring Road
Ghazali Expressway
Sixth Ring Motorway
King Faisal Motorway
King Fahad Motorway
Fahaheel Motorway
King Fahad Motorway
Roadway Type
SRN Ring
PRN Ring
LR
SR Radial
SR (Inside a district)
SR Commercial
SR
PRN Ring (Mid Town)
SRN Ring
SRN Radial (Port Traffic)
SRN Ring (Out Town)
SRN Radial
SRN Radial
SRN Radial (Sub-urban)
SRN Industrial
Note: * PTCS Nos. 1, 7 ,8 ,9, 12 & 13 are established by MOI.
ADR 3000 Automatic Data Recorder provides optimum functionality as permanent instrument. In
its basic configuration as a counter/classifier, it can monitor up to four lanes of traffic with a
combination of loops and Piezos. However, with the addition of easily fitted modules, the ADR3000 can monitor up to 64 lanes simultaneously or classify vehicles on 32 lanes of traffic. The ADR
can perform up to three selectable classifications simultaneously; e.g., speed by class by lane or
length by gap by speed. The type, configuration and format of data to be collected are selected by
the user from intuitive menu driven choices or may be custom-programmed. Available data types
include per-vehicle records, per-lane data, and binned vehicle classification by axle, speed, length,
gap, headway or combinations of the above. Vehicles may be classified according to FHWA/EEC
classification or user classification.
3.3.
Installation of equipment
Installation of the ADR-3000 counters is preceded by the construction of concrete foundations,
installation of protection posts and security housings, cutting out of 2mx2m loops and piezos in
asphalt in all the lanes of the selected locations. Figures 1 to 3 show the operational sequence of the
setup procedures. After completing the setup, the programming of counters follows to collect the
information on traffic by lane, by class and by speed.
Figure 1 Loop Cutting in the Pavement
Figure 2 A Close-up of Loop Cutting Machine
3.4.
Quality Assurance Surveys
Manual classified traffic count surveys are
performed during randomly selected periods to
validate the data. Live feeds of traffic will be
captured using video technology for one hour each
at the randomly selected time period and manual
count will be performed in the office. Pre-designed
data formats will be used to collect the data.
Figure 3 The Counter and Solar Panel in the Housing
4.
Results and Analysis
Preliminary results from count stations 1 and 9 (First Ring Road and Fifth Ring Road - refer to
Table 1) which were established by the Ministry of Interior were obtained by analysing the data
for various traffic characteristics. Summary statistics and, distribution of traffic by lane, by direction
and by speed are presented below.
100.0
100.0
90.0
90.0
80.0
70.0
60.0
West Bound
50.0
East Bound
40.0
Total
30.0
20.0
% Peak Hour Volume
% Peak Hour Volume
4.1.
First Ring Road
The Average Daily Traffic (ADT) was observed to be 10956 per lane, and the Average Weekday
Travel (AWDT) was 11971. The AWDT was about 9% more than the ADT. The total peak hour
volume varied from 2700 to 3800 vph during the study period. A typical hourly variation is shown
in Figure 4. Two peak periods typical to Kuwait are very clear from this figure. Morning peak (7-8
am) occurs in westbound direction and afternoon (1-2 pm) peak occurs in the eastbound direction
on this road. Due to these two peaks in opposite directions the total traffic is uniform between 6am2pm. Friday traffic is very different from a weekday as shown in Figure 5. There is only one peak
occurring between 6-9pm. The average peak hour volume at this station was in the range of 20002400 vph. The peak hour factor was about 0.96. Typical daily variation in a week is shown in
Figure 6. The daily volume is rather uniform throughout the week and low during weekends, being
the lowest on Friday as expected. It should be noted that in Kuwait, Thursday is a rest day in
Government but is a working day in most of the private sector, hence the lowest traffic is observed
on Friday, rather than both weekend days. On average, about 58% on weekdays and 56% on
Fridays travel westbound. Typical urban tidal flow on this road is evident from the results, where
peaks shift directions between morning and afternoon peak hours. Speed distribution by lane is
illustrated graphically in Figure 7. On average, about 77% of drivers travel below the speed limit of
100 kph. More than 30% in Lane 3 travel at high speeds (>120 kph.).
80.0
70.0
60.0
West Bound
East Bound
Total
50.0
40.0
30.0
20.0
10.0
10.0
0.0
0.0
0
2
4
6
8
10
12
14
16
18
20
22
24
Hour of Day
0
2
4
6
8
10
12
14
16
18
20
22
24
Hour of Day
Figure 4 First Ring Road – Weekday Hourly Variation
Figure 5 First Ring Road –Friday Hourly Variation
First Ring Road - Spe e d Distribution (Pe ak Hour)
>130
100%
60000
121-130
111-120
80%
50000
West Bound
30000
East Bound
Total Traffic
% Traffic
Vehicles/Day
100-110
40000
91-100
60%
81-90
71-80
40%
20000
61-70
20%
10000
51-60
40-50
0%
0
Thu.
Fri.
Sat.
Sun.
Day
Mon.
Tue.
<40
1
Wed.
2
3
Total
Lane No.
Figure 6 First Ring Road – Typical Daily Variation
Figure 7 First Ring Road – Speed distribution (Peak Hour)
4.2.
Fifth Ring Road
The Average Daily Traffic (ADT) was observed to be 16923 vehicles per lane, and the Average
Weekday Travel (AWDT) was 2113 vehicles per lane. The AWDT was about 24% more than the
ADT. The total peak hour volume was about 9000 vph. A typical hourly variation is shown in
Figure 8. Two peak periods typical to Kuwait are very clear from this figure. Morning peak (7-8
am) occurs in both east and west bound directions and afternoon (1-2 pm) peak also occurs in both
directions of this road. This is due to the location of the road, which serves both the residential and
work areas at either end of this road. Friday traffic is very different from weekdays as shown in
Figure 9. There is only one peak occurring between 7-9pm. The peak hour factor was about 0.958.
Typical daily variation is shown in Figure 10. The daily volume is rather uniform throughout the
week and low during weekends, being the lowest on Friday. Typical urban tidal flow on this road is
not observed at this location. Speed distribution by lane is presented in Figure 11. At an average
about 98% of drivers travel below the speed limit of 120 kph, however in the inside lane of east
bound direction more than 15% travel at greater than the speed limit of 120 kph during peak hour
period. More than 20% in Lane 5 travel at high speeds (>120 kph.).
7000
10000
9000
6000
8000
5000
6000
West Bound
5000
East Bound
4000
Total
3000
Veh./Hr.
Veh./Hr.
7000
West Bound
4000
East Bound
3000
Total
2000
2000
1000
1000
0
0
0
2
4
6
8
10
12
14
16
18
20
22
24
Hour of Day
Figure 8 Fifth Ring Road – Weekday Hourly Variation
5.
0
2
4
6
8
10
12
14
16
18
20
22
24
Hour of Day
Figure 9 Fifth Ring Road –Friday Hourly Variation
Conclusion
This research establishes permanent traffic counting stations for the first time in Kuwait. A review
of literature has indicated that in several countries around the world, the continuous traffic counting
programs are used to collect various traffic characteristics such as traffic volumes, composition,
loads, traffic variation by day, week, month and year. One of the main objectives of this research is
to develop road-specific expansion factors, which will be used to estimate AADT from shot-term
100%
160000
>130
80%
121-130
140000
Vehicles/Day
120000
100000
West Bound
80000
East Bound
60000
Total Traffic
% Traffic
111-120
100-110
60%
91-100
81-90
71-80
40%
61-70
51-60
40-50
20%
<40
40000
20000
0%
1
0
Thu.
Fri.
Sat. Sun.
Day
Mon
.
Tue
.
We
d.
Figure 10 Fifth Ring Road – Typical Daily Variation
2
3
4
5
Total
Lane No.
Figure 11 Fifth Ring Road – Speed distribution (Peak Hour)
counts in Kuwait. The preliminary analysis of two stations has clearly indicated the locational
variation in the traffic characteristics.
Acknowledgement
The authors wish to acknowledge the support of the Research Administration of Kuwait
University, and the co-operation of both the General Traffic Department (Ministry of Interior) and
the Maintenance Division of the Ministry of Public Works.
References
1. VERMONT DOT, “Continuous Traffic Counter Grouping Study and Regression Analysis
Based on 2001 Traffic Data” ,Vermont, USA, 2001.
2. NEVADA DOT, “The Annual Traffic report”, Traffic Information Division, Carson city,
Nevada, 2002.
3. STEFANUK, G., “City of Lloydsminster – Annual Traffic Volume Survey”, The City of
Lloydsminster, Engineering and Public Works Department, B.C., Canada, Dec. 2000.
4. FLORIDA DOT, “Project Traffic Forecasting Handbook”, Florida, October 2002.
5. HAMPSHIRE COUNTY COUNCIL, “Hampshire Local Transport Plan 2001-2006 Section
9.1 – Monitoring and performance Indicators”, Hampshire, UK 2001.
6. CHARI, S.R., B.P. Chandrasekhar “Report on Project Establishment of permanent Traffic
Counting Stations in A.P.”, Regional Engineering College, Warangal, India, 1998.
7. AL KATHARI, A. S., R. K. MUFTI, A. M. GHARIB, F. B. WILLIAMS, “Traffic
Characteristics in the Arab Gulf Region – A Case Study in Abu Dhabi, United Arab
Emirates”, First Gulf Conference on Roads, Kuwait, March 2002.
8. ERHUNMUNSEE, PAUL O., “Estimating Average Annual Daily Traffic Flow from Short
Period Counts”, ITE Journal, V. 61, No. 11, November 1991, pp. 23-30.
9. SHARMA, SATISH C., Peter Kilburn and Youngquiang Wu, “Precision of Average Annual
Daily Traffic Volume Estimates from Seasonal Traffic Counts: Alberta Example”, Canadian
Journal of Civil Engineering, V. 23, pp. 302-304, 1996.
10. GRANATO, S., “The Impact of Factoring Traffic Counts for Daily and Monthly Variation
in Reducing Sample Counting Error”, Crossroads 2000 Conference, Iowa State University
and Iowa Department of Transportation, 1998.
11. AUNET, B., “Wisconsin's Approach to Variation in Traffic Data”, North American Travel
Monitoring Exhibition and Conference (NATMEC), Middleton, Wisconsin, 2000.
12. ALJASSAR, AHMAD H., ABDULAZIZ A. AL-KULAIB, EL-SAYED W. METWALI, &
KHALED N. HELALI, “Performance of Roads in Kuwait”, Proceedings, 1st International
Conference on Performance of Roads, Bridges and Airport Pavements in Hot Climates,
Dubai, United Arab Emirates, April 28-29, 1998.
13. U.S DEPARTMENT OF TRANSPORTATION, “Traffic Monitoring Guide”, Federal
Highway Administration, Washington, DC, USA, 1992