Dan Sullivan 2015 Session 6
Dan Sullivan 2015 Session 6
Dan Sullivan 2015 Session 6
DR DAN SULLIVAN
Managing Director, Solutions in Transport
dan.sullivan@solutionsintransport.com.au
DR OWEN ARNDT
Director Road Design, Transport & Main Roads
owen.k.arndt@tmr.qld.gov.au
DAVID GOUGH
Principal Engineer, Transport & Main Roads
david.p.gough@tmr.qld.gov.au
1. Introduction
Unsignalised intersections remain the most common form of intersection on the road network and the
preferred layout and control is based on a range of factors. Austroads (2013) provides guidance for the
selection of the preferred intersection turn treatment from the major road using warrant graphs. Further
details on the development of these warrants graphs are explained in Arndt & Troutbeck (2006). The
preferred treatment is determined through economic analysis comparing the difference in crash costs
between two turn treatments with the relative difference in construction costs.
The Austroads (2013) warrants most appropriately apply to 2-lane, 2-way (2L2W) roads for greenfield
intersections on new roads and assist in the selection of the major road turn treatments types for;
right turn treatment - basic, short channelised or full channelised (refer Figure 1), and
left turn treatment - basic, short auxiliary or full auxiliary / channelised (refer Figure 1).
These warrants have been widely used for a number of years. Inevitably queries regarding their
application in a variety of situations not explicitly considered in Austroads (2013) have been raised.
Based on these queries, TMR has established a need to provide additional warrants / guidance for;
A treatment for very low traffic volumes (i.e. a simple intersection treatment below BAR / BAL).
Application to multi-lane roads: 4-lane, 2-way and 6-lane, 2-way roads.
Intersections subject to a limited period (e.g. 6 months to 1 year) of heavier traffic due to
construction or mining activities.
This paper summarises the findings of Sullivan & Arndt (2014) which supported the recommendation of
an expanded set of turn warrants for the selection of the preferred turn treatments from the major road at
unsignalised intersections. The methodology and results presented herein are based on the same
methodology in Arndt and Troutbeck (2006). New warrants and their application would then readily
supplement the existing warrants and provide consistency in outcomes.
Note:
Short versions of the AUL, CHR and CHL turn treatments, AUL(s), CHR(s) and CHL(s) respectively can also be
used
2. Existing warrants
The current Austroads (2013) turn treatments warrants for unsignalised intersections were developed by
Arndt & Troutbeck (2006). The methodology applied evaluated the crash rates at unsignalised
intersections against a range of parameters and were developed using a methodology based on crash rates
to determine the appropriate turn treatment warrants.
The following summarises the methodology contained in Arndt and Troutbeck (2006) and describes the
process used to develop the warrants.
BCR
ARM C A TDL
CTT
(1)
where
BCR = benefit/cost ratio of providing the higher level right-turn treatment over the provision of the
minimum turn treatment (BAR or BAL)
ARM = reduction in Rear-End-Major vehicle accidents per year per major leg according to Equation 2
below number of Rear-End-Major vehicle accidents per year for a minimum turn treatment
(BAR or BAL) minus the number of Rear-End-Major vehicle accidents per year for the higher
level treatment (accidents per year)
CA
= average cost of a Rear-End-Major vehicle accident = $38,974 (Arndt 2004)
TDL = design life (years)
CTT = additional construction cost - the cost of construction of the higher-level turn treatment minus
the cost of construction of the minimum turn treatment (BAR or BAL)
ARM 2.75 10 12 Qi
0.406
QM
0.912
S MT
2.94
(eTTM eTTA )
(2)
where
Qi
QM
SMT
TTM
TTA
QT1
QR
QT2
QL
Turn Type
Splitter Island
QM (veh/h)
Right
No
= QT1 + QT2 + QL
Right
Yes
= QT1 + QT2
Left
No/Yes
= QT2
Substituting Equation 2 into Equation 1 and rearranging produces Equation 3 to solve for Qi.
Qi
0.912
2.94
TTM
TTA
S MT (e e )
C A TDL QM
2.46
(3)
Two warrant curves were developed for ease of use with the first warrant for high speed roads and the
second warrant for intermediate and low speed roads. These warrants were adopted in Austroads (2009)
and have been retained in Austroads (2013).
1
80
60
CHR
AUL or CHL
A
40
CHR(s)
AUL(s)
20
BAR*
BAL
0
0
200
400
600
800
1000
1200
120
100
A
80
CHR
AUL or CHL
60
CHR(s)
AUL(s)
40
20
BAR*
BAL
0
0
200
400
600
800
1000
1200
1400
1600
3. International Practice
Prior to developing new practice, a review was undertaken of guidance used internationally for the
selection of turn treatments at unsiganalised intersections.
Canada
Canadian practice is documented in the Geometric Design for Canadian Roads (Transportation
Association of Canada 1999) and defines two separate warrant methodologies for determining the
requirements for left-turn lanes from two-lane and four-lane roads.
The first, a volume warrant, is based on a capacity / conflict assessment and refers to the practices
documented in the Highway Capacity Manual (Transportation Research Board 2010). This assessment
determines the level of interference to through traffic and the probability a left turning vehicle from the
major road, or queue of vehicles, will obstruct following through vehicles.
The second, a collision warrant, is based on a the historical crash rates at each intersection where four or
more collisions related to left turns occurs per year or where six or more occur within a period of two
years, provided the collisions are of a type which could reasonably be expected to be eliminated by
provision of a left turn lane.
It appears that the Transportation Association of Canada (1999) warrants do not take into consideration
the speed of through traffic on the major road in determining the preferred turn treatment.
United States of America
USA practice in the Green Book (AASHTO 2011) notes that, left turning traffic should be removed
from through lanes whenever practical and the provision of left-turn lanes can reduce intersections crash
rates by 20 to 65%. On two-lane highways, a tabulated guide is provided detailing the traffic volumes
where left-turn lanes should be considered. This practice is based on the work undertaken by Harmelink
(1967) considering the rate of conflicts between turning vehicles and approaching through vehicles.
The NCHRP document Development of Left-Turn Lane Warrants for Unsignalized Intersections
(Fitzpatrick et al 2014) provides a comprehensive evaluation of warrants across the US for left-turn lanes
including the current practices documented in the AASHTO (2011). This report considers the various
methodologies undertaken across the USA and provides a detailed evaluation of each.
Fitzpatrick et al (2014) further recommends a new approach based on a benefit-cost assessment. This
benefit-cost assessment considered both the delay costs (operations) and the crash costs (safety) for an
intersection which then allows the selection of the most appropriate treatment. Crash costs have been
based on research into crash rates at a range of intersections and cost values for typical crash types. Delay
costs are estimated from a simulation model approach. Construction costs have been estimated from a
series of similar roads projects across the USA.
Similar to the Australian warrants, these new warrants are presented in the form of tables and graphs for
arterial roads and high-speed rural roads and are documented for recommended inclusion in the next
edition of the AASHTO (2011). The use of multiple warrants figures takes into consideration the speed
of through traffic on major road with higher-order turn treatments required at lower traffic volumes at
high-speed locations.
United Kingdom
UK practice is documented in Highways Agency report Geometric Design of Major/Minor Priority
Junctions (Highways Agency 1995) and provides guidance for the selection of the appropriate initial Tjunction option from a figure using the design year traffic flows on the major and minor roads. The
analysis behind the preferred intersection types are stated to take into account geometric and traffic
delays, entry and turning traffic flows and accident costs. Designers can select a Simple, Ghost Island,
Single Lane Dualling or Roundabout intersection form.
The element of this guidance relevant to the current paper is the boundary between a Simple intersection
type and a Ghost Island. This is represented as a straight line relationship with a 2-way minor road
volume of 300 vehicles per day allowable up to a major road two-way volume of 13,000 vehicles per day.
There is no apparent consideration of the speed of through traffic on the major road in these warrants.
Comparison of Warrants Methodologies
Two parameters have historically been used to determine warrants for left-turns from unsiganalised
intersections:
Much of the early work was based on conflict rates at intersections with Harmelink (1967) being one of
the earliest documented processes. Practice in the US (AASHTO [2011]) and Australia have for a
number of years been based on conflict rate based warrants. Canadian practice has been based on traffic
operations performance together with a consideration of actual crash rates at an intersection relating to the
left turn from the major road. UK practice is stated to taken into account both traffic operations
performance as well as crash costs but the warrant provided only allows a simple treatment at very low
volumes for the turning movement.
Practice adopted in 2009 in Australia (Austroads 2014) and now proposed for the USA (Fitzpatrick
2014), is to move to a warrant based on the forecast crash performance of the turn treatment, and a
Benefit/Cost Ratio (BCR) consideration of the resultant crash costs against the construction costs to
upgrade the intersection. The proposed US warrants are presented in Figure 4 for rural highways and for
urban arterials, while the Australian warrants are presented for major road speeds greater than 100km/h
(62 mile/h) and less than 100km/h.
The USA practice goes further to include the costs associated with traffic delays in the BCR analysis. It
is interesting to note that in (Fitzpatrick 2014) the delay costs in the examples given range from less than
one percent to just 2 percent of the total intersection benefits associated with the left-turn lane. The
remainder of the benefits are associated with reduced crash rates at the intersection.
An analysis of the cost of traffic delays was undertaken for the left-turn movement for Australian
intersections based on a desired Level of Service C operation (Part 3 of the Austroads Guide to Traffic
Management [2010]). The outcomes revealed that the costs associated with reduced delays were much
lower than the costs associated with reduced crash rates. Due to the relative impact of each element on
the BCR analysis, Australian warrants were therefore determined to require only an analysis of the safety
performance of an intersection.
Figure 4 provides a comparison of the current Australian and proposed USA warrants and demonstrates
remarkable similarities between the shapes of the relationships. approximate UK warrant is also shown
for comparison.
Design Speed: a lower design speed reduces the differential crash rate at the intersection and
allows a higher turning traffic volume to be accommodated for the same BCR outcome
CTT
Additional Construction Cost for the higher order turn treatment: a lower design speed reduces
the length of turn tapers and deceleration lengths, lowering the costs of the CHR(s) and CHR
treatments. To achieve the same BCR outcome a reduced turning volume is required.
The resultant warrant for a design speed <70km/h is presented at Figure 5. The warrant curves shift to the
right compared with the existing curves as shown in Figure 3. The warrants for design speed between
70km/h and 100km/h are depicted as dashed lines. This relationship mirrors that between the current
curves for high speed and mid speed roads.
Notes:
Curve 1 represents the boundary between a BAR and a CHR(s) right turn treatment and between a
BAL and an AUL(s) left turn treatment
Curve 2 represents the boundary between a CHR(s) and a CHR right turn treatment and between a
AUL(s) and an AUL/CHL left turn treatment
Area A represents traffic volumes scenarios where the turning volume is greater than the average
through volume in each direction.
5. Brownfield Sites
The existing turn treatment warrants (Arndt & Troutbeck 2006) were developed based on the generalised
construction cost at greenfield sites where the construction of the intersection occurs as part of the same
works for the through road. A greenfield site is a location on which a new road is being built where there
is no development and compliance with full design standards is relatively easy to achieve. The additional
construction cost just relates to the incremental costs associated with the turn treatments, in comparison
with the costs to construct the through carriageway only.
A significant volume of requests received by TMR relate to the use of the existing warrants in brownfield
situations. A brownfield site is a location where infrastructure (e.g. roads and buildings) exists and
removing, altering or adjusting this infrastructure can be very expensive.
Practice has generally been to apply the existing warrants in brownfield sites. As an alternative, TMR
(2006) provides a method of calculating the safety cost of the various turn treatments for comparison to
actual construction costs. This methodology however has not been widely applied and has been adopted
in TMR (2014) as a Design Exception process. Therefore a warrant based approach for application at
brownfield sites was considered a necessary addition to the current warrants.
At brownfield sites, the construction costs for the intersection turn treatments can vary significantly. At
some sites, costs relating to relocation of utilities / services, to underground drainage systems and to
structures (e.g. bridges over watercourses) can be significant and highly variable. Where these large
incremental costs apply, the warrants approach developed in this paper would not be appropriate.
The incremental costs for the construction of the intersection pavement can be relatively well defined and
at most intersections may be considered to sufficiently describe the incremental intersection construction
sites. For the development of brownfield warrants, the analysis has been undertaken on the basis that
upgrades would in most cases be planned to avoid large variable cost items.
The brownfield warrants have been developed through application of Equation 3 for each of high speed
(110km/h), mid speed (80km/h) and low speed (60km/h) situations.
The only parameter in the analysis for brownfield sites that varies, from that for greenfield sites, is the
construction cost (CTT) for the turn treatment. As the intersection works need to be built around an
existing road, rather than constructed as part of the works on the through road, construction activities have
a lower level of efficiency and higher construction rates apply. The costs for management of existing
traffic on the through road also need to be included. The estimated constructing costs for brownfield
intersections are therefore higher than for greenfield sites.
The increased construction cost for the turn treatment at brownfield sites results in an increase in the
combined traffic volumes that can be accommodated to achieve the same BCR result. Consequently the
warrant curves shift to the right of the existing greenfield warrants. The resulting brownfield warrants for
the high speed situation (design speed 100km/h) are shown in Figure 6. The greenfield warrants are
also shown for comparative purposes.
Figure 5 demonstrates that there is a significant shift of the warrants curves to the right for brownfield
intersections in comparison with greenfield intersections. The graphs for mid speed and low speed
situations revealed similar results.
80
CHR(s)/AUL(s) - CHR/AUL
brownfield
greenfield
60
40
BAR/BAL - CHR(s)/AUL(s)
brownfield
greenfield
20
0
0
200
400
600
800
1000
1200
1400
Major Road Traffic Volume 'QM' (Veh/h)
Design Speed 100km/h
1600
The expectation by drivers that two intersections with the same traffic conditions should be
designed and built to the same standard.
The issues associated with the resulting likely higher crash rate at brownfield intersections and
potentially an earlier requirement to upgrade the intersection.
In considering these outcomes, the approach adopted by TMR is for the existing greenfield warrants to be
applied as Normal Design Domain (NDD) criteria. The brownfield warrants are therefore alternative
design criteria for application at sites subject to site constraints or where the costs of the intersection
upgrade, to greenfield criteria, are considered impractical. These therefore represent Extended Design
Domain (EDD) criteria. Refer to Austroads (2006) for a discussion on Normal and Extended Design
Domain.
10
2.75 10 12 Qi
0.406
QM
0.912
S MT
2.94
TT
(4)
This equation describes the average annual crash rate if a BAR turn treatment was retained at a site for
each of the existing warrant boundaries, CHR(s) and CHR. Table 1 details the results of this analysis.
In applying this approach to establish crash rates for intersections on very low volume roads, an analysis
was undertaken firstly for high speed roads to establish the equivalent curves equating to 2%, 1% and
0.5% annual average crash probabilities (one crash every 50yrs, 100yrs and 200yrs respectively) during
the design hour as shown in Figure 7.
Table 1: Indicative annual average crash probability
Crash Probability(3)
(% per year)
Design Speed
(1)
CHR(s) Boundary
CHR boundary
High
SMT 100km/h
5.5
11.0
Medium(1)
3.2
7.0
SMT 70km/h
2.1
4.7
(2)
Low
Notes:
The curve representing a 0.5% average annual probability of a crash during the design hour provided a
close fit to the right hand section of a proposed guideline currently used in TMR regional areas for high
speed roads. A similar analysis undertaken for low and mid speed roads (refer Sullivan & Arndt 2014)
established that crash rates of 10% of the expected crash rate at the BAR / CHR(s) boundary closely
approximated the proposed guidance.
In undertaking this analysis, the following issues were considered;
The data collected by Arndt (2004) did not include a significant number of sites at which traffic
volumes fell into the category of very low volume roads. Any use of this approach to therefore
predict crash rates is based on an extrapolation of the existing data.
The crash rates at very low volume intersections are considered likely to be so low that other
factors may be of more significance in determining site specific crash rates.
Further discussion regarding these results is presented following the discussion on Method 2.
11
Note:
Area A represents traffic volumes scenarios where the turning volume is greater than the
average through volume in each direction.
The rates of conflict for each of these three event types is calculated through application of the following
theory described in Austroads (2008) Guide to Traffic Management Part 2: Traffic Theory.
a)
The traffic distribution has been assumed to be a displaced negative exponential distribution where
the proportion of right turning vehicles delayed by opposing through vehicles is given by Equation
5.15 from Austroads (2008).
Proportion Delayed 1 e
b)
(5)
The average stopped delay to those vehicles stopped as per Equation 5 is given by Equation 5.19 of
Austroads (2008)
d av (d 0)
qe
c)
q (T )
1 q
q (T )
1 q
(T )
(1 e
q (T )
1 q
(6)
The time taken by right turning vehicles to decelerate from the design speed is calculated from:
v u at
(7)
12
=
=
dav =
v =
u =
a =
t =
g =
The probability of a through vehicle, following a right turning vehicle, needing to react in some way
(slow, stop or manoeuvre) to avoid a conflict with the right turning vehicle can then be determined from
the headway distribution behind a right turning vehicle.
Pr( h ) 1 e
q ( )
1 q
(8)
Where t is determined as follows with the second term as either dav for a vehicle following a stopped right
turn vehicle, or zero for a vehicle following a right turning vehicle that does not stop.
u
(d av ,0) 1.0
2 0.36 g
(9)
It is acknowledged in these calculations that use of the average delay for all right turning vehicles needing
to stop is a generalisation and that there will in fact be a distribution of delays. However, for the purpose
of this review to test the overall methodology, the use of a single uniform delay is considered sufficient.
This methodology was applied to the existing warrants curves and revealed that the CHR(s) / BAR
warrant boundary for high speed situations represents the range of conflict rates in Table 2.
Table 2: Indicative conflict rates for design hour traffic volumes
Conflict Rate(2)
(conflicts per hour)
Design Speed
High(1)
Notes:
SMT 100km/h
CHR(s) Boundary
CHR boundary
1.8 4.0
4.4 14.0
Based on this process and these results, an analysis was undertaken for high speed roads to establish the
equivalent curves equating to peak hour conflict rates of 1/hr, 0.5/hr, 0.2/hr and 0.1/hr (equivalent
conflicts = 2200/yr, 1100/yr, 450/yr, 225/yr respectively). The results of this analysis are presented in
Figure 8.
13
Note:
Area A represents traffic volumes scenarios where the turning volume is greater than the
average through volume in each direction.
High Speed Roads - 0.5% average annual crash probability during the peak hour
Mid Speed Roads 0.33% average annual crash probability during the peak hour
Low Speed Roads 0.2% average annual crash probability during the peak hour
In considering the use of intersection treatments below a BAR / BAL, it is also considered appropriate the
concepts of NDD and EDD should apply. Under this concept the following recommended position
proposed;
The form of these warrants is discussed in the recommendations section of this paper.
14
7. Multi-lane Roads
Arndt (2004) incorporated data collected from a range of sites, including a number of four and six lane
roads. However, the warrants developed in Arndt & Troutbeck (2006) did not explicitly discuss
application to the number of through lanes on the roadway. The issue of warrants for turn facilities on
multi-lane roads has regularly generated queries. This has therefore driven a need to more explicitly
examine these situations.
Arndt (2004) found that the crash rate for a Multi-lane No Right turn facility (MNR - refer Figure 1) on
four and six lane roads was significantly higher than for a similar volume BAR on a two lane road. The
Turn Treatment constant (refer Equation 2) for an MNR was calculated at 4.59, compared with 3.83 for a
BAR representing a doubling of the expected crash rate at similar traffic volumes when used in Equation
1. By comparison, there was no significant difference in the crash rate for a CHR(S) or CHR turn
treatment on a four / six lane road compared to a two lane road with a CHR(S) or CHR.
In considering the development of warrants for application on multi-lane roads, the current TMR policy
on these treatments is a relevant consideration. Consequently, it is considered unlikely that greenfield
multi-lane roads would be constructed as undivided roads. This together with the markedly higher crash
rate from MNR treatments results in the TMR preferred position that new four or six lane roads will be
designed with some form of median. For the purpose of turn treatment warrants, the minimum right turn
treatment allowed from multi-lane roads is therefore a CHR(S) treatment. This paper does not consider
the development of a warrant which describes a MNR facility.
The warrants for multi-lane roads have therefore been developed applying the same methodology outlined
in Appendix A and using the following parameters.
= 4.59
= $0
The resultant warrant curves for high speed roads showed very little difference compared to the existing
warrants for 2L2W roads. Due to the variability in crash rates at intersections the separation between the
warrants was not considered significant. In addition the left turn warrants are based on the right turn
warrants and consequently they contain an additional level of conservatism. Therefore the left turn
warrants do not signify any particular accuracy with respect to combined traffic volumes based on a BCR
analysis.
It is therefore recommended that the warrant for a 2L2W road be adopted to provide consistency in
application. This would ensure that the warrant for a CHR treatment is required at the same traffic
volume levels for all roads by using the traffic volumes in the closest through lane to the turning lane
(refer Figure 9) for multilane roads.
A similar analysis was undertaken for brownfield intersections on 4 and 6 lane roads. The results of this
analysis demonstrated results similar to that for the greenfield intersections. As for greenfield sites the
recommended approach is therefore to adopt the 2L2W warrant to provide for a level of consistency in
application.
The resulting recommendations with regards to these warrants are therefore contained in the
recommendations section of this paper.
15
(10)
The crash reduction benefit arising from the analysis of the upgraded intersection treatment, subject to a
short period of high traffic demand, across the design life of the intersection can therefore be given as
C RM C RM ( normal)
(TDL TH )
T
C RM ( high) H
TDL
TDL
(11)
where
CRM(normal) = crash reduction benefit ($) for normal traffic conditions for design life
CRM(high) = crash reduction benefit ($) for high traffic conditions assumed across design life
TH
= Period (yrs) for which higher traffic volume conditions occur
The analysis in Appendix A can then be shown to reveal the following relationship between the traffic
volumes during the high period and traffic volumes during normal conditions.
Qi ( high)
406
CTT TDL C QM0.912 Qi0(.normal
) (TDL TH )
C QM0.912 TH
(12)
where
Qi(high)
Qi(normal)
Within this equation a relationship between the turning volume during the high traffic conditions and the
through traffic flow on the major road (QM), can therefore be determined for a given normal turning
traffic volume and the period for which the higher traffic volumes exist. All the remaining parameters are
typically given a single value. Examples of the resultant warrants are shown in Figure 9 for high speed
roads. In these curves points to the left of the curve require a BAR, and points to the right of curve
require a CHR(s)
16
150
100
50
0
0
100
200
300
400
500
600
700
800
150
100
50
0
50
100
150
200
250 300
350
400
Major Road Traffic Volume 'QM' (Veh/h)
150
450
500
100
50
0
0
50
100
150
200
Major Road Traffic Volume 'QM' (Veh/h)
1
150
250
300
3 4 5
50
0
0
Notes:
Curve 1:
Curve 3:
Curve 5:
50
5 yrs <
1 yr <
100
150
200
Major Road Traffic Volume 'QM' (Veh/h)
TH
TH
TH
Curve 2:
2 yrs
Curve 4:
6 months
2 yrs <
6 months <
250
TH
TH
5 yrs
1 yr
Figure 9: Warrant analysis for short term high volume use of intersection High speed
AITPM 2015 National Conference, Brisbane, Australia
17
7. Conclusion
In considering the results in each of the previous sections, the new warrants detailed in the Appendix are
recommended for application in determining the appropriate form of intersection turn treatments.
The warrants recommended are based on the crash performance of the different turn treatments.
Consideration must also be given to the operational performance of the intersection which may require a
higher level turn treatment, or alternative intersection control, particularly for higher traffic volumes.
The existing warrants have proven to be the favoured tool for application in determining the preferred
intersection turn treatments from the major road at unsignalised intersections. However, there is now a
need to supplement the existing warrants to provide additional guidance to address a wider range of
situations.
The warrants developed in this paper provide additional guidance for lower speed roads, brownfields
situations, simple intersection turn treatments, multi-lane roads and short term high volume use of
intersections. This new set of warrants now provides more complete guidance for designers in selecting
the preferred intersection turn treatment.
18
REFERENCES
AASHTO (2011) A Policy on Geometric Design of Highways and Streets. American Association of State
Highway and Transportation Officials, Washington D.C., USA,
Arndt OK (2004) Relationship between unsignalised intersection geometry and accident rates. Doctor of
Philosophy Thesis, Queensland University of Technology and Queensland Department of Main
Roads, Brisbane, Australia.
Arndt OK and Troutbeck RJ (2006) New Warrants for Unsignalised Intersection Turn Treatments. 2006
ARRB Conference, Melbourne, Australia.
Austroads (2006) Guide to Road Design Part 2: Design Considerations, Austroads, Sydney, Australia
Austroads (2008) Guide to Traffic Management Part 2: Traffic Theory, Austroads, Sydney, Australia
Austroads (2009) Guide to Road Design Part 4a: Unsignalised and Signalised Intersections, Austroads,
Sydney, Australia
Austroads (2010) Guide to Road Design Part 3: Geometric Design, Austroads, Sydney, Australia
Austroads (2013) Guide to Traffic Management Part 6: Intersections, Interchanges and Crossings,
Austroads, Sydney, Australia
Austroads (2013a) Guide to Traffic Management Part 3: Traffic Studies and Analysis, Austroads, Sydney,
Australia
Fitzpatrick, K., M. A. Brewer, J. S. Gluck, W. L. Eisele, Y. Zhang, H. S. Levinson, W. Von Zharen, M.
R. Lorenz, V. Iragavarapu and E. S. Park. (2014) Development of Left-Turn Lane Warrants for
Unsignalized Intersections. NCHRP Web Only Document 193, Transportation Research Board,
Washington DC, US, 2010.
Harmelink, M. D. (1967) Volume Warrants for Left-turn Lanes at Unsignalized Grade Intersections.
Department of Highways, Ontario, Canada.
Highways Agency (1995) Geometric Design of Major/Minor Priority Junctions. Highways Agency, TD
42/95, London, UK.
Sullivan & Arndt (2014) Expanded Warrants for Unsignalised Intersection Turn Treatments (TMR
Internal Report), Queensland Department of Transport & Main Roads, Brisbane, Australia
Transport and Main Roads (2006) Road Planning and Design Manual (1st Ed) Chapter 13, Queensland
Department of Transport & Main Roads, Brisbane, Australia
Transport and Main Roads (2014) Road Planning and Design Manual (2nd Ed) Part 4A: Unsignalised and
Signalised Intersections, Queensland Department of Transport & Main Roads, Brisbane, Australia
Transportation Association of Canada (1999) Geometric Design Guide for Canadian Roads,
Transportation Association of Canada, Ontario, Canada.
Transportation Research Board (2010) Highway Capacity Manual. Transportation Research Board,
Washington D.C., USA.
19
An additional warrant for roads with speeds 70km/h is added to the existing warrants in Transport and
Main Roads (2006) and Austroads (2013).
At greenfield intersections on 2L2W roads, the minimum turn treatment shall be a BAR / BAL.
Intersections without any widening (Simple intersection treatments) should not be constructed at new
intersections on new roads.
In considering 4-lane, 2-way (4L2W) and 6-lane, 2-way (6L2W) situations the same warrant curves as for
2L2W roads should be applied. The major road traffic volume (QM) is calculated as per note 5 to Figures
8 and 9. At new four and six lanes roads, it is assumed that a median of sufficient width would be
included to accommodate a CHR(s) or CHR treatment at every intersection where a right turn from the
major road is allowed.
2.
At brownfield sites, the incremental costs for the construction of either a new intersection on an existing
road, or for the upgrade of an existing intersection are significantly increased due to the piecemeal nature
of the works and the management of existing traffic.
The warrants for brownfield sites represent alternative design criteria for application at sites subject to site
constraints or where the costs of the intersection upgrade are considered impractical. These criteria are
therefore proposed as Extended Design Domain (EDD) warrants.
An additional consideration is that many intersections, especially lower volume intersections, have been
constructed without any widening and hence do not meet the minimum design layout for a BAR / BAL.
These intersections are referred to in this paper as Simple Intersections (SR/SL). Warrants have therefore
been proposed which allow the retention of these intersections at brownfield intersections.
3.
Intersections in regional areas across Queensland are at times subject to a limited period of heavier traffic
due to the development of Coal Seam Gas (CSG) infrastructure, short term agricultural activities, mining
activities or other developments. During these periods the traffic volumes may be substantially higher
than the base traffic conditions. However, in many cases it is not considered appropriate to upgrade the
intersection just for the short period of high use.
Using a balanced design life assessment of the Crash Reduction potential of the upgraded intersection
treatment, an alternative set of warrants were developed including parameters for a shorter period TH of
high turn volumes.
20
80
60
CHR
AUL or CHL
A
40
CHR(s)
AUL(s)
20
BAR*
BAL
0
0
200
400
600
800
1000
1200
120
100
A
80
CHR
AUL or CHL
60
CHR(s)
AUL(s)
40
20
BAR*
BAL
0
0
200
400
600
800
1000
1200
1400
1600
150
125
A
CHR
AUL or CHL
100
75
CHR(s)
AUL(s)
50
BAR*
BAL
25
0
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Figure A1: Warrants - major road turn treatments - Normal Design Domain.
21
80
A
60
CHR
AUL or CHL
40
CHR(s)
AUL(s)
20
BAR*
BAL
0
0
200
400
2L2W only
SR
SL
600
800
1000
1200
1400
1600
120
100
A
80
CHR
AUL or CHL
60
CHR(s)
AUL(s)
40
20
BAR*
BAL
0
2L2W only
SR
SL
150
200
400
600
800
1000
1200
1400
1600
1800
2000
125
A
CHR
AUL or CHL
100
75
CHR(s)
AUL(s)
50
BAR*
BAL
25
0
0
2L2W only
SR
SL
200
400
600
800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000
Major Road Traffic Volume 'QM' (Veh/h)
Design Speed 70km/h
22
Curve 1 - For 2L2W roads represents the boundary between a BAR and a CHR(S) turn
treatment and between a BAL and an AUL(S) turn treatment. For 4/6L2W roads represents
the boundary between a BAL and an AUL(S) turn treatment only. The minimum right turn
treatment is a CHR(s) on 4/6L2W roads.
2.
Curve 2 represents the boundary between a CHR(S) and a CHR turn treatment and between
an AUL(S) and an AUL / CHL turn treatment. The choice of CHL over an AUL will depend
on factors such as the need to change the give way rule in favour of other manoeuvres at the
intersection and the need to define more appropriately the driving path by reducing the area
of bitumen surfacing.
3.
Curve 3 represents the boundary between a Simple Intersection Treatment and a BAR turn
treatment. This curve applies to EDD only.
4.
The warrants apply to turning movements from the major road only (the road with priority).
For turns from the minor road, turn treatments are determined through an operational
performance evaluation applying gap acceptance analysis and an evaluation of acceptable
delays and queues.
5.
QM
a. For 2L2W roads, Figure A3 is to be used to calculate QM.
b. For 4/6L2W roads, QM for right turns uses the full opposing flow QT2 and only the traffic
flow in the nearest lane of the following flow (50% of QT1 for a four lane road, and 33%
for a six lane road).
QT1
QR
Road Type
2 Lane
2 Way
4 Lane
2 Way
6 Lane
2 Way
Turn Type
Right
Left
Right
Left
Right
Left
QT2
QL
Splitter Island
No
Yes
Yes / No
No
Yes
Yes / No
No
Yes
Yes / No
QM (veh/h)
= QT1 + QT2 + QL
= QT1 + QT2
= QT2
= 50% x QT1 + QT2 + QL
= 50% x QT1 + QT2
= 50% x QT2
= 33% x QT1 + QT2 + QL
= 33% x QT1 + QT2
= 33% x QT2
Figure A3: Calculation of the Major Road traffic Volume Parameter QM.
23
6.
Traffic flows applicable to the warrants are peak hour flows, with each vehicle counted as
one unit (i.e. do not use equivalent passenger car units [pcus]).
7.
If more than 50% of the traffic approaching on a major road leg turns left or right,
consideration needs to be given to possible realignment of the intersection to suit the major
traffic movement. However, route continuity issues must also be considered (for example,
realigning a highway to suit the major traffic movement into and out of a side road would be
unlikely to meet driver expectation).
8.
If a turn is associated with other geometric minima, consideration should be given to the
adoption of a turn treatment of a higher order than that indicated by the warrants.
9.
At higher traffic volumes, consideration should also be given to the operational performance
of the intersection which may require a higher level turn treatment, or alternative intersection
control, than required by these warrants based on crash analysis.
10.
24