ROUNDABOUT ~’

~DESIGN
GUIDELINES

STATE Of MARYLAND
DEPARTMENT OF TRANSPORTATbN
STATE HIGHWAY ADMINISTRATION
 ROUNDABOUT DESIGN GUIDELINES

LIST OF FIGURES

LIST OF TABLES

FORWARD i

1.0 INTRODUCTION l-2

2.0 USE OF ROUNDABOUTS 3-5

3.0 PERFORMANCE OF ROUNDABOUTS 6 - 31

4 GEOMETRIC DESIGN 32 - 44

5.0 LANDSCAPE DESIGN 45 - 48

6.0 SIGNING AND PAVEMENT MARKING 49 - 54

~ 7.0 LIGHTING 55

8.0 PEDESTRIAN AND BICYCLE CONSIDERATIONS 56 - 59

9.0 WORK ZONE TRAFFIC CONTROL 60 - 64

APPENDIX

BENEFIT/COST ANALYSIS I - IV

REFERENCES
 LIST OF FIGURES
Figure 1.1 Yield at entry, Deflection, and Flare 1
Figure 3.1 Typical turning movement diagram 8
Figure 3.2 Roundabout entry and circulating tlows 8
Figure 3.3 Required number of entry and circulating lanes 9
Figure 3.4 Entry capacity fat a single lane roundabout with a 13 foot wide 17
entry lane and one circuiating lane
Figure 3.5 Entry capacity for a roundabout with two 13 foot wide entry lanes 18
and two circulating lanes
Figure 3.6 Proportion of vehicIes stopped on a singie lane roundabout 25
Figure 3.7 Proportion of vehicles stopped on a multi-lane entry roundabout 25
Figure 3.8 Definitions of the Terms used in Tables (a) and (b) 26
Figure 4.1 Flare design at entry 32
Figure 4.2 Typical Roundabout Entrance/Exit Conditions for Urban Areas 33
Figure 4.3 Oval roundabout 33
Figure 4.4 Turning templates for over-dimensional vehicles 34
Figure 4.5 Typical Rural Roundabout Design (with High Speed Approach 35
Roads)
Figure 4.6 Alternative methods for providing vehicle deflection (not to scale) 38
Figure 4.7 Illustration of the detlection criteria for a single lane roundabout 39
Figure 4.8 Illustration of the deflection criteria for a multi-lane roundabout 39
Figure 4.9 Sight distance requirements 41
Figure 4.10 Roundabout on a road with a very wide median 43
Figure 5.1 Example of the use of landscaping to reinforce the funnelling 46
effect at the entrance to roundabouts
Figure 5.2 Typical Section of the Truck Apron 47
Figure 5.3 Plan of Central Island 48
Figure 6.1. Typical signing for a state route roundabout 51
Figure 6.2 Typical signing for a local road roundabout 52
Figure 6.3 Typical pavement markings for roundabouts .54
Figure 8.1 Examples of Pedestrian Crossings 58
Figure 8.2 Example of a special bicycle facility 59
Figure 9.1 Roundabout workarea pavement markings 61
Figure 9.1. Work Zone Traffic Control 62
Figure 9.3 Roundabout workarea delineation 64

LIST OF TABLES

Table 3.1 Dominant Stream Follow-up Headways (tr,). (Initial values) in 11

seconds.
Tabie 3.2 Ration of the Critical Acceptance Gap to the Follow-up Headway 12
(LlkJ
Table 3.3 Adjustment Times for the Dominant Stream Follow-up Headway 14
Table 3.4 Sub-dominant Stream Follow-up headway t, 14
Table 3.5 Average headway between bunched vehicles in the circulating 16
traffic (r-) and the number of effective lanes in the circulating
roadway.
Table 3.6 Proportion of Bunched Vehicles, 0 19
Table 3.7(a) Geometric Delay for Stopped Vehicles (Seconds per vehicle) 22
Table 3.7(b) Geometric Delay for Vehicles Which Do Not Stop (Seconds per 23
vehicle)
Table 3.8 Typical Casualty Accident Rates for Different Urban 27
Intersections with Moderate to High Volumes in Victoria,
Australia
Table 4.1 Deflection Curve Radii 37
Table 4.2 Approach Sight Distance (ASD) 40
FORWARD

The mission of the Maryland State Administration is to build and maintain a safe and efficient
highway system. There is evidence that roundabouts reduce accidents. Maryland has a highway
system that is well maintained, well planned, well organized and has a steadily declining accident
rate. The future holds many challenges for the Administration in continuing to fulfill its mission.
These chaknges include performing the same or more work with less staff, maintaining a high
level of service with limited resources, enhancing the environment, preserving the highway
system and fulfUng the requirements of the Intermodal Surface Transportation Act of 1992
(ISTEA). Given the above, we must constantly review m ideas, processesand technologies
while remaiuin g focussed in accomplishing our mission. This guide brings us one step closer
to where we want to be.

This guide was developed to set forth a standard approach to the planning, design, and
construction of roundabouts in the State of Maryland, given that no federal guidelines exist. This
text borrows information from recognized experts in the planning, design and construction of
ro~dabouts; namely the Australian Design Guide (3). Generally, major conceptsfor safety and
design should follow AASHTO Design Guide. This guide supplementsthese fderal Guidelines
until such time when formal guidelines are established.

We believe Austroads is a leader in the planning, design and implementation of roundabouts.

We commend Australia, the European Community and others for recognizing the true benefits
of roundabouts early and enhancing their design such that they operate safely and efficiently.
We have selected the Australian Design guide as a model because its design procedure most
closely represents current procedures already adopted by the Maryland State Highway
Administration. We would like to extend our sincere thanks to those individuals at Au&roads
responsible for developing “Guide to Traffic Engineering Practice - “Roundabouts”. Most
importantly we thank them for allows us to reprint much of what they had developed.

It is intended that this u be used to standardize the approach to roundabouts in Maryland.

It is only a guide. It is being issued as an interim document for one year from April 1,
1995. We expect many changes in the upcoming months and subsequent years to cane.
We encourage users to suggestimprovements and send them to :

Mr. Thomas Hicks, Director

Office of Traffic & Safety
7491 Connelly Drive
Hanover, Maryland 21076
1.0 INTRODUCTION

Well designed roundabouts have proven to be safe and efficient forms of inteisection control
in the countries that have adopted modern guidelines. These countries include Great Britain,
Australia, France, Germany, Spain, Norway, The Netherlands, among other countries.

These guidelines are intended to be temporary and updated as ofteq as is necessary. Design
guidelines from other countries form the basis of this document. Over time, these guidelines
will be analyzed as to their applicability to the driving conditions in Maryland. Input into these
guidelines from users is encouraged.

Roundabouts operate by gap acceptance, in that approaching drivers must give way to
circulating traffic in the roundabout. The proven safety performance of most roundabouts is due
to the low relative speeds of all vehicles and the relative simplicity of decision making required
of drivers.

Conditions will be encountered wherein the procedures highlighted in these guidelines cannot
be fully implemented. It is expected that the designer make modifications as necessary while
ensuring the major concepts of safety and design.

The aims of these guidelines are as follqws:

a. to give guidance on where roundabouts may be used

b. to describe the performance and operation of roundabouts
C. to give guidance on design standards for roundabouts so that high standard and
uniform design will be encouraged.

The designer of a modem roundabout should be fully aware of the difference between a
roundabout and a traffic circle. Basically, there are three main differences; Yield-at-entry,
Deflection, and Flare. These are illustrated in Figure 1.1..‘.

FIGURE 1.1 YieId at entry, Deflection, and Flare

Yield-at-entry allows vehicles in the roundabout to continue through the roundabout to their
appointed exit and. eliminates gridiock which occurs when entering vehicles are afforded the
right-of-way. Yield-at-entry also enables the traffic engineer to design smaller roundabouts.

1
 Deflection is the physical slowing of vehicles through the roundabout which is achieved by
causing the driver to curve around thexentral island. Deflection increases the safety of the
intersection by lowering the entry and circulating speeds.

Flare is the widening of the approach to the roundabouts to increase capacity.

2.0 USE OF ROUNDABOUTS

2.1 GENERAL

Roundabouts should be considered at a wide range of intersection types including but

not limited to; freeway terminal interchanges, state route intersections, and state
route/local route intersections. Roundabouts perform better at intersections with roughly
similar traffic volumes and at intersections with heavy left turning movements.
Roundabouts can improve safety by simplifying conflicts, reducing vehicle speeds and
providing a clearer indication of the driver’s right-of-way compared to other forms of
intersection control.

2.2 SITE SELECTION CRITERIA

The following site selection guidelines are intended as general guidelines only. The
designer should determine the applicability of a roundabout at a particular intersection
by considering the following items:

- Capacity Analysis of all methods under consideration

- Cost/Benefit Analysis
- Percentage of Truck Traffic
- Bicycle and Pedestrian Traffic
- Right-of-Way Consideration
- Parking Requirements
- Compatibility with Adjacent Intersection
- Safety Aspects
- Effect of Possible Traffic Growth
- Speed of Traffic
- Installation and Maintenance Costs

ROUNDABOUT SHORTLISTING GUIDELINES

Introduction

The following guidelines are based on existing design manuals from England, Australia, and
other countries, and video tapes of existing roundabouts. The guidelines are not meant to be
rigid but should be used in conjunction with engineering judgement, and traffic analysis. For
example, it could be stated that a roundabout should not be placed where there is an existing .
signal in close proximity (i.e. Chevy Chase Circle) because the queues from the signal may
extend temporarily into the roundabout. Intuitively, this would not seem to be an appropriate
place for a roundabout, however traffic analysis may indicate that a roundabout may work better
than any other solution. The proposed intersection treatment, therefore, should be chosen based
on the advantages/8disadvantages, benefits/costs that it provides.

3
Location

l High Accident Location (with left turn or right angle accidents)

0 Capacity/Delay Problem Intersection

l Intersection in which traffic signal was requested but not warranted.

l 4-Way Stops

Traffic Volume and Composition

l Heavy Delay on Side Street

l Flow Distribution with Heavy Left Turn Movement (makes signals less efficient -no
impact on roundabout)

0 DHV of 7000 or Less (initially)

Right-of-Wav

l Generally take no more right-of-way than comparable solution using signals:

Appropriate Sites for Roundabouts

0 Heavy delay on minor road.

0 Traffic signals result in greater delay.

0 Intersection with heavy left turning traffic.

0 Intersection with more than four legs or unusual geometry.

0 At rural intersections (including those in high speed areas) at which there is an accident
involving crossing traffic.

0 Where major roads intersect at “Y” or “T” junctions.

0 At locations where traffic growth is expected to be high and where future traffic patterns
are uncertain or changeable.

0 At intersections where U-turns are desirable.

0 At Freeway Interchange Ramps.

0 High accident intersection where right angle accidents are prominent.

4
Inannronriate Site for Roundabouts

Where a satisfactory geometric design cannot be provided.

Where a signal interconnect system would provide a better level of service.

Where it is desirable to be able to modify traffic via signal timings.

Where peak period reversible lanes may be employed.

Where the roundabout is close to existing signals and queueing from the signal could be
a problem.

2.3 COST/BENEFIT ANALYSIS

A cost/benefit analysis shall be completed for all intersections in which a roundabout

is being considered. See procedure outlined in Appendix A.

5
3.0 PERFORMANCE OF ROUNDABOUTS

3.1 GENERAL

Until more data is gathered concerning the performance of roundabouts in Maryland,

the Maryland State Highway Administration recommends that designers use the
Australian practice at this time.

Australian practice for determining the capacity and delay of roundabouts is based in
gap acceptance theory, and the techniques have been researched in Horman and Tumbull
(1974), Avent and Taylor (1979) and Troutbeck (1984, 1986 and’1990). The most recent
methodology developed by Troutbeck (1989), uses the empirical results of field
observations made in four Australian capital cities.

These field studies identified a number of driver behavior aspects that affect the
analysis of capacity and delay. These are:

0 That entering vehicles generally give way to all circulating vehicles. Entering
drivers are often unsure that a circulating driver to their left may intend to leave
at the next exit and travel across their paths. Consequently, entering drivers tend
to give way to all circulating vehicles, even where the circulating roadway is two
or more lanes wide. An exception is when vehicles are entering from an
auxiliary “right turn only” lane. If this auxiliary lane and the entry curve is
designed so that entering drivers are protected from the circulating traffic, they
will generally proceed without “giving way” to any circulating vehicle.

0 That at multi lane entries, vehicles are prepared to enter simultaneously alongside
other entering vehicles at the same approach.

l That drivers entering in different lanes of the same approach will behave
differently.

0 That exiting vehicles have no effect on drivers entering at the same leg unless
the negotiation speed is high or the roundabout is small and the entering drivers
have difficulty in determining whether a vehicle is exiting or not.

These findings influence the capacity and delay calculations. The principal departure
from the 1986 Australian guide is that the drivers in each entry lane on a particular
approach behave differently. This means that each entry lane will have a different
capacity and vehicle delay. As a consequence, if the number of entry lanes is doubled
then the capacity is not quite doubled.

The usual terms used to define gap-acceptance behavior are the critical acceptance
gap, t,, and the follow-up time, tf. The critical acceptance gap is the minimum
acceptable gap that will be acceptable to a homogeneous and consistent population of
drivers. The follow-up time is the minimum headway between minor stream vehicles
which enter in the longer gaps in the circulating traffic. In both cases the units are in

6
 seconds. The gap acceptance terms represent the average for all drivers as the average
predicted capacity and average delay values are required. In the theory, it is assumed
that all drivers will accept a gap greater than the critical acceptance gap. It is also
assumed that drivers are consistent and behave exactly the same each time a gap is
offered. That is, a driver does not reject a gap only to accept a shorter one later.
However in practice, drivers are different from one another and often act inconsistently
because they are not always able to make accurate assessments of gap durations. Also,
drivers do occasionally reject a gap then accept a shorter one. Calculations based on
these assumptions nevertheless give estimates of capacity which are reasonably consistent
with observations (Troutbeck 1989, Catchpole and Plank, 1986 and Plank and Catchpole,
1984).

As the drivers in each entry lane behave differently, each entry lane will be given
different critical gap and follow-up headway parameters.

3.2 TRAFFIC BALANCE

Roundabouts operate best when the traffic flows are balanced. This does not mean
that all movements must be of the same magnitude but simply that the predominant
movements are “broken up” by circulating traffic so that gaps are provided to allow
vehicles waiting on adjacent legs to enter the roundabout without major delays.

3.3 ANALYSIS OF THE CAPACITY OF ROUNDABOUTS

This section provides an analytical technique which can be expected to give quite
accurate results which reflect current Australian experience and practice. SIDRA
Software is recommended and is available through McTrans at the University of Florida.

In situations where a high degree of accuracy is not required, Figures 3.3, 3.4 and
3.5 may be used to obtain general estimates of the capacity of a roundabout.

3.3.1 PROCEDURE

The capacity .of a roundabout is influenced by its geometry through the critical gap
parameters. The procedure for capacity analysis of each approach is as follows:

Assemble Traffic Data

Cyclic and stochastic variations in traffic flows should be taken into account when
assembling the traffic data into the turning movement flows to be used in the analysis.

Figures 3.1 and 3.2 show the conversion of typical traffic turning movements at a
cross-road type intersection into entry and circulating flows on a roundabout.

7
 L II 1 I

FIGURE 3. I Typical turning FIGURE 3.2 Roundabout entry

movement diagram and circulating flows

Where the truck flows are less than 5 percent the total vehicle flow is considered to be
passenger car units (pcu’s). For truck volumes greater than 5 percent the truck flows should be
converted to passenger car units. A single unit truck is assumed to be equivalent to 2 pcu’s and
an articulated vehicle, 3 pcu’s. Equivalencies for other vehicle types (such as bicycles or large
combination vehicles) may be estimated and used if necessary.

Number of Entry and Circulating Lanes

The number of entry lanes will generally be determined from the number of lanes on
the approaches. However, an entry may be widened or flared, particularly if there
are heavy turning movements.

It is usually assumed that the number of circulating lanes will equal the number of
entry lanes at any approach. This assumption can be relaxed later if required.

Figure 3.3 is a plot of approach (entry) flows and circulating flows and the
acceptability of a single or multi-lane roundabout. The shaded bands indicate the
limits for a single lane roundabout and a two lane roundabout. For instance, if
conditions at a roundabout give a point in the upper shaded area of Figure 3.3, then
a two lane roundabout may be acceptable depending on the entry geometry and the
acceptable degree of saturation. The user may need to evaluate both a two lane and
a three lane roundabout in this case. Figure 3.3 is based on the acceptable degree
of saturation being less than 0.8.

8
 2500
* entry an
clrcul atlng
Three lane * carriageway
roundabout 1

FEY
(veh/h)
Two lane l
roundabout

” I I 11-1 ’ I ’ I I I
0
8 8 8
E Ti z E
- Ts N is iz

Circul sting FI ow (veh/h)

FIGURE 3.3 Required number of entry and circulating lanes

Record the geometric values

From the functional plans of the roundabout or from actual measurement, record the values
for:

0 the inscribed diameter, Di.

l The number of entry lanes, %.

n, is 1 for entry widths less than 18 feet
n, is 2 for entry widths between 18 and 33 feet
n, is 3 for entry widths greater than 33 feet

l the number of circulating lanes, n,.

n, is equal to 1 for circulating roadway widths less than 33 feet
n,is 2 for widths greater than or equal to 33 feet and less than 50 feet
n, is 3 for widths greater than 50 feet.

9
NOTE: For some circulating roadways between 25 feet and 33 feet wi’de and with
circulating flow rates greater than 1000 veh/h, there may be two effective lanes and
n, may be set to 2. See the discussion later in this Section.

l the average entry lane width, (or the entry width divided by the number of the
entry lanes).

Refer to Section 4 for a procedure for the geometric design of roundabouts.

Classify the entry lane type

Classify the entry lanes as either dominant or sub-dominant. Where there are two
or more entry lanes, one entry lane “dominates”. That is the drivers in this lane tend
to influence the behavior of drivers in other entry lanes at the approach. The entry
lane with the greatest flow is chosen to be the dominant stream; other entry lanes will
then be sub-dominant. If there are 3 entry lanes, two will be sub-dominant and only
one will be a dominant stream. If there is only one entry lane at an approach then
this lane is considered as a “dominant” lane (Troutbeck 1989).

Estimate the critical gap acceptance parameters

Gap acceptance parameters are affected by geometry of the entry. Geometries which
offer an easier entry path give lower gap acceptance values. These parameters are
also a function of the circulating flow. At higher circulating flows, the circulating
speeds are lower and drivers are more willing to accept smaller gaps. Also at higher
circulating flows, more circulating drivers slow and allow entering drivers to move
in front of them. This leads to priority sharing or even a reversal of priority.
Further discussion of the interactions is given in Troutbeck (1989 and 1990).

10
 Circulating flow (veh/h)

From Troutbeck (1989)

Note: The values of the follow-up headway are given to two decimal places to assist in
interpolation. The adopted value may be rounded to one decimal place.

Flows above about 1700 vph are not applicable to single lane circulating roadway
(shaded area in table).
’ The ratio of the critical acceptance gap to the follow-up headway (t&J, is given in
Table 3.2. The critical acceptance gap is the product of the appropriate values from
Table 3.1 and Table 3.2.

TABLE 3. I Dominant Stream Follow-up Headways ($A.

(Initial values) in seconds.

11
Number of
circulating
lanes one more than one

Average
entry lane
width (ft) 10 13 16 10 13 16

Circulating
flow
(veh/h)
0 2.32 1.98 1.64 2.04 1.70 1.36
200 2.26 1.92 1.58 1.98 1.64 1.30
400 2.19 1.85 1.52 1.92 1.58 1.24
600 2.13 1.79 1.45 1.85 1.51 1.18
800 2.07 1.73 1.39 1.79 1.45 1.11
1000 2.01 1.67 1.33 1.73 1.39 1.10
1200 1.94 1.60 1.26 1.67 1.33 1.10
1400 1.88 1.54 1.20 1.60 1.26 1.10
1600 1.82 1.48 1.14 1.54 1.20 1.10
1800 1.48’ 1.14 1.10
2000 1.41 1.10 1.10
2200 1.35 1.10 1.10
2400 1.29 1.10 1.10
2600 1.23 1.10 1.10

From Troutbeck (1989)

’ For single lane circulating roadways, if the critical gap calculation from Tables 3.1
and 3.2 is less than 2.1 s, use 2.1 s.
o For multi-lane circulating roadways, the minimum value of critical gap should be
1.5 s.

NOTE: Values of the ratio may be interpolated for intermediate widths of entry lane.

TABLE 3.2 Ratio of the Critical Acceptance Gap to the Follow-up Headway (t&fd)

12
(4 For a single lane entry

Table 3.1 lists the dominant stream follow-up headway (tf,). If there is one
circulating lane (nC= l), these values are used for the entry stream. If there are 2 or
more circulating lanes (n, =2 or 3), then the values in Table 3.1 should be increased
by 0.39.

The ratio of the critical acceptance gap to the follow-up headway (t&J, is given in
Table 3 -2. The critical acceptance gap is the product of the appropriate values from
Table 3.1 and Table. 3.2.

For Multi-lane Approaches

To estimate the entry lane flows at approaches with two or more lanes, it can be
assumed that drivers wishing to turn right will use the right hand entry lanes and the
drivers turning left will use the left hand lanes. However in some situations lanes
may be marked with signs or pavement arrows to restrict them to particular traffic
movements and the lane arrangement so marked would be used in the analysis. The
through traffic then needs to be proportioned to the appropriate lanes to finalize the
lane entry flows. While the above provides the most accurate assessment, it is
pointed out that estimates of approach capacity are not significantly affected by the
distribution of traffic in the lanes.

The entry lane with the greatest flow at an approach is termed the “dominant” lane
and traffic in this lane is termed the dominant stream. Other lanes contain
subdominant streams.

The critical gap parameters for an approach with two or more entry lanes are
estimated using Tables 3.1, 3.2, 3.3 and 3.4.

Table 3.1 gives values for the follow-up headway for the dominant stream. These
values. are adjusted if the number of entry lanes differs from the number of
circulating lanes. The adjustment values are given in Table 3.3.

Table 3.4 gives the values of the sub-dominant stream follow-up headway (t;) as a
function of the dominant stream follow-up headway (Q and the ratio of dominant
stream entry flow to the sub-dominant stream entry flow.

The critical acceptance gap values for each lane are given by the product of the
follow-up headway (from Tables 3.1 and 3.4) and the ratios in Table 3.2. As stated
above, critical acceptance gap values need to be calculated separately for each entry
lane. Refer to Troutbeck (1989) for an example of these calculations.

13
 Number of
circulating lanes Number of entrv lanes

Note: Add or subtract these factors from the initial values from
Table 3.1.

TABLE 3.3 Mjustment Times for the Dominant

Stream Follow-up Headway

From Trou tbeck (I 989)

TABLE 3.4 Sub-dominant Stream Follow-up headway tfi

Estimate the characteristics of the circulating traffic

As the entering drivers give way to all circulating vehicles, the circulating traffic can be
considered as if it were all in one lane. There are, however, circulating stream characteristics
that change with flow and the number of circulating lanes.

The greater number of circulating lanes, the shorter will be the average headway between
bunched vehicles in all lanes. If there are two or more circulating lanes then the average
headway (r) between bunched vehicles is about 1 s and if there is only one lane the average
headway is 2 s.

If a circulating roadway equal to or greater than 33 feet wide carries a circulating flow
greater than 1000 veh/h it can be assumed to effectively operate as two streams and the average
headway between bunched vehicles (z) will be 1 s. (see Table 3.5). Under these conditions the
vehicles might travel in an offset pattern and users should decide whether or not the circulating
roadway will be considered to have one or two effective lanes. It may be preferred to consider
all single lane roundabouts to have only one effective lane regardless of the circulating flow and
hence an average headway between bunched vehicles of 2 s. This action would be conservative.
Note that if it is considered that there will be two effective circulating streams, then the number
of circulating lanes (nJ should be set to 2. Table 3.3 may then need to be consulted when
estimating the follow-up headways.

The operation of the circulating stream also affects the average percentage of vehicles which
are in bunches. As the flow increases, more vehicles are in bunches.

The proportion of bunched vehicles, (O), is evaluated from the circulating flow, the number
of effective circulating lanes (characterized by the average headway between bunched vehicles)
and the proximity of the roundabout to signalized intersections or other situations which increase
bunching. Troutbeck (1989) gives equations for estimating the proportion of free vehicles, i.e.
those not in bunches. Values for the proportion of bunched vehicles have been developed from
these equations and the revised values are listed in Table 3.6. (Also see Akselik and Troutbeck
(1991)). It is suggested that the values given in this Table be then adjusted according to the
proximity of the roundabout to nearby signalized intersections or other situations which will
influence the approaching traffic conditions and the circulating flow at the roundabout. Values
should be increased or decreased by no more than 0.2 based on judgement of the extent of
bunching caused.

The proportion of bunched vehicles is expected to range from 0 for random traffic to about
0.8 for heavily platooned traffic. Values as high as 0.8 to 0.9 have been observed in extreme
cases. This is equivalent to an average platoon length of 1 to about 3 or 4 vehicles in most
conditions and up to 10 vehicles under the worst conditions.

15
 TABLE 3.5 Average headway between bunched vehicles in the circulating trafic (z)
and the number of effective lanes in the circulating roadway.

Calculate Absorption Capacity and Degree of Saturation

The absorption capacity of each entry lane is calculated from the entry lane gap acceptance
parameters (t, and tJ applicable to the dominant lane and to each sub-dominant entry lane and
the circulating flow characteristics (Qc, z, and e). The appropriate equation is:

c= 3600 (1-e) qce-"'ta-T'

pe-"tr
(3.1)

Where:

C = the absorption capacity of an entry lane in veh/h

e = the proportion of bunched vehicles in the circulating streams

qc= the flow of vehicles in the circulating streams in veh/s

t, = the critical acceptance gap relevant to the dominant or sub-dominant lanes respectively.

tr = the follow on headway relevant to the dominant or sub-dominant lanes respectively.

r = the minimum headway in the circulating streams, and these are related by:

16
 A- w-wqc (3.2)
1-zc;Tc
Note that the capacity predicted by Equation 3.1 is the expected steady-state capacity, or the
maximum entry flow rate and it is not the “practical capacity”. See discussion on degree of
saturation below.

The above analysis method and equations are more comprehensive than may be necessary
for some purposes. Figures 3.4 and 3.5 may be used to obtain a quick estimate for use in the
planning and preliminary layout of a roundabout at a particular site.

Figure 3.4 refers to a single lane roundabout with a 13 foot wide entry lane and one
circulating lane. The results in Figure 3.5 reflect the operating conditions of a roundabout with
two 13 feet wide entry lanes and two circulating lanes,

1000
Entry
zx Y
500

Circulating Flow (veh/h)

FIGURE 3.4 Entry capacity for a single lane roundabout with a 13 foot wide entry lane
and one circulating lane

17
 2500

2000
Entry
$P$c~y 1500
Both lanes
(veh/h) 1000

500

Circulating Fl6w (veh/h)

FIGURE 3.5 Entv capacity for a rounahbout with two 13 foot wide entry lanes and two
circulating lanes

For very high circulating flows (exceeding about 1700 veh/h for single lane circulating flows,
or about 3400 veh/h for multi-lane circulating flows), the entry capacities from equation 3.1
approach zero. In such cases, a minimum entry capacity may be assumed.

Calculate Degree of Saturation

The degree of saturation of an entry lane is the arrival flow divided by the entry (absorption)
capacity of the lane:

x=Q, (3-3)
C

Where:

Q, = entry lane arrival flow in vehk, and

C = entry lane capacity in veh/h (from equation 3.1 or figures 3.4 and 3.5)

18
 The degree of saturation during the design period for an entry lane should be less than about
0.8 to 0.9 for satisfactory operation, although this may not always be practicable.

Within this range of degree of saturation, designers should consider using the delays as a
more appropriate measure of performance.

The maximum (practical) degree of saturation corresponds to the concept of “practical

capacity “. For example, if practical degree of saturation (x) is 0.85, practical capacity is 0.85
C, where C is the entry capacity from equation 3.1. The practical degree of saturation is also
used for “spare capacity” calculations.

Judgement may be exercised in the assessment of the acceptability of the degree of saturation
or delays taking into consideration factors such as environment, locality, possible alternative
intersection treatments, cost and the period that the roundabout can be expected to have less
satisfactory performance characteristics than an alternative treatment.

Number of effective
circulating lanes one more than one
Average headway
between bunched vehicles 2.0 1.0
z9 (9
Circulating flow (veh/h)
0 0.250 0.250
300 0.375 0.313
600 0.500 0.375
900 0.625 0.438
1200 0.750 i, 0.500
1500 0.875 0.563
1800 1.000 0.625
2000 0.667
2200 0.708
,240O 0.750
2600 0.792
q

TABLE 3.4 Proporh’on of Bunched Vehicles, Q

19
3.4 Delays at Roundabouts -.

There are two components of the delays experienced at roundabouts, namely queuing and
geometric delay.

Queuing delay is the delay to drivers waiting to accept a gap in the circulating traffic.

Geometric delay is:

(0 The delay to drivers slowing down to the negotiation speed, proceeding through
the roundabout and then accelerating back to normal operating speed; or

(ii) The delay to drivers slowing down to stop at the end of the queue and, after
accepting a gap, accelerating to the negotiation speed, proceeding through the
roundabout and then finally accelerating further to reach normal operating speed.
It excludes the time to wait for an acceptable gap.

In some instances it may be appropriate to consider only the queuing delay, e.g. when
approximate results only are required, or when making a comparison with a “STOP” or
“YIELD” controlled approach at an intersection. In these cases, the geometric delay for
traffic entering from the side (controlled) road approach would experience about the same
geometric delay as at a roundabout. In most cases it may be desirable to consider the total
delay e.g. when the results are required for a comparison with traffic signals or in an
economic analysis.

Total delay is the sum of the queuing delay and the geometric delay.

3.4.1 Queuing Delay

To calculate the average queuing delay, first calculate the minimum delay for the
conditions when the entering traffic flow is very low using:

w_e*(t,-d-t i hT2-2te
II--(l-ws, =x+2 (nz+l-43)
where the gap acceptance parameters, t,, z, 8 and A are as in Equation 3.1 and the
circulating flow qCis in veh/sec.

For all practical purposes the average queuing delay per vehicle is given by:

w,+w*+900T[Z+ z2+
{73

20
where :
*.
w, = average delay per vehicle in seconds

Wh = minimum delay in seconds when entering traffic is very low (from Equation 3.4)

T= duration of the flow period in hours, i.e. the time interval during which an average
arrival demand Q, persists (use 1 h or 0.5 h)

z = x-l

X = degree of saturation of the entry lane (= Q,/C as in Equation 3.3)

c= entry lane capacity in vehicles per hour

m = a delay parameter given by

m = Wb Cl450

The second term of Equation 3.5 accounts for the queuing delays due to the presence of
a queue in the entry lane. Equation 3.5 is a time-dependent formula (Akqelik 1991, Akcelik
and Troutbeck 1991) derived from the steady-state formula given by Troutbeck (1989). It
is applicable for near-capacity and oversaturated conditions. The flow period parameter
becomes important for high degrees of saturation, i.e. the delays are insensitive to the flow
period for low degrees of saturation.

21
 Negotiated speed through roundabout
Vn (MPH)

* Refer to Figure 3.8 for the definitions of the dimensions.

TABLE 3.7(a) Geometric Delay for Stopped Vehicles (Seconds per vehicle)

22
Approach Distance*
Speed V, around Negotiated speed through roundabout
N-9 roundabout Vn (MPH)
D (ft)

9 12 16 19 22 25 28 31

25 65 7 4 2 1 0
25 195 17 11 7 4 0
25 325 19 13 8 4
25 460 13 8
25 590 12

37 65 11 8 5 4 3 2 1 1
37 195 20 15 11 8 4 2 1 1
37 325 22 17 13 9 5 1 1
37 460 17 13 8 4 1
37 590 17 12 .7 2

50 65 14 11 9 7 6 5 4 3
50 195 24 19 15 11 8 5 4 3
50 325 26 20 16 13 9 5 3
50 460 21 17 13 9 4
50 590 21 16 12 7

62 65 18 15 12 10 9 8 7 6
62 195 27 22 18 15 12 9 7 6
62 325 29 24 20 16 13 10 6
62 460 25 20 17 13 12
62 590 25 20 16

4
* Refer to Figure 3.9 for the definitions of the dimensions.

TABLE 3.7(b) Geometric Delay for Vehicles which Do Not Stop (Seconds per vehicle)

23
3.4.2 Geometric Delay

The geometric delay for vehicles differs depending on whether the vehicles have to stop
or not. George (1982) developed a method for calculating the average geometric delays as
follows:

Average Geometric delay:

d, = P, d, + (l-P,) d, (3.6)

where :

ps = the proportion of entering vehicles which must stop,

4 = the geometric delay to vehicles which must stop,

(l-P,) = the proportion of entering vehicles which need not stop,

d, = is the geometric delay to vehicles which need not stop

This equation has also been documented by Middleton (1990).

The proportion of entering vehicles which must stop, P,, can be estimated using Figures
3.6 and 3.7 depending on the number of circulating lanes. This proportion depends on the
entry and circulating lane flows. Increase either of these flows and the proportion of
entering drivers stopped will increase. The near linear lines in these Figures result from the
gap acceptance parameters and the level of bunching in the circulating stream being a
function of the circulating flow.

Tables 3.7(a) and (b) have been developed to allow d, and d, to be estimated. These
enable the geometric delay to be calculated for each approach to a roundabout.

Geometric delay is different for each traffic movement - left turn, right turn and straight
on, at each approach and each should be calculated separately.

3.4.3 Total Average Delay

Total average delay is the sum of the queuing delay and the geometric delay. Again the
total delay will not be the same for vehicles making various turns and using different entry
lanes. The total average delay per vehicle from an approach must then be estimated using
the proportion of vehicles making each movement and their respective delays.

24
 I
8
s!
Circulating Flow (vehlh)

FIGURE 3.6 Proportion of vehicles stopped on a single lane roundabout

Proportion
3f vehicles
stopped

Circulating Fldw (veh/h)

FIGURE 3.7 Proportion of vehicles stopped on a multi-lane entry

roundabout.

25
 at negotiation speed

Negotiation speed ;

“,mph
~~~~~~
FIGURE 3.8 Definitions of the Terms used in Tables (a) and (b)

3.5 Entry Queue Lengths

The average entry queue length, 12y,under steady state under-saturated conditions is given
by the product of the average queuing delay, w,, and the entry lane flow, Q,.

n, = w, Q, (for x =Z1) (3-7)

As a rule of thumb, the 95 percentile queue length is three times the average, i.e. a
queuing space for 3n,., vehicles will be exceeded about 5 per cent of the time. Values of
queuing space should be rounded up to the next vehicle.

Note that the distribution is not a Normal distribution and the usual limits do not apply.

3.6 Safety of Roundabouts

The safety performance of roundabouts has been documented in a number of Australian

and UK studies. “Before” and “after” type accident studies carried out at intersection
involving a wide range of site and traffic conditions at which roundabouts have been
constructed, indicate very significant reductions in casualty rates.

26
Intersection Treatment Mean Casualty Typical Range of
Accident Rate Casualty Accident
Rates
T-Intersections -Unsignalized 1.5 1.3 - 1.7
-Signalized 1.4 1.2 - 1.6

Cross-intersections -Unsignalized 2.4 2.1 - 2.7

-Signalized 1.7 1.6 - 1.8

Multi-leg Intersections -Signalized 3.2 2.8 - 3.6

Roundabouts (high volumes) 0.8 0.6 - 1.1

Roundabouts (low volumes) 0.4 0.1 - 1.0

TABLE 3.8 Typical Casualty Accident Rates for Diflerent Urban Intersections
with Moderate to High Volumes in Victoria, Australia

The previous tabulation (Table 3.8) illustrates the result of comparative studies carried
out in Australia. Similar results have been obtained in the UK.

The good safety record of properly designed roundabouts can be attributed to the
following factors:

The general reduction in conflicting traffic speeds (limited to less than 30 MPH)
passing through the intersection on all legs.

Elimination of high angles of conflict thereby ensuring low relative speeds between
conflicting vehicles.

Relative simplicity of decision making at the point of entry.

On undivided roads, in high speed areas, long splitter islands provide good “advance
warning” of the presence of the intersection.

Splitter islands provide refuge for pedestrians and permit them to cross one direction
of traffic at a time.

Roundabouts always require a “conscious action” on the part of all drivers passing
through the intersection, regardless of whether other vehicles are present or not.

An understanding of the above factors and their implications, in respect to the geometric
design of roundabouts is essential to the full achievement of the safety benefits of
roundabouts.

27
 The safety record of roundabouts with more than three circulating lanes has not been well
established. Maycock and Hall (1984) analyzed the influence of the geometry of roundabouts
on their accident performance. They did not find circulating roadway width to be a
significant factor. The accident potential at roundabouts with 3 circulating lanes would be
influenced by the drivers’ entry curvature. At this time, there is insufficient data to quantify
the safety performance of roundabouts with 3 entry and 3 circulating lanes.

Standards for deflection of vehicle paths through roundabouts were developed in the
United Kingdom from the safety performance of a large number of roundabouts. At some
of these sites, the size of the central island was reduced to provide greater circulating
pavement widths and thus increase capacity. In doing so, deflection through these
roundabouts was reduced. A study of 23 of 26 such sites showed a 91 percent increase in
casualty accidents after deflection through the roundabouts was reduced. This increase,
which was statistically significant at the 0.001 level, was attributed to higher vehicle speeds
through the roundabouts and supports the design criteria outlined in this guide.

3.7 The Cost of Rolindabouts

The cost of roundabout installation varies a great deal between sites depending on factors
such as the area of pavement construction and other road works, the cost of land acquisition
and the relocation of services. Roundabouts may be either more or less expensive than
traffic signals depending on the particular site. There are many sites where traffic signals
can be installed with little or no change to the existing pavement and curb lines, whereas this
is rarely possible with a roundabout. For those situations traffic signals will generally be
less costly to install than roundabouts. When completing a benefit/cost for a roundabout,
the life cycle costs for the anticipated duration of the improvement should be considered.

Maintenance costs associated with roundabouts will normally include:

0 Pavement Maintenance: The “scrubbing fction” of heavy vehicles turning through

a roundabout makes it necessary to carefully consider the type of surface
treatment required and it may influence the frequency of resurfacing.

0 Curb and gutter and drainage systems,

l Traffic signs,

0 Pavement markings,

l Street lighting,

0 Landscaping

In general there is little difference between the cost of maintaining these items at a
roundabout compared with that at other forms of channelization of equivalent pavement
area. However the additional cost of maintaining and operating traffic signals, which may
be required in conjunction with other forms of channelization, is not required at a

28
 roundabout, except in the unusual situations where “metering” traffic signals are required.

3.8 Environmental Issues

Roundabouts can offer considerable scope for environmental enhancement and are
sometimes favored over other forms of intersection treatment in environmentally sensitive
areas.

The central island can be landscaped and planted provided:

l the treatment does not block any of the sight triangles (refer Section 4)

0 any landscaping will yield to out-of-control vehicles and not be a hazard;

l the treatment does not constitute an unnecessary distraction to drivers.

Planting can be used to discourage pedestrians from crossing at undesirable locations.

Compared to traffic signals, roundabouts may operate with reduced queue lengths and
shorter average delays. This results in:

- less air and noise pollution;

- lower fuel consumption;

- less parking restrictions;

- better access to private driveways.

In addition, the use of a roundabout eliminates potential traffic safety and disruption
problems associated with the malfunction of traffic signals.
2.
Roundabouts can be used on local streets to discourage high traffic speeds and the
intrusion by very large vehicles. Provisions for emergency and service vehicles need to be
considered in the design of these roundabouts.

3.9 Means of Improving the Performance of Roundabouts

3.9.1 Continuous (Slip) Lanes

Where there is a heavy right turn traffic movement, this may be either separated from
the operation of the roundabout by providing a separate right turn slip lane or by providing
an auxiliary lane for this traffic. In the latter case, the right turn entry conditions can be
improved by positioning the splitter island past the entry lanes thereby shielding the right
turn entry movement.

Where a separate right turn slip lane is used, right turning traffic can be excluded from
the capacity and delay calculations. To be fully effective, the layout must ensure that the

29
 circulating traffic and the right turning traffic does not conflict. The exclusion of this right
turning traffic will increase the capacity of the roundabout.

3.9.2 Flaring (tapering) of the Entries

*.
Kimber (1980) has established equations for the performance of roundabouts (in the UK)
with flared entries. The capacity is increased by about 20 percent if an entry is flared from
two lanes to three lanes over a short tapering distance from the entry. However studies in
the UK (Maycock and Hall 1984) showed that as the entry widths are increased, so do the
accident rates between entering and circulating vehicles. The increase in capacity is not
insignificant, but it is substantially less than can be achieved by adding another full length
entry lane. It has been observed that drivers’ perception and behavior at roundabouts is quite
variable and at many locations drivers do not use multi-lane roundabouts to their maximum
effectiveness. In Australia, the introduction of a short flaring on an approach generally does
not realize increased benefits. However if designers wish to investigate the potential for
flared approaches, they should consult the design equations given in Kimber (1980).

3.9.3 Grade Separation

The performance of any intersection can be improved with the grade separation of some
of the conflicting traffic movements. On heavily travelled roads, roundabouts incorporating
grade separated movements can offer significant benefits. It is usual for the through traffic
movements of the more important road to be separated from the roundabout, e.g. as in a
“bridged rotary” type interchange on a freeway. However in special situations, the traffic
movements separated out from the roundabout can also be a right turn movement.

The analysis of a grade separated roundabout is the same as for other forms of
roundabout except that the traffic movements that are removed from the conflict are not
included in the calculation of the performance characteristics. It is obvious that grade
separation substantially increases the capacity and reduces delay and accident potential of a
roundabout.

3.9.4 Entry Metering

Roundabouts will not function efficiently if there are insufficient acceptable gaps in the
circulating traffic stream. If there is one approach with a very heavy through or right turn
traffic movement, which is not interrupted sufficiently by the circulating flow, then this
stream will present few acceptable gaps to drivers at the next entry. The capacity of this
next entry will be very low and the delay to this traffic excessive. Often this situation occurs
only during peak flow periods and at these times the operation can be dramatically improved
by artificially interrupting the high flow approach.

Entry metering is usually done by installing traffic signals to meter the flow as it
approaches the roundabout. This has the effect of bunching the flow of traffic and
introducing more of the longer duration gaps. It is important not to locate the signals too
close to the entry as this may confuse the “right of way” requirements at the entry. While
there has been concern expressed about the confusion of vehicle priority where signalized
pedestrian crossings are located close to roundabouts, Thompson et al (1990) concluded that
“there was no evidence to support the view that drivers who proceed through these traffic

30
signals strongly associate them with the control of entry onto a roundabout”.

Pedestrian operated traffic signals, set to operate on a regular cycle during the peak
periods or activated by a detector placed in the approach suffering excessive delay, have
been successfully used at some sites to meter traffic into a roundabout. Two aspect signals
(red and yellow only) have also been used. Wherever such metering is used, it is important
to provide signing at the signals to advise drivers that the flow is being metered. the “STOP
HERE ON RED SIGNAL” sign is usually also provided.

Where a signalized pedestrian crossing is used for traffic metering, care needs to be taken
with its location to ensure that sufficient space is provided between it and the exit from the
roundabout to avoid traffic queuing back into the circulating roadway.

Metering can be applied to more than one entry at the same roundabout.

31
4 GEOMETRIC DESIGN OF ROUNDABOUTS

4.0 GENERAL

AASHTO guidelines should be followed for turning radii, superelevation, grades, etc.
If they are not followed, justification must be documented and approved at the P.I. submittal.

4.1 GENERAL PRINCIPLES

4.1.1 FLARE DESIGN AT ENTRY

Flare is the widening of the approach road to increase the capacity of the roundabouts.
The approach should never be widened such that there are more approach lanes than
circulating lanes. The length of flare should be between 100 and 300 feet. See Figure 4.1.

4.1.2 ENTRY WIDTH

Entry width can vary depending on the design vehicle and approach roadway width. In
general, the entry width should be between 11 feet and 15 feet per entry lane. The entry
width should be less than or equal to the circulating width.

In the end, the entries should be designed to accommodate the design vehicle while
ensuring adequate deflection.

The approach curve to the roundabout should be the same radius or smaller than the
radius of the curved path that a vehicle would be expected to travel through the roundabout.
It is better to give approaching drivers a clear indication of the severity of the curve they will
have to negotiate, since the speed at which drivers negotiate is dependent on their perception
of the sharpness of the first curve. The entry radii should be designed tangential to the
central island.

FIGURE 4.1 Flare design at entry

32
 The entry radius should be a minimum of 50 feet for single lane roundabouts and 100
feet for multi-lane roundabouts. Small entry radii results in drivers reducing speed to a
degree that drivers consider unreasonable or have difficulty in negotiating, or in drivers
ignoring lane lines and cutting off vehicles in adjacent lanes. Figure 4.2 illustrates the
components of entry design.

Straight tangential
departure is prefe
- Continuation of
Varies, IT-I 5’ curved approach

Splitter island
envelope
12’45

to I’

. .

* 2’ acceptable on small islands

FIGURE 4.2 Typical Roundabout Entrancelfiit Conditions for Urban Areas

4.1.3 CIRCULATING WIDTH

The circulatory width should be constant and should be between 1.0 and 1.2 times the
maximum entry width.

The circulating roadway should generally be circular in plan. Oval shaped roundabouts are
acceptable ( and preferred on roadways
with wide med.&s or with unusual
geometry ) as long as deceptively tight
bends are avoided. See Figure 4.3 for an
example of an oval shaped roundabout.

It is a good design to avoid short

lengths of reverse curve between entry
and exits. It is difficult to achieve this
on three-legged roundabouts or on
roundabouts with skewed entries.
FIGURE 4.3 Oval rourudabout

33
4.1.4 INSCRIBED CIRCLE DIAMETER

The size of the roundabout is a compromise between making it small enough to provide
adequate deflection while making it large enough to provide for the appropriate design
vehicles.

Maryland SHA has determined that the smallest inscribed circle diameter for a single lane
roundabout is loo-feet on a state highway based on a WI%-50 design vehicle. Roundabouts
on smaller subdivision roads may be smaller depending on the maximum design vehicle. In
all cases, the layout should be verified using the appropriate design vehicle template. See
Figure 4.4.

FIGURE 4.4 Turning templates for over-dimensional vehicles

4.1.5 EXITS

The exit from a roundabout should be as easy to negotiate as possible. Whereas entries
are designed to slow vehicles, exiting vehicles should be able to accelerate out of the
circulating roadway. Therefore, the exit radii should generally be greater than entry radii.
Straight paths are preferred, if possible.

4.1.6 SPLI’ITER ISLANDS

Splitter islands should be provided on all roundabouts. They provide shelter for
pedestrians, guide traffic into the roundabout, and deter left turns from dangerous short-cuts
through the roundabout.

34
 On arterial road roundabouts, the splitter islands should be of sufficient size to shelter
a pedestrian and be a reasonable target to be seen by approaching traffic.

A long length of curve on the approach island allows drivers to more easily recognize
the degree of curvature ahead. This is particularly so on divided roads and when approach
speeds are high. However, care should be taken not to provide unnecessarily large radius
curves as this could encourage high speeds through the roundabout.

The entry and exit curves from a roundabout form the spitter island envelope. Pavement
markings and a raised island should be constructed within the spitter island envelope as
shown in Figure 4.2.

In high speed areas the splitter island should be relatively long (200 feet 2) to give early
warning to drivers that they are approaching an intersection and must slow down. Preferably
the splitter island and its approach pavement markings should extend back to a point where
drivers would be expected to normally start to reduce their speed. The lateral restriction and
funnelling provided by the splitter island encourages speed reduction as vehicles approach
the entry point. Curbs should be placed on the right-hand side for at least half the length
of the splitter island to strengthen the funnelling effect. See Figure 4.5.

Typically 65’ to 100’ diameter

FIGURE 4.5 Typical Rural Roundabout Design (with High Speed Approach Roads)

35
4.1.7 DEFLECTION

Adequate deflection of the vehicle entering a roundabout is the most important factor
influencing their safe operation. Roundabouts should be designed so that the speed of all
vehicles is restricted to less than 30 MPH within the roundabout. This is done by adjusting
the geometry of the entry and by ensuring that “through” vehicle paths are significantly
deflected by one or more of the following means:

l The alignment of the entry and the shape, size and position of approach splitter
islands (see Figure 4.6);

l Provision of a suitable size and position of central island;

l Introduction of a staggered or non-parallel alignment between any entrance and exit.

Deflection at Roundabouts with one Circulating Lane

The maximum desired “Design Speed” is obtained if no vehicle path (assumed 7 feet
wide) has a radius greater than 430 feet. This radius of curvature corresponds approximately
to a vehicle speed of 30 MPH assuming a sideways force of 0.2 g. The required deflection
for a si.ngle lane roundabout is shown in Figure 4.7.

Deflection at Roundabouts with two or three Circulating Lanes

For multi-lane roundabouts (two or three circulating lanes), it is generally difficult to

achieve the full deflection recommended above for single lane roundabouts. Where this is
the case, it is acceptable for the deflection to be measured using a vehicle path illustrated in
Figure 4.8. This differs from that used at single lane roundabouts in that the fastest
(maximum radius) vehicle path is assumed to start in the right entry lane, cut across the
circulating lanes and pass no closer than 5 feet to the central island before exiting the
roundabout in the right lane.

36 ’
 Deflection at Roundabouts for various desim speeds

For most state highway applications, design of the entries for 25-30 MPH deflection is
acceptable. However, on minor state roads, county roads, and local roads, the designer may
wish to create a slower entry condition. The following table illustrates the deflection curve
for different entry conditions.

1 I 11
1 De yii 1 Deflec~tiCurve 1

20 180
25 290
30 430

. Table 4.1 Deflection Curve Radii

4.1.8 SIGHT DISTANCE

Several sight distance criteria should be applied to the combination of vertical and
horizontal geometries at roundabouts. Those criteria which greatly influence the safety
performance of a roundabout and also affect the positioning of signs and landscaping etc.,
are illustrated in Figure 4.9.

Criterion 1

The alignment on the approach should be such that the driver has a good view of both
the splitter island, the central island and desirably the circulating carriageway. Adequate
approach stopping sight distance should be provided, to the “Yield” lines and, as an absolute
minimum, to the nose of the splitter island.

37
 a) By Central Island

I) Wlthout ‘blisters’ 2) Wlth ‘blisters’

Low speed (<35 mph) (Preferred treatment)
Urban local streets

(when disturbance of exlstlng

curbs Is to be avolded)

430’ radius

cl By Entry Roadway Alignment

and Posltlon of Splitter Islands

FIGURE 4.6 Alternative methais for providing vehicle defection (not to scale)

38
FIGURE 4.7 Illustration of the defection criteria for a single lane roundabout

FIGURE 4.8 Illustration of the dejlection criteria for a multi-lane roundabout

39
Table 4.1 indicates the required approach sight distances.

Approach Speed Stopping

(MPH) Distance* (ft)
25 98
31 131
37 180
43 230
50 344
56 426
62 525
68 623
75 754

* measured 4.0 ft to zero

TABLE 4.2 Approach Sight Distance (ID)

Criterion 2

A driver, stationary at the “yield” line, should have a clear line of sight to approaching
traffic entering the roundabout from an approach immediately to the left, for at least a
distance representing the travel time equal to the critical acceptance gap. A critical gap
value of 5 s, giving a distance of 225 feet, (based on an entry speed of 30 MPH), would be
typical for arterial road roundabouts operating with low circulating flows. At sites with
higher circulating flows or in local streets, criterion 2 sight distance could be based on a
critical gap of 4 s.

The criterion 2 sight distance should also be checked in respect to vehicles in the
circulating roadway having entered from other approaches. The speed of these vehicles can
be expected to be considerably less than 30 MPH and the corresponding sight distance to
them (e.g. across the central island) should also be based on a critical gap of 4 s to 5 s.
This represents a distance much less than 225 ft because of the low circulating speed of these
vehicles. This is illustrated in Figure 4.9.

Criterion 3

It is also desirable that drivers approaching the roundabout are able to see other entering
vehicles well before they reach the “yield” line. The 125 ft - 225 ft sight triangle shown in
Figure 4.9 allows an approaching driver, slowed to 30 MPH, time to stop and avoid a
vehicle driving through the roundabout at 30 MPH. It is desirable that this sight triangle be
achieved, although in urban areas it may not always be possible. At roundabouts, the speed
of vehicles is more controlled in the circulating roadway than on the approaches and if
Criterion 3 sight distance is available to an approaching driver then any circulating driver
in this zone would also be able to see an approaching vehicle.

40
 Note that within the zones subject to Criteria 2 and 3, it is acceptable to allow momentary
sight line obstructions such as poles, sign posts and narrow tree trunks.

CRITERION 2 (essential) 1
m
1111
Provide adequate sight distance
for drivers to detect occsptablc CRITERION 2
%%%staod % 5s gap and Sight dist-based an 4s to
30 mph for arterialroads 5s gap and max ri ht tun
115’based on 5s gap for circulotlng speed (13 mph
local roads. I II to 20 mph t pical far at&al road8.
4s 7 ap and g nrph to 10 mph
for ocal roads).

CRITERION 3 /
Provide sight triangle ’ CRITERION 1
to allow comfortable Provide CQproach Sit Distance
race nitiin of potentid
conf I?ct. (desirable)

FIGURE 4.9 Sight distance requirements

4.1.9 OTHER VISIBILITY CONSIDERATIONS

At any roundabout, designers must provide the sight distance quantified and described
above. A driver must also be provided with sufficient visibility to readily assess the driving
task. The sight distance required for this is not precisely quantified and only general
guidance can be given.

41
 To enhance the prominence of the roundabout, the curbs on both the splitter island and
central island may be light colored or painted white. To improve driver recognition, the
central island may be mounded and/or reflectorized chevron pavers may be used, provided
the overall height does not obstruct visibility or hide the drivers view of the overall layout.

As with other types of channelization, it is better to position a roundabout in a sag

vertical curve rather than on a crest. Unlike other cross intersections, roundabouts require
all drivers to change their path and speed, thus it would be important to avoid locating
roundabouts just over a crest where the layout is obscured from the view of approaching
vehicles.

At grade separated roundabouts, particularly where there may be a structure in the central
island or a bridge railing which might obstruct a drivers’ visibility, care must be taken to
ensure that the sight distance requirements are met. Any guard fences used to protect piers
and structures may also interfere with visibility.

Where there is a light rail crossing incorporated in to the roundabout, care needs to be
taken to ensure that the negotiation speeds are slower and that drivers are aware of the
presence and the location of the rail tracks.

Light rail can be successfully incorporated into a roundabout. As the tracks will pass
through the central island, eliminating part of it, care needs to be taken to ensure that
residual central island remains large enough to be recognized.

4.1.10 SUPERELEVATION AND DRAINAGE

Normal curve superelevation through the roundabout is generally not necessary as speeds
are constrained and drivers tolerate higher values of the sideways force and utilize higher
values of the coefficient of sideways friction when travelling through an intersection.

Above all, it is important that the layout of the roundabout be clearly visible to
approaching drivers and this is best achieved by sloping the crossfall away from the central
island. This generally means accepting negative superelevation for left turning and through
vehicles in the circulating carriageway, but avoids depressing the central island thereby
reducing its visibility to approaching traffic.

As a general design practice, a minimum pavement crossfall of 0.025 to 0.03 ft/ft should
be adopted for the circulating carriageway. A crossfall as low as 0.02 ft/ft has been found
adequate to allow for pavement drainage and would also provide additional driver comfort.

Designing superelevation to slope away from the central island often simplifies the
detailed design of pavement levels and avoids inlets around the central island.

42
 Exceptions to this approach include:

0 Roundabouts on a generally sloping topography, in which case the crossfall

should approximately match the slope across the whole of the roundabout. At
these roundabouts the crossfall may vary around the circulating carriageway but
it should stay within the range of 20.04 ft/ft. Locating a roundabout on grades
greater than 3 to 4 percent should be avoided. Where the general slope of the
land is greater than 0.04 ft/ft, it will be necessary to “bench” the area for a
roundabout, modifying the grade to not exceed a maximum negative crossfall of
0.03 ft/ft.

l Large roundabouts where vehicles will travel on the circulating roadway for some
distance. In such cases, a crown following the center line of the circulating
roadway may be satisfactory or it may be positively super-elevated by sloping the
roadway toward the central island. This improves driver comfort but tends to
increase vehicle speed within the roundabout and reduces the visibility of the
circulating roadway and the central island.

4.1.11 WIDE MEDIANS AND STREETS OF UNEQUAL WIDTH

Particular problems in roundabout design occur at locations where one intersecting street
is considerably wider
than the other and/or
where a wide median
exists. This situation
can occur with local,
collector or arterial
streets or, as is often
the case, where the
intersecting streets are
not of the same
functional
classification. very
often a roundabout will
not be the appropriate
type of treatment.
However, where the FIGURE 4.10 Roundabout on a road with a very wide median
volume of traffic on the
narrower street is greater or equal to that on the wider street and if there are heavy left turn
flows, a roundabout could be suitable.

Where a roundabout is proposed, special care should be taken to ensure that the design
is in accordance with the guidelines listed in Sections 4.2.1 to 4.2.8. In particular,
sufficient deflection for through traffic entering the roundabout is most important.
Generally, a low cost solution which does not require improvements encroaching onto
existing medians will not be possible. Figure 4.10 is an example of a roundabout designed
to adequate standards for an in undivided road crossing a divided road with a wide median.

43
 In these situations the central island is not circular and will involve different circulating
speeds for different sections of the circulating roadway. Left turning drivers entering from
the narrow road in Figure 4.10 will find that the radius of their turning path decreases and
becomes more difficult. This may create a higher accident risk.

4.1.12 WIDE UNDIVIDED STREETS AND “T” INTERSECTIONS

Where a roundabout is to be constructed at an existing “T” junction, it is generally

necessary to build out the curbline to provide deflection of the traffic movement across the
top of the “T” opposite the terminating road. This practice has also been adopted at certain
cross road intersections where one cross street is wider than the other and/or where there
is space for more than one lane of traffic on a particular approach.

Where curblines are to be built out on approaches to roundabouts, special care should
be taken to ensure that adequate delineation is provided, particularly in instances where
there are no parked vehicles on the approach.

4.2 LOCAL STREET ROUNDABOUTS

The major differences in the geometric treatment of local street roundabouts, compared
with roundabouts on state routes, arise from the differing design aims, the generally
narrower street widths, the lower traffic speeds applicable, and the smaller class of vehicle
using the facility. Typical vehicles to be designed for include cars and occasional single unit
trucks.

Roundabouts are usually installed in the local street system as a traffic management
device to improve safety and amenity by controlling vehicle speed and, due to their
restrictive geometry, to create a deterrent for large vehicles and high traffic flows. The
traffic demand in local streets is usually low and capacity and delay calculations are not
required. In most if not all cases, only one entry lane (on each approach) and one
circulating lane is provided.

The geometric design principles for local street roundabouts differ slightly from those
used for rural and arterial road roundabouts. The control of vehicle speed on entry and
through the roundabout remains an important objective as with roundabouts on arterial roads.
However, in the layout of local street roundabouts, the pavement space provided for vehicle
maneuvers usually comprises areas normally required for cars and light commercial vehicles
and specially paved encroachment areas (which may be slightly raised), to cater for the few
larger vehicles which may need to use the site. These larger vehicles may be required to
drive over traversable islands or rumble strip treatments in the splitter islands to negotiate
the roundabout.

44
 5.0 LANDSCAPING

5.1 GENERAL

A roundabout creates new design opportunities within an intersection:

l Roadside Planting.
0 Planting in the central island.
l Creative pavements in truck aprons, splitter island, circling lanes and pedestrian
crosswalks.

However certain goals should be kept in mind as the landscaping design of a roundabout
is developed.

0 Improve the aesthetics of the area.

0 Avoid introducing hazards to the intersection.
l Avoid obscuring the form of the roundabout layout to the driver.
I 0 Maintain required stopping and turning sight distances (see Figure 4.9).
Clearly indicate to the driver that they cannot pass straight through the
intersection.
0 Minimize oncoming headlight glare.
0 Discourage pedestrian traffic through the central island.

5.2 ROADSIDE PLANTING

Plant material can be added to the approach roadways to help in the ftmnelling effect.
The funnelling effect is discussed in Section 4.1.6 on arterial road roundabouts. The lateral
restriction and funnelling provided by the splitter island encourages speed reduction. Plant
material on the right an left side of the approaches reinforces this effect. This illustrated in
Figure 5.1.

45
 L
FIGURE 5.1 Ekampte of the use of lana!scaping to reinforce the funnetting efect at the entrance .
to roundabouts

46
5.2 LANDSCAPE DESIGN FEATURES

As previously mentioned, the introduction of the roundabout to the state highway system
creates new opportunities for landscaping.

In addition to the central island planting, plant material can be added to the approach
roadways to help in the funnelling effect. The funnelling effect is discussed in Section 4.1.6
on arterial road roundabouts. The lateral restriction and furmelling provided by the splitter
island encourages speed reduction. Plant material on the right and left side of the approaches
reinforces this effect. This is illustrated in Figure 5.1.

Where truck aprons are used (As shown in section 4.1.4) and in the splitter islands,
creative pavement colors, textures and markings can be used to blend the roundabout in with
the existing surrounding area. This is illustrated in Figures 5.2 and 5.3.

Finally, when planting in an old road bed, as for the central island, it is necessary to
excavate the old road bed out (to a depth two feet below the original road surface) and
backfill with approved topsoil to provide for adequate growing conditions.

STANDARD TYPE ‘C’ COMBINATION CURB AND GUTTER

OPTIONAL STANDARD TYPE ‘A’ CONCRETE CURB

8’ JOINTED REINFORCED CONCRETE PAVEMENT, MIX NO. 6 MODIFIED

(HIGH EARLY STRENGTH) WITH OPTIONAL STAMPED CONCRETE FINISH
7’ GRADED AGGREGATE BASE

FIGURE 5.2 Typical Section of the Truck Apron

47
 /-LANDSCAPED CENTRAL 1
ISLAND
/

CONCRETE TRUCK
APRON

STANDARD T
COMBINATION
AND GUTTER
ATTERNED

FIGURE 5.3 Ptan of Central Island

48
6.0 SIGNING AND PAVEMENT MARKING

6.1 SIGNING

The general concept for--Roundabout signing is similar to signing any other new
geometric feature along a highway. Proper advance warning, directional guidance and
regulatory control are required to avoid driver expectancy related problems. The guidelines
set forth in the Manual on Uniform Traffic Control Devices for Streets and Highways, the
State of Maryland - Standard Highway Signs booklet, and the.various memorandums issued
by the Maryland Department of Transportation - State Highway Administration govern the
design and placement of signs along all roads in the state. Standards developed by the
individual counties that suppiement the state guidelines may also apply to the design and
placement of signs along roads within those jurisdictions.

Roundabout signing varies based upon the type of roadways intersecting at the
roundabout. The division is made between Highways and Local roads, as explained below.

l HIGHWAY - State highways and County collector roads

l LOCAL ROAD - Other County roads, commercial and residential

6.11 HIGHWAY

Approaches:

1. Junction Assemblies should .be used.

3-. “ROUNDABOUT AHEAD” Warning signs with “YIELD AHEAD” plates should
be used.

3. Destination Guide signs should be used. For higher speed multi-lane approaches
(> 45 MPH) Diagrammatic Guide signs should be considered.

4. “YIELD AHEAD” (W3-2A) signs in combination with Advisory Speed (W13-1)

plates should be used.

5. Where possible, the designer should attempt to reuse any appropriate existing
signs.

6. Other guide signs, such as Advance Route Marker Turn Assemblies may be used
as described in the MUTCD and SHA sign guidelines.

Intersections:
-.
1. “YIELD” (Rl-2) signs in combination with “TO TRAFFIC ON LEFT” (RX-X)
plates should be used.

3d. “ONE WAY” (R6-1R) signs in combination with obstruction markers (WlS-2)
should be used.
,
49
 3. Exit Guide signs should be used.

The locations of these signs. are shown in Figure 6.1.

In high speed, rural areas extraordinary warning signs may be used. These signs are to be used
only if approved by the Director, Office of Traffic & Safety. In cases where high speeds are
expected, and the normal signage and geometric features are expected to have less than desired
effect on vehicle speeds, the following measures may be considered:.

0 Addition of Hazard Identification Beacons to approach signing (Diagrammatic,

Roundabout Ahead).
l Speed Display Signs (actuated by speed detectors).

6.12 LOCAL ROADS

Each approach should include:

1. “ROUNDABOUT AHEAD” Warning sign with “YIELD AHEAD” plate.

Each approach may include:

2. Destination Guide sign. .

The intersection should include:

1. “YIELD” (Rl-2) sign in combination with “TO TRAFFIC ON LEFT” (RX-X)

plate.

The intersection may include:

3-. “ONE WAY” (R6-1R) sign in combination with obstruction marker (W15-2).

3. Exit Guide sign with “DO NOT ENTER” (R51) sign mounted on back.

The locations of these signs are shown in Figure 6.2.

Alternate location of
Exit Guide sign

A Diagramatic Guide sign may be

*substituted for the Destination
Guide sign shown above.

An Advance Route Marker

Turn assembly may be
considered as an alternate.
or supplemental sign.
I

FIGURE 6.1 Typical signins for a state route rounduhout

51
 l OPTIONAL

FIGURE 6.2 Typical signing fbr u locul roud roundabout

.
52
6.2 PAVEMENT MARKINGS

Pavement markings for a roundabout intersection consist of the yield line, hatch markings
in the splitter island envelope, raised retlective pavement markings and typical edge and lane
line striping as shown in Figure 6.3.

The pavement markings at the entrance to the roundabout consist of an 8” to 16” wide
stripe with 3-foot segments and 3-foot gaps. A yield line placed 8’ to 10’ in advance of the
entrance to the roundabout is shown in Figure 6.3. There shall be no painted lines across
the exits from roundabouts.

While there is not conclusive evidence as to whether there should be lane lines
delineating the circulating lanes within the roundabout, it is felt that such pavement markings
may confuse rather than help drivers in the performance of their task of negotiating the
roundabout. With multi-lane roundabouts, pavement markings may also mislead drivers into
thinking that vehicles must exit exclusively from the outer lane of the roundabout. The use
(or lack of use) of pavement markings will be made on a case by case basis.

The pavement marking may be emphasized by the placement of raised reflective

pavement markings as shown in Figure 6.3. Thermoplastic markings may be used to
increase the pavement marking visibility.

Rumble strips may be utilized to reduce approach speeds and to call the drivers’ attention
to warning or destination guide signs. They are especially useful on high speed rural
approaches. An example of an application of rumble strips is shown in Figure 6.1.

The rumble strips consist of 10 foot crosswise stripes of 4 inch permanent preformed
pavement marking tape. Double applications of the 4 inch permanent preformed pavement
marking tape are typically required to produce the desired effect (i.e. one layer of the tape
is applied to the roadway and the second layer is applied directly on top of the first layer).
The spacing of the strips varies directly with the approach speed.
 a”-16” BROKEN WHITE

8” SOLID WHITE

(MATCH EXISTING)
MATCH EXISTING WIDTH

FIGURE 6.3 Typical pavement markings for roundabouts

.
54
7.0 LIGHTING

The satisfactory operation of a roundabout relies heavily on the ability of drivers to enter
into, and separate safely and efficiently from a circulating traffic stream. To do this, it is
important that the driver must perceive the general layout of the intersection in sufficient
time.

It is not possible to provide detailed recommendations on lighting layout for roundabouts

because of the great variety of possible geometric layouts. It is important, however, to
recognize certain desirable features:

A. Lights should be located so that they provide good illumination on the approach nose
of splitter islands, the conflict area where traffic enters the circulating stream, and
at places where traffic streams separate at points of exit.

B. Particular attention should be given to the lighting of the pedestrian crossing areas
if applicable.

C. Lighting poles should not be placed within splitter islands, on the central island
directly opposite an entry roadway, or on the right-hand perimeter immediately
downstream of an entry point.

55
8.0 PEDESTRIAN AND BICYCLE CONSIDERATIONS
8.1 PEDESTRIANS

In the planning and design of roundabouts, special thought should be given to the
movement of pedestrians. In general, research indicates fewer pedestrian accidents at
roundabout intersections when compared to signalized and unsignalized intersections. This
is so for two reasons. First, the speed of all vehicles is slower at a roundabout intersection.
Second, pedestrians use the splitter island as a refuge area. In so doing, the pedestrian only
crosses one stream of traffic at a time. The normal pedestrian crossing location should be
at the yield line as shown in Figure 8.1 (a).

Pedestrian crossing lines should not be painted on the entrances and exits of roundabouts
as they may give pedestrians a false sense of security. Pedestrians should be encouraged to
identify and accept gaps in traffic and to cross when it is safe to do so.

The normal placement of pedestrian crossings at roundabouts should be 20-25 feet from
the yield line. Crosswalk striping should not be used because the driver may confuse the
crosswalk limit lines with the yield lines. The pedestrian crossing could be reinforced with
handicapped ramps and/or colored and patterned concrete.

Consideration should be given to providing priority crossings for pedestrians where

pedestrian volumes are very high, where there is a high proportion of young, elderly or
infirm citizens wanting to cross the road, or where pedestrians are experiencing particular
difficulty in crossing and are being delayed excessively. It is desirable that these crossings
be placed at least 75 feet downstream of the exit from the roundabout (and possibly be
augmented by pedestrian signals). This will reduce the probability that vehicles delayed at
the pedestrian crossing will queue back into the roundabout and gridlock the whole
intersection. The locations of these pedestrian crossings are shown in Figure 8.1 (b).

8.2 CYCLISTS i.

In most circumstances, roundabouts provide satisfactorily for cyclists, although it has

been found that multi-lane roundabouts are more stressful to cyclists than single roundabouts
owing to the greater chance of conflicts between vehicles and cyclists. It has been found that
generally, cyclists use roundabouts in a similar way to motor vehicles. Special provisions
for cyclists are not normally required.

To provide a satisfactory level of safety for cyclists at roundabouts, particular attention

will need to be taken in the layout design to:
0 ensure that adequate deflection and speed control is achieved on entry and through
the roundabout.

l avoid larger than necessary roundabout (inscribed) diameter, thus reducing travel
speed through the roundabout.

l avoid excessive entry widths and alignments which can also increase vehicle entry
speeds.

56
 l ensure that sight lines are not obstructed by landscaping, traffic signs or poles which
may even momentarily obscure a cyclist.

l provide adequate lighting.

Normally, special bicycle lanes will not be required as the cyclist would be able to proceed
through the roundabout in the travel lane. If high volumes of bicycle traffic exist, a special
bicycle/pedestrian facility could be constructed as shown in Figure 8.2.

i.

57
PAVED SIDEWALK
\

PAVED SIDEWALK -/
Ill lb-dr

FIGURE 8.1 Examples of Pedestrian Crossings

58
FIGURE 8.2 Example of a special bicycle facility

59
9.0 WORK ZONE TRAFFIC CONTROL

During the construction of a roundabout it is essential that the intended travel path be
clearly identified. This may be accomplished through pavement markings, signing,
delineation, and guidance from police and/or construction personnel depending on the size
and complexity of the roundabout. Care should be taken to minimize the channelizing
devices so that the motorist has a clear indication of the required travel path. Each
installation should be evaluated separately as a definitive guideline for the installation of
roundabouts is beyond the scope of this policy.

9.1 PAVEMENT MARKINGS

The pavement markings during construction should be the same layout and dimension
as those used for the final installation. Because of the confusion of a work area and the
change in traffic patterns, additional pavement markings may be used to clearly show the
intended direction of travel. In some cases when pavement markings cannot be placed,
channelizing devices should be used to establish the travel path.

Temporary raised pavement markers (TRPM’s) should be used to supplement pavement

markings. Spacing and positioning of the TRPM’s should be such as to clearly delineate the
intended travel path but not be in such quantities as to detract from the pavement markings.
(See Figure 9.1)

9.2 SIGNING

The signing during construction shall consist of all necessary signing for the efficient
movement of traffic through the work area as described in Figure 9.2, pre-construction
signing advising motorists of the planned construction, i.e. “CONSTRUCTION OF A
ROUNDABOUT TO BEGIN.. . “, and any regulatory and warning signs necessary for the
movement of traffic outside of the immediate work area.

9.3 LIGHTING

Permanent lighting, as described in section 7.0, should be used to light the work area.
If lighting will not be used, delineation, as described in section 9.1, should be used.

9.4 CONSTRUCTION STAGING

As is the case with any construction job, before any work can begin, all traffic control
devices should be installed as indicated in the traffic control plan or recommended typical.
This signing shall remain in place as long as it applies and then removed when the message
no longer applies to the condition.

60
 RPM SPACING SHOULD GIVE
A CLEAR INDICATION OF
THE TRAVEL PATH WITHOUT
DETRACTING FROM THE
PAVEMENT MARKINGS.
SPACING SHOULD BE ON THE
4VERAGE IO’, WITH A MIN-
MUM SPACING OF 5’ FOR
SHARPER CURVES AND A
MAXIMUM OF 20’.

FIGURE 9. I Roundabout workarea pavement markings

61
 WORK ZONE TRAFFIC COMROL lYPICAL
IMPORTANT- THIS DRAWING SHALL BE USED IN
COM8INATION WITH THE GENERAL NOTES.

KEY:
I I CHANNEUZlNG OEWES
I
& ---SIX
-FAac
SUPPORT
of SIGN

Q MREC710IJ OF TRAFFIC

&iJ WORK SITE

FUG (OpnoNk)

SPEmm aEGof?Y
tax IrEus
I Maryland Department of Transportation
ApPRova) STATE HIGHWAYADMlNlSTF?ATlON
m.ctamuff1wn:m,wm
M0u.m ROUNDABOUT FLAGGING OPERATION
1 GREATER THAN 40 MPH/OVER 24 HRS.

4 STANDAJ3D NO.
FIGURE 9.2 Work Zone Trafic Control

62
 Prior to the work which would change the traffic patterns to that of a roundabout, certain
peripheral items may be completed. This would include permanent signing (covered),
lighting, and some pavement markings. These items, if installed prior to the construction
of the central island and splitter islands would expedite the opening of the roundabout and
provide additional safety (lighting) during construction.

When work has commenced on the installation of the roundabout, it is desirable that it
be completed as soon as possible to minimize the time drivers are faced with an unfinished
layout or where the traffic priority may not be obvious. If possible, all work, including the
installation of splitter islands and linemarking should be done in one day.

If it is necessary to leave a roundabout in an uncompleted state overnight, the splitter

’ islands should be constructed before the central island. Any portion of the roundabout that
is not completed should be outlined with pavement markings and delineation in such a way
as to clearly outline the intended travel path. (see Figure 9.3).

In general, the order of construction should be as follows:

1. Install and cover proposed signing.

2. Construct outside widening if applicable.
3. Reconstruct approaches if applicable.
4. Construct Splitter Islands and delineate central island. At this point the signs
should be uncovered and the intersection should operate as a roundabout.
5. Finish construction of the central island.

9.5 Public Education

It is important to educate the public anytime there is a change in traffic patterns. It is

especially important. for a roundabout because a roundabout will be new to most motorists.
The following are some suggestions to help alleviate initial driver confusion.

1. Public meetings prior to construction.

2. News releases/handouts detailing what the motorist can expect before, during, and
after construction.
3. Variable Message signs during construction.
4. Travellers Advisory Radio immediately prior to and during construction to
disseminate information on “How to drive” etc.

63
 v\
00
00
0 0
000

(IONE-
R6-I R

R4-7R

I-.

FIGURE 9.3 Roundabout workarea delineation

APPENDIX
BENEFIT/COST ANALYSIS

1. Accident data is collected for a ;hree year period. This allows a large amount of data to be
collected and insures an high probability that collision types related to the design and operational
features of the location are represented proportionately to their occurrence.

2. Traffic volume figures are developed for each year studied for the purpose of computing accident
rates by type of collision. Based upon the three year average accident rate by each collision type
and using linear regression to project the traffic volume to the improvement completion year we
can estimate what the projected accident frequency by collision type will be for the improvement
completion year.

3. The proposed improvements are based upon the accident experience that has manifested itself
during the three year study period. After an improvement is selected, the effectiveness of the
improvement is measured by estimating the extent it will reduce a specific type of accident.
Since roundabouts are a relatively new traffic control device, the accident reduction factors are
assumed and will be monitored upon completion of Before/After studies when a statistically
significant number of roundabouts have been installed. Primarily roundabouts have been found
to reduce angle and left turn accidents. The accident reduction factor for each type of accident
is as follows:

Angle : 0.5
Left Turn : 1.0

4. The effectiveness quotients are summed to determine the total number of accidents reduced. The
composite figure is then multiplied by the average accident cost for each collision. This figure
is what we identify as the First Year’s Benefit (FYB). The first year’s benefit derived from the
improvement ‘cannot in itself be used to evaluate the merits of the improvement. This cost does
not reflect the increases in the benefit brought about by an increase in the accidents reduced and
interest saved over the service life of the improvement.

5. The cost of the improvement should reflect all those costs that will have to be paid out. This
includes interest on money spent derived by the use of a Capital Recovery Factor (CRF) based
upon the current interest rate and the service life of the project, and the maintenance costs paid
in addition to normal maintenance. The capital recovery factors to be used for this analysis are
as follows:

Proiect Life CRF

10 0.1490
15 0.1168
20 0.1019
25 0.0937

I
 In addition, those costs returned to the motorist should be considered. For instance, if the
equipment at a particular site was removed and replaced with new, the old equipment’s salvage
value should be considered and reduced from the initial cost of the improvement. Therefore,
the initial cost of the improvement multiplied by the capital recovery factor plus the maintenance
cost minus the salvage value multiplied by the sinking fund factor gives us an Equivalent
Uniform Annual Cost (EUAC).

EUAC = initial cost x CRF + maintenance cost - (salvage value x SFF)

6. The first year’s benefit, as stated, is an inadequate measure of benefit. This recognizes only the
benefit expected to be derived during the first year. In reality we would expect an annual
benefit derived by the cumulative total of accidents being reduced during the entire service life
of the improvement. This also takes into account those accidents reduced that would have
occurred as a result of increased traffic volumes. The benefit is actually a measure of cost that
would have been paid out by the motorist in accident occurrence, but has been foregone by the
improvement. This same cost, if it has been paid, would also collect interest. Since the cost
has been eliminated, the interest paid on this cost is also eliminated and therefore able to be
invested productively elsewhere. Thus taking into consideration the First Year’s Benefit, the
interest saved over the entire service life of the improvement, the Capital Recovery Factor and
normal growth in traffic volumes, we calculate the Equivalent Uniform Annual Benefit (EUAB).

The EUAB is calculated as follows:

Where: EUAB = Equivalent Uniform Annual Benefit

CRF = Capital Recovery Factor
FYB = First Years Benefit
i = Average Annual Interest Rate
j = Annual Growth ’
N = Service Life

7. The Benefit/Cost ratio is a rather simple concept. The EUAB is divided by the EUAC, the
resulting number indicates how much return (benefit) can be expected for each dollar invested
(cost). If the number exceeds one, the project can be considered cost effective. On a very
simplistic basis, the higher the B/C ratio, the better the investment. A sample spread sheet is
attached for the users benefit.

II
LE
.:.: :::,, ::>::j::::I:
,.:.:.:.:.:.:., ,.,. .,
‘:i:i:i’:‘:i:i:i::,i;~: PROJECTED ACCIDENT AVERAGE FORECASTED FIRST
MANNER OF C()LLlSl()N ACCIDENT REDUCTION ACCIDENT REDUCTION YEARS
EXPERIENCE FACTOR COST BENEFIT

2.3 0.5 10200 1.2 11730

11.6 1 10700 11.6 124120

TOTAL 12.8 135850

SERVICE CRF INITIAL EUAC
I
LIFE COST
5
JNDABOUT IO 0.149 300000 44700

*- (ANN % INC IN TRAFFIC) 0.03

J-
i= (INTEREST RATE) 0.08 EXP= 8.15
N= (SERVICE LIFE) 10

:ORECASTED REDUCTION 12.8 EUAC 44700

4VERAGE ACCIDENT COST 10653.13 EUAB 152749

:IRST YEARS BENEFIT 135850

BENEFIT COST 3.42

III
’ ‘SST-ROSCHE ENGINEERS, INC.
SCOTT ADAM ROAD, SUITE 103
~ICOCKEYSVILLE, MD 2 I 030 MARYLAND STATE HIGHWAY ADMINISTRATION
‘EL: 41 O-683-1 683 SAFETY IMPROVE MENT WORKSHEET
MD 1701 MD 174 A UNE ARUNDEL COUNTY

:.:::::j :,.,., ::::i’i;:,:,:,. . ...\.:.:.: ,.:: .;:.‘::..:..:,:.:‘:‘(~:‘::‘:::::,::,:(’~

MANNER OF 1990 1991 1992 ,. .\..(. . ...\. :* .(.( ,. . . . . . . .::?y:.:,
~....,..((.> .\.(.(.(\l,.,
‘(‘(0’.
,.,.,.,.:~
.il ,.:. . ..I...., ,A),., : ::
.,. . ..v ,,. .,.,. :, ,+:,,.:,. . . . . . . . . :,:,::,
,/...~.3.00. . . . . .,.,.,.,.,\. .., ,.(, :,:,,,:,:,:::: :,:,:, :,:,>::,;>:
..:.::::.:
::>:)... . . . . . ...? . . ..A.
,.... .i.

COLLISION ACC # RATE ACC # RATE ACC # RATE RATE PROJ.

ANGLE 3 0.22 2 0.14 1 0.07 0.14 2.3

REAR END

FIXED OBJECT

OPP. DIRECTION

SIDESWIPE

LEFT TURN 13 0.97 8 0.55 9 0.62 0.71 11.6

PEDESTRIAN

-‘ARKED VEHICLE

TVI’H COLLISION

NIGHTTIME ACC.

ADT 36700 ADT 40000 ADT 39600

AVM 13395%x AVM 1460000C AVM 144936OC
:::. A...,..:...:,
‘,‘,‘,““‘.
‘V.‘.‘. .,.,.(.
.,.,.,.,
.:.:.:.:‘:‘,.i:‘:‘:‘:.:.:...:.:.:..:::~::~;:::,~:::
:,:,:,::
:.:,:,:.
::::, . . (((.(.‘,‘.‘..
.,,.,.,~.‘..:.:.:L
“‘,‘,‘.’ ._...
:.,.:,:.,,:.:.:.:.:.~....
....,,,.,:
.~,~,.,.,.,.......,:,:,:,
..........._......
....:.:.:.:;,:,:::::
:,,,:, :.:.:.:.:.;
c~,:,:
,(,(.(.:.:
:,:,:,:,:,.,:,
” ”::,
“.‘~:‘~‘+:.:e
‘:““‘~:‘.- ‘,‘,‘,‘,:.:.::.:.:.:i:::::::.:.:.:.:
:”:=+t:
‘.‘.‘. ,,,.,:
._._.,.,...,.... ,.,.,,.,...._(,(.(,(.(,.,
~,.,.,.,.,..~...,................::
,:,:,:,:,lt::‘:‘:’
.,.,.,.,...,......
(,.,.,,,_ ::::::
::,:,::,:,
‘:Bsj:!:::I::::::::::~:~:~,~.~,~:~:::~:~:~~:~:~:~~:::::::::::::::::
,,./.:.:,~
,:,.,.,:,
““‘.‘.‘O...,...;,:,:,:,:.:..
::::~~:~:::.:.::::::.:,:;:.:.:.:.:.:,:,,~::::
::~:,:.:t.:::::.:.~~~ .:.:.:,:.:
::::::::;~::~:~~:~:::~,~~~i:18
........... .......~.../C..
“.“:.‘.:.:.:~:f:::::.:.:.:~:‘:::.:::
:?::I:::.::::.: :.:.:.:.:,
:,:,:Mjij,;,~.:,:,::~.::
:...:
.:.i::.,..
,. :.:..:.:+:p:::
::::::::,:::,>:.:
.A....
...\. :(,:,~,..:.
~:.:.:.:.,.,,,,(,(,,,)
,,,,,:(,)~.,:.
,.,..
,.:.:.:,:.:.:.:,;:;:;
.,.,.,)).,.....,.,...,.,.,.,.,.,~,.,.,.,.,
,:.,.,.,.,
., : FORECAST 1994 1635923:
DABOUT
REFERENCES

1. AKCELIK, R (1991). “Implementing roundabout and other unsignalized intersection

analysis methods in SIDRA ” Australian Road Research Board Working Document
WDTE91/002.

2. AKCELIK, R AND TROUTBECK, R.J. (1991). “Implementation of the Australian

Roundabout Analysis Method in SIDRA. In Highway Capacity and Level of Service, (Ed U.
Brannolte) Proceedings of the International Symposium on Highway Capacity. ” Karkxuhe
Balkena Rotterdam July, 1991.

3. AUSTROADS (1993)a. “Trafic Engineering Practice, Part 6 - Roundabouts, Sydney. ”

4. AVENT, A.D AND TAYLOR, R.A. (1979). “Roundabouts - Aspects of their design and
operations. ” Institution of Engineers Australia (Queensland Division. Technical Papers).
Vol. 20(7), July, 1979.

5. CATCHPOLE, E.A. AND PLANK, A. W. (1986). “The capacity of a Priority

Intersection. ” Transportation Research Board. Vol. 20B(6), pp 441-456.

6. CENTRE FOR RESEARCH AND CONTRACT STANDARDIZATION IN CIVIL AND

TRAFFIC ENGINEERING - THE NETHERLANDS. - ’ Sign up for the Bike-Design
Manual for a cycle-friendly infrastructure” - August, 1993.

7. Department of Transport, Roads and Local Transport Directorate, Departmental Standard

TD 16/84 - “The Geometric Design of Roundabouts,” August 1984. Published by The
Department of Transport.

8. Department of Transport, Roads and Local Transport Directorate, Department Standard TA

42/84- “The Geometric Design of Roundabouts,” August 1984. Published by The
Department of Transport.

9. GEORGE, A.L. (1982). ‘Estimation of Geometric Delay at a Roundabout. ’ Unpublished

report to the Country Roads Board (now Roads Corporation), Victoria.

10. HORMAN, C.H. AND TURNBULL, H.H. (1974). “Design andAnalysis ofRoundabouts. ’
Proceedings 7th Australian Road Research Board Conference, Vol 7(4), pp. 58 - 82.

11. KIMBER, R.M. (1980). “The Trajjk Capacity of Roundabouts. ’ Transport and Road
Research Laboratory, Laboratory Report LR 942.

12. MAYCOCK, G. AND HALL, R.D. (1984). “Accidents at 4-arm Roundabouts. ” Transport
and Road Research Laboratory, TRRL Laboratory Report 1120.
REFERENCES (Continued)

13. MIDDLETON, G. (1990). “Geometric Delays at Roundabouts. ” Roundabouts and

Unsignalized Intersection Workshop, Queensland University of Technology, November,
1990. (Ed: R.J. Troutbeck).

14. NAASRA (1986)a. “Rourulabouts, a Design Guide. ”

15. PLANK, A.W. AND CATCHPOLE (1984). “A general capacity formula for an
Uncontrolled Intersection. ” Traffic Engineering and Control. Vol. 25(6), pp 327-329.

16. TROUTBECK, R.J. (1984). “Does gap acceptance theory adequately predict the capacity
of a roundabout?” Proceedings 12th Australian Road Research Board Conference, Vol.
12(4), pp 62 - 75.

17. TROUTBECK, R.J. (1986). “Average delay at an unsignalized intersection with two major
streams, each having a dichotomised headway distribution. ” Transportation Science, Vol.
20(4), pp 272-286.

18. TROUTBECK, R.J. (1989) “Evaluating the Performance of a Roundabout. ” Australian

Road Research Board Special Report 45.

19. TROUTBECK, R. J . (1990) “Traffic Interaction at Roundabouts. ” Proceedings 15th

Australian Road Research Board Conference, Vol. 15(5), pp 17-42.