Improving Intersection Design Practices

Adam Kirk, Chris Jones, and Nikiforos Stamatiadis

This study aims to use operational characteristics to help determine the engineers from considering one or more of the alternatives, even though
size and the design of intersections on the basis of a targeted level of oper- such designs may be appropriate. Unwarranted operational or safety
ation. This approach will allow for a preliminary evaluation of a broader problems or unwarranted costs may be incurred when suboptimal
range of possible designs, by screening out those designs considered less designs are constructed.
desirable or inappropriate on the basis of operational performance. This The goal of this project is to improve intersection design practices
approach will also allow for a more objective comparison of all alterna- by (a) expanding the scope of intersection design alternatives consid-
tives because all options target the same operational service level. The use ered and (b) providing a structured and objective evaluation process
of the critical lane analysis method was considered an appropriate to compare alternative design concepts. This goal is achieved through
approach for developing such size estimates for intersections. Similar the development of a process capable of evaluating 13 alternative
methods for stop-controlled and yield-controlled intersections were also traffic control designs through the use of more than 12,000 unique
identified because it was necessary to expand these methods to include lane configurations for a given location. To facilitate this evaluation,
unsignalized designs as well. The Intersection Design Alternative Tool the methodologies described in this paper were developed into a
developed through this effort is capable of evaluating 13 intersection software-based tool identified as the Intersection Design Alternative
alternatives and identifying preferred lane configurations from more Tool (IDAT). The tool provides designers with a list of potential solu-
than 12,000 available configurations. tions that are based on the minimum number of lanes required to
achieve a targeted level of operation.

Intersections are a critical component of the roadway system and fre-

quently act as choke points on the transportation system. Intersection
design is a balancing act of various elements and constraints aiming Alternative Intersection Designs
to produce a solution that addresses mobility, safety, environmental,
A comprehensive list of alternative intersection designs was obtained
and financial aspects of the project. To achieve this balance, alterna-
from the Maryland State Highway Administration’s Unconventional
tive strategies and options must be identified, developed, and evalu-
Arterial Intersection Design Tool, which provides conceptual infor-
ated in a systematic manner. Significant advances in transportation
mation and considerations for a wide range of alternative intersection
engineering have identified new traffic control measures, practices,
designs (1). A list of the at-grade intersections in this tool is provided
and design concepts capable of further increasing the operational effi-
below:
ciency and safety of intersections. Moreover, no significant efforts
have been undertaken to provide an effective comparison between
• Unsignalized inside left turn;
various types of intersection designs or traffic control measures that
• Median U-turn, signalized;
could demonstrate their effectiveness. For example, the Highway
• Median U-turn, unsignalized;
Capacity Manual uses two different delay metrics for estimating level
• Superstreet, unsignalized;
of service (LOS) for unsignalized and signalized intersections, thus
• Superstreet, signalized;
rendering comparisons between these two types of intersection
• Continuous flow;
control infeasible. Current practice, while achieving great strides
• Continuous green T;
in improving the efficiency, lacks a systematic, objective, and
• Jughandle;
well-defined approach to evaluating individual design alternatives.
• Bowtie;
A review of current literature identified 13 alternative intersection
• Modern roundabout; and
designs. No systematic process can be identified that compares these
• Paired.
alternative designs. Most guidelines identify the need for comparative
studies but do not identify the factors or methods that one should
Several of these alternative intersection designs have been used
apply in determining the most beneficial design. The lack of specific
throughout the country and aim to improve intersection operation
guidance at both the national and state levels is likely to discourage
and safety. The median U-turn design has been used in Michigan
extensively for years, the jughandle design in New Jersey, and the
continuous flow in New York and Maryland. The use of modern
A. Kirk, Kentucky Transportation Center, and N. Stamatiadis, Department of Civil
Engineering, University of Kentucky, 176 Raymond Building, Lexington KY 40506. roundabouts is perhaps the most adopted alternative, and its use is
C. Jones, Stantec, 1409 North Forbes Road, Lexington KY 40511-2050. increasing rapidly throughout the United States.
Corresponding author: N. Stamatiadis, [email protected]. Despite their use, limited guidance is available on the evaluation,
design, and implementation of these designs. AASHTO’s A Policy on
Transportation Research Record: Journal of the Transportation Research Board,
No. 2223, Transportation Research Board of the National Academies, Washington, Geometric Design of Highways and Streets contains guidelines on the
D.C., 2011, pp. 1–8. design of standard intersections and some guidance on the median
DOI: 10.3141/2223-01 U-turn, jughandle, and roundabout alternatives (2). However, this

1
2  Transportation Research Record 2223

guidance is limited and does not address when and how to use these • Superstreet, unsignalized;
alternatives. A recent report by FHWA addressed the restricted cross- • Superstreet, signalized;
ing U-turn intersections (superstreet) and median U-turn at-grade • Continuous flow;
intersections (3). Most documented is the use of modern roundabouts. • Continuous green T;
Twelve states have developed roundabout guides that address the • Jughandle; and
planning, design, and operations of roundabouts, primarily based on • Bowtie.
the FHWA Roundabouts: An Informational Guide (4), although
several states have much more comprehensive guides. These intersections may be broadly grouped in two categories:
In addition to published research, a review of design guides used signalized and unsignalized control.
by each state was undertaken to determine the factors considered
in intersection design and how decisions on control type and size
are reached. Of the 41 state transportation agencies reviewed, only Intersection Design Procedures
Florida, Missouri, New Jersey, New York, Texas, and Washington have
developed their own intersection design guidelines contained within a A basic problem in comparative analysis of intersection designs is
separate intersection design manual or included within their roadway ensuring that the alternatives examined all deliver a similar level of
design manuals. All states reviewed have intersection design guidance operational performance. For instance, depending on the volume
that adheres to or follows the AASHTO guidance and the Manual on distribution, a signalized intersection with two approach lanes may
Uniform Traffic Control Devices for determining traffic control service the same volume as a single-lane roundabout. However,
(mainly for signalization). Of the states with independent guides, the during the initial concept development, both alternatives may be
most frequently considered design factors are operational analysis compared with two-lane approaches. This method leads to comparison
and construction cost (five of six states with specific guidance). of alternatives with vastly different operational performance in
These two factors are considered controlling for designing and eval- addition to costs and right-of-way and environmental impacts. The
uating intersection options because the factors define the operational approach taken by this project was to identify the minimum foot-
and construction efficiency of the intersection. A few manuals men- print for each intersection design for a given traffic demand, capable
tion alternative intersection designs but do not provide any guidance of meeting targeted operational parameters. A volume-to-capacity
for when they could be considered as viable alternatives. Moreover, ratio (v/c) of 0.90 was chosen as the threshold in the development
no manual provides specific guidance for selecting appropriate of IDAT. If an alternative meets this threshold, it is considered to be
intersection design or control types; most manuals simply note that feasible and no further operational evaluation should be carried
comparisons among alternatives should be performed. It is apparent forward in consideration and development. This approach allows
that there is a lack of tools that provide designers or planners with an for full comparison of other design factors such as construction costs,
estimate of appropriateness for different intersection designs. right-of-way, and environmental impacts. An example of the benefits
The review of state design practices revealed that there is not a provided by this approach is provided in the following section.
significant amount of research on alternative intersection designs or The critical aspect of this approach is to determine the optimum
design procedures. The limited guidance available is provided by design scenario that can meet the desired operational threshold. The
state agencies that have developed their own intersection design optimum design in this instance is that with the smallest footprint,
guidelines. No state has developed a systematic process that com- and subsequently, smallest construction costs and impacts. Capacity
pares these alternative designs. Most manuals identify the need for analysis software may be used for design and sizing intersections;
comparative studies but none identify the factors that one should however, an iterative process is required for each alternative to
consider in weighing alternatives and determining the optimal achieve the desired level of capacity. More than 12,000 feasible lane
design. Maryland is the only state that is in the process of develop- configurations have been identified for signalized intersections
ing such an approach, but not much progress has been made since alone, most of which can be developed for each of the 13 alternative
2005 when the concept was initiated. The development of separate designs considered. Using traditional capacity analysis software or
manuals for roundabouts by a few states is a step in the right direc- microsimulation models to design the appropriate lane configuration
tion for identifying and considering alternative intersection designs; for each alternative would be time-consuming and limit the range of
however, these manuals do not provide a means for comparison and alternatives to be considered. Therefore, the research team set out to
may further segregate alternative designs from traditional or other identify methodologies that could be used to directly link the traffic
alternative designs. demand (i.e., design hour turning movement volumes) to the opti-
mum lane configuration for each alternative. It was understood that if
achieved, the methodologies would have to apply to signal-controlled,
Research Findings stop-controlled, and yield-controlled (roundabout) alternatives.
The most promising approach that could be used in designing and
From the literature review, the following 13 different intersection evaluating comparative alternatives was critical lane analysis (CLA).
alternatives were identified for consideration in our research: This method allows for the automation of the design process of
signalized intersections by systematically linking traffic demand,
• Signalized; geometric design, and operational LOS. CLA, as developed by Messer
• Roundabout; and Fambro, uses the geometry of the intersection and intersection
• All-way stop; traffic volumes as the basis for establishing a measure of potential
• Two-way stop; performance and, by extent, of capacity (5). CLA uses the volumes
• Unsignalized inside left turn; of the approaches for an intersection to estimate their distribution
• Median U-turn, signalized; among the available lanes. Once volumes are apportioned to each of
• Median U-turn, unsignalized; the lanes, phasing plans are developed that allow for the appropriate
Kirk, Jones, and Stamatiadis 3

intersection movements. Critical volumes for each phase are deter- TABLE 1   Intersection Approach Volumes,
mined on the basis of certain rules, and these volumes are summed Major Streets
to determine the total critical lane volume for the intersection. This
Total Major Street
sum can then be directly related to the LOS definition for signalized Volume (vph) Eastbound Westbound
intersections.
Similar techniques (i.e., estimates of capacity) have been developed 1,800 1,080 720
for unsignalized intersection designs as well. The Highway Capacity 1,400 840 560
Manual provided intersection capacity estimates based solely on 1,000 600 400
conflicting movements and reserve capacity while considering 600 360 240
intersection geometry (6).
A recent report offered another consideration for estimating capac-
ity for roundabouts (7). This report develops control delay models TABLE 2   Intersection Approach Volumes,
for single-lane and multilane roundabouts by using the critical lane Minor Streets
methodology shown in Equations 1 and 2, respectively.
Total Minor Street
Volume (vph) Northbound Southbound
Single-lane roundabouts:
1,200 600 600
ccrit = 1130 ext ( −0.0010 × vc ) (1) 800 400 400
400 200 200
Multilane roundabouts: 860 600 260

ccrit = 1130 ext ( −0.0007 × vc )

570 400 170
(2)
285 200 85
where ccrit is entry capacity of critical lane [passenger car unit (pcu)/h]
and vc is conflicting flow (pcu/h).
The methods discussed for estimating intersection capacity present the north–south directions were the minor street. A combination of
simple estimates based on intersection geometry and turning volumes. different volumes and turning percentages was determined for the
These methods, although not as refined as current microsimulation east–west direction and the north–south direction. The volumes
models or more complex macro models, allow for direct linkage were selected to be greater than the minimum volumes to satisfy the
between intersection design and operation. Such simple models allow 4-h signalization warrant (8). The volume combinations used are
for manipulation through computational models, which permit the shown in Tables 1 and 2.
automation of preliminary designs for establishing the basic geometry The east–west street used two different turning percentages. The
needed to achieve a desired intersection capacity. CLA and unsignal- first was 10% left turns and 10% right turns, and the second was 15%
ized intersection LOS methods have served as the foundation for the left turns and 15% right turns for each of the four different volumes.
calculating procedures used in the current version of the Highway The turn percentages used for the north–south street were not uniform
Capacity Manual. These approaches are viewed as a basic, funda- and were based on the total northbound approach volume (Table 3).
mental process for evaluating intersection design alternatives. The Ninety-six scenarios were created on the basis of these approach
focal point of all these approaches is that they provide the potential volumes and turn percentages. For each scenario, different calcula-
for a common basis of comparison (i.e., v/c or unused capacity), tions were needed to determine the total critical volume for the all-
which can be used in targeting design options and provide a common way stop-controlled and signalized intersections, the approach and
basis for comparisons. conflicting volumes for roundabouts, or a critical approach volume
for two-way stop-controlled intersections. The process followed for
each of the traffic control options is described next.
Evaluation Approach CLA was used to determine the lane configuration for signal-
controlled intersections, and approach and conflicting volume totals
The concepts presented above have not been proven, and the first step were used to determine the lane configuration for roundabouts.
in this effort was to verify the relationship between delays and inter- The lane configuration used for signalized intersections was also
section design on the basis of CLA. Verification of these relation- used for two-way and all-way stop-controlled intersections. Auxiliary
ships was achieved through a simulation effort of various intersection
designs and traffic control strategies. Different scenarios of volumes
were identified to be simulated and obtain estimates of control TABLE 3   Minor Street Turn Percentages
delays. These delays were then used in determining the relationship
Northbound
between traffic volumes and delays for each intersection control option Approach Left Turn Right Turn
evaluated. This study considered only four-leg intersections, and the Volume (vph) (%) (%)
control types examined are two-way stop, all-way stop, signal, and
roundabout. The corridor simulation (CORSIM) software was cho- 600 10 10
sen because of its microscopic nature and ability to simulate traffic 20 30
conditions in various traffic control environments. 400 10 10
20 30
The first step of this work was to determine the different traffic 30 60
volume scenarios to use in the simulation models. The east–west
200 30 60
cardinal directions were considered the major street approaches, and
4  Transportation Research Record 2223

70

60

Intersection Control Delay

50

40

30

20

10

0
0 200 400 600 800 1000 1200 1400 1600
Critical Volume

FIGURE 1   Signalized intersection delay and critical volume.

left- and right-turn lanes were provided when turning volumes value for each volume scenario and traffic control was recorded for
exceeded 100 vehicles per hour (vph). each approach and calculated for the entire intersection.
For each type of control evaluated, either the corresponding key
volume or total volume was used to determine the relationship
Simulation Results between control delay and this volume metric. Regression analysis
was used to find a line of best fit for the data that could correlate
Each of the 96 volume scenarios for each of the four traffic control delays to the corresponding volume metric. For the signal-controlled
options was evaluated by using CORSIM. Default values were used and all-way stop-controlled intersections, critical volume was used as
for all parameters that were not modified among the various runs. the predicting variable. For the two-way stop-controlled intersections,
Control delay per vehicle was measured for each scenario and con- the approach critical volume was used, and for the roundabout, the
trol type. The multiple run feature was used to run each simulation approach and conflicting volume was used. A definitive capacity
four times using a different random number (i.e., representing a threshold was identified in which delay is shown to increase expo-
different traffic volume arrival pattern). The average control delay nentially for all control types examined. Figures 1 and 2 correlate the

60

50
Approach Control Delay Per Vehicle

40

30

20

10

0
0 200 400 600 800 1000 1200
Approach and Conflicting Volumes

FIGURE 2   Roundabout approach delay and approach and conflicting volume.
Kirk, Jones, and Stamatiadis 5

effect of the predictor variables on intersection delay. This threshold example, the median U-turn operates as a signalized intersection at
was identified as the capacity of the intersection for the purposes of its center, paired with two adjacent intersections to accommodate
determining intersection operations and is summarized below: left-turning movements. Each alternative has advantages and dis­
advantages as well as differing turn movement arrangements that will
• Signalized intersections: 1,400 vph (based on critical volume), optimize efficiency of each design. Furthermore, each alternative may
• All-way stop-controlled: 1,200 vph (based on critical volume), also be manipulated to accommodate a wide range of alternative lane
• Two-way stop-controlled: 900 vph (based on approach critical configurations to meet the unique demands of each project. To evalu-
volume), and ate the full range of alternatives and lane configuration combinations
• Roundabouts: 1,000 vph (based on approach and conflicting available, the authors decided that evaluation with a software appli-
volume). cation was appropriate. The following sections discuss the specific
requirements used to develop IDAT software.
Regression models were developed for each of the four intersection
controls to determine whether the relationship noted between the
volume metrics used and the delay estimated was statistically sig-
Intersection Lane Configuration Combinations
nificant. All tests indicated that the relationships were significant, and
all coefficients and intercepts were significantly different than zero. The 12,000 different lane configurations evaluated for each alter­native
Therefore, it can be concluded that the volume metrics used for each include eight different left- and right-turn auxiliary configurations
intersection control are capable of capturing the changes in the delays for each of one, two, and three through-lane combinations for a total
and can be used as indicators of the capacity and level of operation of 24 combinations for each approach. All combinations of each
of the intersection as a result of the traffic control used. approach were evaluated with each approach, except that the major
On the basis of this analysis, the derived critical volume procedures and minor streets were restricted to the same number of through lanes.
were validated. Capacity was identified by significant deflection iden- Figure 3 shows the eight different approach combinations for a single
tified in the delay curve. These critical volume capacities can be used through-lane alternative.
to establish the ultimate threshold for the targeted performance values These eight approach configurations developed for the signalized
of each alternative design, thus ensuring that designs operate below intersections served as the basis for the other intersection alternatives,
capacity and at an acceptable LOS. Therefore, these volumes are which were modified to meet the unique demands of the differing
1,350 vph for signal-controlled, 1,200 vph for all-way stop, 1,000 vph designs.
for two-way stop, and 1,000 vph for roundabout. These values were Each of the eight approach configurations was scored according
used in determining the appropriate intersection design in IDAT. to (a) the total number of lanes used in the design and (b) the desir-
ability of the configurations from an operational, safety, and driver
expectancy (i.e., commonality of design used) standpoint. Lane
Intersection Design Alternative configurations were rated as follows:
Tool Development
• 1: 8 (highest score),
Most of the designs considered in this study manipulate a traditional • 2: 6.5,
design, evaluated above, through redirected or channelized turn • 3: 6.5,
movements to address problematic or heavy turning movements. For • 4: 5,

ONLY ONLY

Ln Config 1 Ln Config 2 Ln Config 3

ONLY ONLY ONLY ONLY ONLY ONLY

Ln Config 4 Ln Config 5 Ln Config 6

ONLY ONLY ONLY ONLY ONLY ONLY

Ln Config 7 Ln Config 8

FIGURE 3   Approach lane configurations (ln config 5 lane configuration).

6  Transportation Research Record 2223

• 5: 4, between alternatives. The scoring method provides five points for

• 6: 2, an approach with a single lane. Approaches with five or more lanes
• 7: 3, and receive zero points. Turn lanes, such as a left or right auxiliary lane,
• 8: 1 (lowest score). are counted as one-half of a lane because they will likely be required
for only a short length. The average score of all approaches for the
For feasible combinations (i.e., v/c is less than 0.90), a total inter- design is used in the final scoring. Two points were deducted from
section score was then developed as the sum of the individual the overall score for jughandle and bowtie designs because of their
approach scores. The combination with the highest score was chosen increased space requirements. Even though intersection size may be
as the preferred configuration for that alternative. For each alterna- disaggregated into components, including number of approach lanes,
tive design, a preferred configuration was developed for single-lane, intersection number of lanes (including auxiliary lanes), and physical
two-lane, and three-lane approaches on the major street. intersection area, such a detailed approach was not appropriate for the
Volume transformations were used for each of the innovative inter- level of anticipated use of the evaluation tool.
section designs, and they were decomposed to the appropriate signal-
ized and unsignalized elements to estimate the lane requirements.
For example, in the median U-turn (unsignalized), all left-turning Example Application
volume was added to the through movement and used the left turn
at the U-turn location. The required number of lanes to serve the To test the capabilities of IDAT, an intersection project being consid-
reconfigured volumes at the intersection produced the new configu- ered by the Kentucky Transportation Cabinet was evaluated to identify
ration for the intersection, and a similar approach for an unsignalized potential alternative solutions. This section focuses on the practical
intersection produced the requirements for the U-turn. solution developed for the intersection of KY-17 at Old Madison Pike
in Fort Wright, Kentucky, as part of the KY-17 intersection improve-
ment study conducted by the Northern Kentucky Area Planning
Preferred Alternative Selection Commission and KYTC. The existing stop-controlled intersection
on the westbound approach operated at LOS F during the morning
IDAT provides a multiple-level screening analysis to allow the user and afternoon peak periods (9). Figure 4 shows the projected turning
to select the preferred alternative or alternatives to be carried forward movement counts at the intersection.
for further analysis. All design alternatives that have been identified The project was initially designed with a traditional design
as feasible through CLA are carried forward and presented in the approach, which analyzed two alternatives. The first was a traditional
final output. This analysis includes the identification of the highest signal installation, and the second introduced a modified modern
scoring single-lane, two-lane, and three-lane configurations for each roundabout. The design for these alternatives yielded project esti-
alternative. If multiple approach lane configurations are feasible for mates between $5 million and $5.8 million because both alternatives
a given alternative, those with a greater number of through lanes are caused major impacts to the existing bridge structure and rock cuts
identified as “not recommended” to indicate that a configuration with near the intersection. These alternatives are shown in Figure 5.
a smaller footprint is feasible. The feasible alternatives are evaluated IDAT results confirmed the proposed alternatives, which identified
using a weighted scoring scheme to assist in the final evaluation.
a three-lane approach on the major street for signalized operations
The scoring system used to evaluate the intersection alternatives
and identified the roundabout as not feasible, thus requiring addi-
examines the size of the intersection (right-of-way requirements);
tional modification, as shown in Figure 6. Other feasible alternatives
safety performance for vehicles, pedestrians, and bicycles; and access
management capabilities. These criteria are also assigned weights to
indicate their level of importance in the project; IDAT default values
assume an equal weight (0.33) for each criterion. Each alternative is KY 17
scored against these criteria, and a composite total score is developed.
The composite score is then determined as the sum of the products of
the individual scores and criterion weight.
The alternative scoring was developed by an expert panel employed
for this task. The panel consisted of traffic operations, highway safety,
and design engineers who were asked to evaluate and score each
intersection design on their a priori understanding of the level of safety,
pedestrian and bicycle accommodation, and access management
control provided by the intersection. The scoring method used a five-
point scale, where 1.0 represented the lowest score and 5.0 the highest.
For intersection designs with similar control but a greater number of
approach lanes, relative scores were developed according to the exper-
tise of the panel. The safety level was scored individually for vehicular,
pedestrian, and bicycle traffic, creating a unique score for each
category because each travel mode has different safety concerns.
The size of intersection was used to develop the scores for the
right-of-way. Because all alternatives are developed to operate at the
same level of efficiency, the size of the intersection becomes a critical
KY 17
determinant of suitability. Although this metric cannot provide pre-
cise estimates at the preliminary design stage because topographic or FIGURE 4   Design hour volumes, KY-17 at
other constraints on the site, it can provide a relative comparison Old Madison Pike.
Kirk, Jones, and Stamatiadis 7

(a) (b)

FIGURE 5   Initial design alternatives.

SAFETY
MINIMUM LANE

ROW

A.M.

SCORE
OPERATION CONFIGURATION

Bike
Veh.

Ped.
INTERSECTION ALTERNATIVE
EVALUATION
L4 L2

0.3

0.1

0.1

0.1

0.3
L3 L1 L4 L2 U U

2-Way Stop Control* Not Feasible 0 2.5 1.5 2.5 1.0 0

4-Way Stop Control Not Feasible 0 4.0 4.0 5.0 1.5 0

Signalized Intersection (1 lanes) Not Feasible 0 4.0 4.0 3.5 2.5 0

Signalized Intersection (2 lanes) Not Feasible 0 3.0 3.0 3.0 3.0 0

Signalized Intersection (3 lanes) Feasible 4 2.5 2.5 2.0 3.0 3.08

Jughandle A EB (1 Lanes) Not Feasible -2 3.0 2.5 3.0 3.5 0

Jughandle A EB (2 Lanes) Not Feasible -2 2.5 1.5 2.5 4.0 0

Jughandle A EB (3 Lanes) Feasible 2 2.0 1.5 1.5 4.0 2.53

Jughandle A WB (1 Lanes) Not Feasible -2 3.0 2.5 3.0 3.5 0

Jughandle A WB (2 Lanes) Not Feasible -2 2.5 1.5 2.5 4.0 0

Jughandle A WB (3 Lanes) Feasible 2 2.0 1.5 1.5 4.0 2.53

Jughandle A EB-WB (1 Lanes) Not Feasible -2 3.0 2.5 3.0 3.5 0

Jughandle A EB-WB (2 Lanes) Not Feasible -2 2.5 1.5 2.5 4.0 0

Jughandle A EB-WB (3 Lanes) Feasible 2 2.0 1.5 1.5 4.0 2.53

Roundabout Not Feasible 0 5.0 4.0 4.0 5.0 0

Median U-Turn (Signalized) (1 Lanes) Not Feasible 0 4.0 3.5 2.5 5.0 0

Median U-Turn (Signalized) (2 Lanes) Not Feasible 0 3.0 2.0 2.0 5.0 0

Median U-Turn (Signalized) (3 Lanes) Not Feasible 0 2.5 2.0 1.5 5.0 0

Median U-Turn (Unsignalized)* Not Feasible 0 4.0 2.5 2.5 5.0 0

Superstreet (Signalized) Feasible 1.5 3.5 2.5 1.5 4.0 2.64

Superstreet (Unsignalized) Not Feasible 0 3.5 1.5 1.5 4.0 0

Inside Left Turn (Signalized) (NB 'T') (1 Lane) Not Feasible 0 4.0 1.0 1.0 3.5 0

Inside Left Turn (Signalized) (NB 'T') (2 Lane) Feasible 2.25 3.0 1.0 1.0 4.0 2.61

Inside Left Turn (Signalized) (NB 'T') (3 Lane) Not Recommended 2 2.5 0.5 0.5 4.0 2.37

Inside Left Turn (Signalized) (SB 'T') (1 Lane) Not Feasible 0 4.0 1.0 1.0 3.5 0

Inside Left Turn (Signalized) (SB 'T') (2 Lane) Feasible 1 3.0 1.0 1.0 4.0 2.2

Inside Left Turn (Signalized) (SB 'T') (3 Lane) Not Recommended 0.75 2.5 0.5 0.5 4.0 1.95

Inside Left Turn (Unsignalized) (NB 'T')* Not Feasible 0 3.5 1.0 1.0 3.0 0

Inside Left Turn (Unsignalized) (SB 'T')* Not Feasible 0 3.5 1.0 1.0 3.0 0

Bowtie (1 Lane) Not Feasible -2 3.5 3.5 2.5 5.0 0

Bowtie (2 Lane) Not Feasible -2 2.5 2.5 2.0 5.0 0

Bowtie (3 Lane) Not Feasible -2 2.0 2.0 1.5 5.0 0

FIGURE 6   IDAT output (EB 5 eastbound, WB 5 westbound, NB 5 northbound, SB 5 southbound).

8  Transportation Research Record 2223

do not have to involve comparison of operations, but instead can

focus on costs and impacts to the associated project area.
The tool identifies the most efficient design (minimum number
of lanes) that is capable of meeting a targeted level of operation. A
designer is presented with several options that meet the minimum
operational requirements, allowing examination of other trade-offs
such as right-of-way impacts, safety considerations, and the like.
This approach eliminates the need to compare different alternatives
with varying performance levels across different types of traffic con-
trol measures. The proposed approach provides a greater efficiency
in the evaluation and conceptual design of intersection alternatives,
with the intent to achieve greater operational efficiency and improved
safety performance. The approach allows for a more appropriate and
properly customized design for each intersection, avoiding the use of
standard or typical designs.

FIGURE 7   IDAT recommended alternative.

References
identified were jughandle and superstreet designs, which were not 1. Applied Technology and Traffic Analysis Program. Unconventional
feasible because of sight constraints. However, an inside left-turn Arterial Intersection Design. University of Maryland, College Park, Md.,
lane was also identified as feasible; it could operate with a reduced 2005. http://attap.umd.edu/bbs/zboard.php?id=projects&select_arrange=
number of lanes (two) on KY-17 and have minimal impacts on Old headnum&desc=asc&page_num=5&selected=&exec=&sn=off&ss=
on&sc=off&category=10&ss=on&ss=on&keyword=. Accessed June 22,
Madison Pike. This alternative is shown in Figure 7. 2010.
Implementation of the inside left-turn lane has the ability to 2. A Policy on Geometric Design of Highways and Streets. AASHTO,
maintain the existing cross section with minimal widening to Washington, D.C., 2004.
accommodate a small channelizing median for the left turn. This 3. Hughes, W., and R. Jagannathan. Alternative Intersections/Interchanges:
Information Report (AIIR). FHWA-HRT-09-060. FHWA, U.S. Department
unique design fit the roadway within its specific needs and con-
of Transportation, 2009.
straints. The cost for the proposed improvements was estimated at 4. Robinson, B., L. Rodegerdts, W. Scarborough, W. Kittelson, R. Troutbeck,
$275,000, providing a savings of more than $4.5 million for the W. Brilon, L. Bondzio, K. Courage, M. Kyte, J. Mason, A. Flannery,
intersection. Detailed operational analysis of the intersection was E. Myers, J. Bunker, and G. Jacque Mart. Roundabouts: An Informational
estimated at LOS B, which exceeded the target value LOS D or E Guide. FHWA-RD-00-067. FHWA, U.S. Department of Transportation,
2000.
for the initial alternatives. 5. Messer, C. J., and B. D. Fambro. Critical Lane Analysis for Intersection
Design. In Transportation Research Record 644, TRB, National Research
Council, Washington, D.C., 1977, pp. 26–35.
Conclusion 6. Special Report 209: Highway Capacity Manual. TRB, National Research
Council, Washington, D.C., 1985.
7. Rodegerdts, L., M. Blogg, E. Wemple, E. Meyers, M. Kyte, M. P. Dixon,
This paper documents the efforts and approach used to develop the
G. F. List, A. Flannery, R. Troutbeck, W. Brilon, N. Wu, B. N. Persaud,
operational component of IDAT. The tool and this approach provide C. Lyon, D. L. Harkey, and D. Carter. NCHRP Report 572: Roundabouts
a method to evaluate 13 different intersection design alternatives in the United States. Transportation Research Board of the National
with more than 12,000 unique lane configurations each. The ultimate Academies, Washington, D.C., 2007.
applicability of this approach is to identify a wider range of feasible 8. Manual on Uniform Traffic Control Devices. FHWA, U.S. Department of
Transportation, 2009.
intersection design alternatives with significantly less effort than is 9. DLZ Kentucky Inc. Madison Pike (KY 17) Intersection Improvement
currently afforded through the independent evaluation of intersec- Study. Northern Kentucky Area Planning Commission and Kentucky
tion designs through capacity software or microsimulation programs. Transportation Cabinet, Fort Wright, Ky., 2006.
Notable is the approach taken to sizing intersection alternatives to
deliver a targeted performance level so that comparative evaluations The Operational Effects of Geometrics Committee peer-reviewed this paper.