Supplement to Austroads Guide to
Traffic Management
Part 2: Traffic Theory (2015)
OCTOBER 2015
vicroads.vic.gov.au
Summary of Changes to Austroads Guide to Traffic Management, Part 3 (2008) – Edition 1
<month> 2015
1
Table of Contents
1.
Introduction ..................................................................................................... 3
1.1
General ............................................................................................................. 3
1.2
How to Use this Supplement ............................................................................. 3
2.
Classification of Supplement Information .................................................... 4
3.
Details of Supplement Information ................................................................ 5
Section 7.5 Congestion Management Theory ............................................................... 5
7.5.2 Flow Monitoring and Management ....................................................................... 5
7.5.5 Other Models of Flow Breakdowns ....................................................................... 6
SECTION 8 PRINCIPLES UNDERLYING MANAGED MOTORWAYS ............................ 9
8.2 Motorway Operational Capacity ............................................................................. 10
8.2.1 Capacity of Segment ............................................................................................ 10
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
2
1. Introduction
1.1
General
All road agencies across Australia are working towards greater consistency between States/Territories in how
road networks are managed. In order to achieve this, the Austroads Guide to Traffic Management (AGTM)
and Australian Standards relating to traffic management have been adopted to assist in providing that level of
consistency and harmonisation across all jurisdictions. This agreement means that these Austroads Guides
and the Australian Standards are the primary technical references.
Austroads Guide to Traffic Management, Part 2: Traffic Theory (2015) is a nationally agreed guideline
document and has been adopted by all jurisdictions, including VicRoads.
All jurisdictions will be developing their own supplement to clearly identify where its practices currently differ
and to provide additional guidance to that contained within Austroads Guide to Traffic Management, Part 2:
Traffic Theory (2015). This document is the VicRoads supplement and shall be read in conjunction with
Austroads Guide to Traffic Management, Part 2: Traffic Theory (2015).
1.2
How to Use this Supplement
There are two key parts to this document:
Classification of Supplement Information: this table classifies supplement information as a Departure,
Additional Information or both. This information assists with identifying its hierarchy in relation to the
Austroads Guide to Traffic Management, Part 2: Traffic Theory.
•
Details of Supplement Information: this section provides the details of the supplement information.

Departures: where VicRoads practices differ from the guidance in the Austroads Guide to Traffic
Management (AGTM). Where this occurs, these differences or ‘Departures’ will be highlighted in a
box. The information inside the box takes precedence over the AustRoads Guide to Traffic
Management section. The Austroads Guide to Traffic Management, Part 2, section is not applicable in
these instances.

Additional Information: all information not identified as a departure provides further guidance to the
Austroads Guide to Traffic Management, Part 2: Traffic Theory and is read and applied in
conjunction with the Austroads Guide to Traffic Management, Part 2: Traffic Theory, section.
Where a section does not appear in the body of this supplement, the Austroads Guide to Traffic Management,
Part 2: Traffic Theory (2015) requirements are followed.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
3
2.
Classification of Supplement Information
The classification of each clause as a Departure, Additional Information or both is shown in the table below.
Section
Classification
7.5.2
Additional Information
7.5.5
Additional Information
8.0
Additional Information
8.2
Additional Information
8.2.1
Additional Information
The Austroads Guide to Traffic Management, Part 2: Traffic Theory (2015) requirements are followed for
sections not shown in this table.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
4
3.
Details of Supplement Information
Section 7.5 Congestion Management Theory
7.5.2 Flow Monitoring and Management
Density and occupancy
The three primary variables used to describe traffic flow are identified in Section 2 as volume (q), density (k)
and speed (v) , which, in aggregate terms, are related by q =k.v (Equation 2.3), in which the appropriate v is
the space mean speed.
It has long been recognised that density is a fundamental measure of the level of service (LOS) being
provided on a road at any particular time (e.g. HRB 1965) but, until relatively recently, the difficulties of field
measurement of density led to the use of other LOS measures such as volume/capacity ratio. Historically,
density (the number of vehicles in a unit length of lane or road)has been measured in the field by one of four
methods, as follows:
•
Photographic techniques measure density directly using photographs along a length of road, taken
either from a fixed, high vantage point or from an aircraft. From the photographs, the number of vehicles in
each length of road or lane that is of interest are counted and the density is obtained by dividing by the
known length of road or lane. This method generally is no longer used for assessment of complex
networks. Other methods utilising modern detection techniques are used in this instance.
•
Input-output counts enable the number of vehicles in a road section to be updated from an initial, known
number by adding counts of vehicles entering the section and subtracting counts of vehicles leaving. The
passage detectors must be able to ensure accurate countsat both ends of the section and a means of
regularly re-initialising the number of vehicleswithin the section is desirable. Such re-initialisation is difficult
except in the situation of road sections with no intermediate entry or exit points and no lane changing, in
which case the number of vehicles in each lane of the section can be obtained as the count of vehicles
entering between the entry and exit of a specifically identified vehicle.
It is well known that the errors in passive count between two detectors sites accumulate very quickly and
the need for reinitialising (or correction through other means) will be continuous - (every few minutes).
Hence the density calculations obtained by this type of measurement must take into account the sum of
the errors in both input and output measurement points. Practitioners who have used this method for
measuring density for real time ramp metering algorithms have often been oblivious to the error
accumulation issue and the fact that between two consecutive detector sites that may be 500-1000m
apart, the space provides potential for significant levels or traffic bunching to occur resulting in flow
breakdown densities which are not always measurable using this method (i.e. if a bottle neck occurs
between two count locations, the resulting computation will be misleading). The error problem introduced
by this method would calculate a density error that would be outside the error range required to control
traffic for managed motorway operations (refer to VicRoads Ramp signals Handbook).
•
This method is only generally useful for post project performance monitoring and, in this case, should still
be used with care.
•
Occupancy measurement became a viable means of determining density with the introduction of
accurate presence detectors. Occupancy (at a given location over a (usually fairly short) period of time) is
defined as the proportion of time for which the presence of a vehicle over the detector is recorded. Given
that presence is recorded whenever any part of a vehicle length is over any part of the effective length of
the detector, occupancy is related to the average spacing of vehicles by the relationship:
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
5
Traffic density and per cent occupancy ranges corresponding to different levels of service are
shown in Table 7.1, which has been adapted from May (1990) and Transportation Research Board
(2010).
Table 7.1: Density and occupancy level of service indicators
In relation to Table 7.1, the occupancy measurement calculated by the many different detector devices used
in Australia would benefit from having an agreed standard (i.e. the L D measurement = 2m). This would avoid
much confusion amongst practitioners who use different detection types where, for example, Level of Service
E occupancy might vary widely between 8% and 25% (not 14-19%) depending on the technology used e.g
Sensys, loops, CCTV, TIRTL etc.
7.5.5 Other Models of Flow Breakdowns
Six traffic state model
Schonhof and Helbing (2007) investigated 1 min data for the same section on the A5 freeway as Kerner and
Lindgren. They interpreted traffic flow by six states: free traffic (FT), pinned localized cluster (PLC), moving
localised cluster (MLC), stop-and-go waves (SGW), oscillating congested traffic (OCT) and homogeneous
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
6
congested traffic (HCT). The most frequent states at the investigated freeway were the PLC and OCT states.
HCT occurs mainly after serious accidents with lane closures or during public holidays. An adaptive smoothing
method was used to identify the different traffic states. This method interpolates and smoothes traffic data
from successive freeway sections, taking into account the propagation speeds of perturbations in free and
congested traffic.
Flow breakdown and the resulting shockwaves types do not occur purely based on volumes alone but are
strongly influenced by the delta (lane changing) which is associated with weaving, merging and diverging.
Refer diagrams below showing relationship between lane flows, lane changing and resultant traffic states.
All the bottlenecks shown in Figure 8.3 were not only associated with entry ramps but also geometric
constraints such as grades and curvature and occurred just upstream of a mid-block lane gain that induced
lane changing spikes e.g delta lane changing. Many of the wave types identified by Helbing et al are
discernible in VicRoads data and this has lead to improved understanding of the cause of flow breakdown and
assisted with further tuning of managed freeway systems - see VicRoads SVO graphs below:
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
7
Schonhof and Helbing (2007) found that the congested traffic states identified by this model were in good
agreement with prediction of some second-order macroscopic traffic models and some microscopic carfollowing models.
Readers should refer to Austroads (2008), Austroads (2009a) and Han and Luk (2008) for further information
regarding these flow breakdown models and their applications in the identification and analysis of freeway flow
breakdowns.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
8
SECTION 8 PRINCIPLES UNDERLYING MANAGED
MOTORWAYS
In recent years, road agencies have focussed on the management of motorways under congested flow
conditions using e.g. ramp meterings and variable speed limit (VSL) signs. Austroads (2014a) and VicRoads
(2013) described the general principles underlying the use of these managed motorway tools.
The simplified approach to the fundamental relationships provided in Section 2 may mask the fact that the
extreme maxima/minima relating to feasible headway, speed and density outcomes are of limited value for
traffic modelling, road design and traffic operations as, whilst limits on the speed-density-volume diagrams
shown are theoretically and operationally feasible, we now know that these points are in fact unstable by
nature and need to be used with care when informing traffic operations or modelling or design purposes.
The following diagrams from Traffic Flow Dynamics - Treiber and Kesting Figure 4.12 (refer Fig 8.1x) show
that these relationships are not contiguous functions and that the critical points to be used for modelling,
design and operations to provide stable on road flow outcomes can fall well below the maxima/minima
identified in the curves.
Fig 8.1x (From Treiber and Kesting – Traffic Flow Dynamics – Data, Models and Simulation)
For example, in practice it is possible to operate a managed freeway at higher sustainable flows if it is
operated at the stable points which usually always fall significantly below the maximum points indicated by the
speed-volume-density relationship diagrams (refer Fig 8.1x above). Note this is consistent with the Stochastic
Concept comment documented by Brilon in Section 7.5.5 that the optimum degree of saturation is around
90%. This is essential information for modellers, designers and operators. As an example, the optimum
degree of saturation for a 2 lane freeway is 90% of 2100veh/lane = 1890 veh/lane. However, this figure may
vary between locations depending on the types of vehicles and other operational constraints.
In addition, the phase diagrams below (Fig 8.2x) demonstrate the boundary between the traffic states
discussed in Sections 7.5.4 and 7.5.5 and confirms the statement above that the critical control figures to
ensure stable flows occur below the maximum theoretical capacity
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
9
Fig 8.2x (From Treiber and Kesting – Traffic Flow Dynamics – Data, Models and Simulation)
The speed-volume-density diagrams above (Fig 8.2x) and those shown on Figure 8.6, Page 80 clearly show
that the recovery flow typically occurs around Level of Service C or typically 1300 - 1400 vehicles per lane per
hour, which is worth noting so operational practitioners can build algorithms to recover flow in real time etc.
8.2 Motorway Operational Capacity
8.2.1 Capacity of Segment
Capacity of a road segment, as determined for design purposes, is the maximum sustainable hourly rate at
which persons or vehicles can reasonably be expected to traverse a point or uniform section of a lane or
roadway during a given time period under the prevailing roadway, environmental, traffic and control conditions
(TRB 2010).
For motorways, capacity could be expressed in passenger car equivalents (PCE) across all lanes.
The concept of PCE is related to traffic behaviour due to the vehicle mix (i.e. presence of heavyvehicles) in
the traffic flow. These factors include:
•
physical space taken up by a large vehicle
•
longer and more frequent gaps in front and behind heavy vehicles
•
speed of vehicles in adjacent lanes and their spacing.
In motorway capacity analysis, heavy vehicles are converted into an equivalent number of
passenger cars to achieve a consistent measure of flow.
In measuring the capacity, it is generally the maximum sustained 15 min flow rate, expressed in passenger
cars per hour per lane (pc/h/ln), that can be accommodated by a uniform motorway segment under prevailing
traffic and roadway conditions in one direction of flow. The flow rate measured over a short period is generally
not sustained over a longer period. The ratio of maximum hourly volume to the maximum 15 minute flow rate
expanded to an hourly volume is thepeak hour factor (PHF). The PHF is a measure of traffic demand
fluctuation within the peak hour and is typically up to 0.95 in high flow conditions.
Above two lanes, the capacity per lane drops with each additional lane added to a motorway. This
phenomenon may be because the number of lane-changing conflict points increases with each additional lane
(refer Figure below). An additional factor may be that carriageways with four or more lanes have a greater mix
of passive and aggressive drivers in their middle lanes resulting in greater uncertainty and friction within these
lanes. Austroads Guide to Traffic Management Part 3, Section 4.1 (Austroads 2013a) provides further details
on capacity of uninterrupted flow facilities.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
10
Operational capacity is the actual real-time capacity for a road segment, which can vary depending on
prevailing roadway, traffic and control conditions. These variable conditions include the percentage of heavy
vehicles, driver population (passive or aggressive driving, familiar or unfamiliar with road), road geometry,
road surface, time-of-day, weather and light. (Theoretical capacity for a road segment is an average capacity
estimate over a period.)
Operational capacity, which can be either measured in total vehicles per hour or passenger car equivalents
per hour, is particularly relevant to the control of managed motorways. For example, ramp signals maintain
operational capacity by managing occupancy at bottleneck locations while regulating inflow demand to
prevent flow breakdown.
The German Manual for the Design of Road Traffic Facilities (RSRTA 2005) estimates mainline capacities for
freeways with speed limits of 100 and 80 km/h on grades up to 2% as shown in Table 8.1.
Table 8.1: German freeway mainline capacities for two and three lane carriageways
Where four lanes and above are proposed, further reduction in operational capacity (lane flows) occur.
Observation on Melbourne’s motorway network suggests that a reduction in the order of 100 vehicles per lane
for every lane that is added i.e for unmanaged 4 lane, may equal 1700 per lane and 5 lane = 1600 hundred
vehicles per lane and 6 lane 1500 veh/lane etc. This is directly related to the conflict points and the
turbulence/friction effects associated with getting traffic to and from the inside lanes. See conflict diagrams
above. These figures do not reflect the existence of significant weaving movements/lane changing and their
associated impacts on lane flows.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
11
Operational capacity is likely to be further reduced in locations where there are significant weaving
movements/lane changes due to overlapping Origin-Destination patterns between two or more interchanges.
This is particularly relevant on urban freeways with closely spaced ramps with high demands (eg. closely
spaced service and system interchanges).
When designing motorway projects or upgrading existing motorways, operational capacity values should be
used rather than theoretical values to gain an appropriate understanding of how the project will perform after
construction and to ensure that adequate infrastructure is provided for the anticipated demands.
Based on VicRoads (2013) research, operational investigations and further observations, appropriate
maximum capacity values for motorway design are in the table below:
Number of Lanes
2
3
4
5
6
Unmanaged (PCUs)
1800
1700
1600
1500
1400
Managed (PCUs)
2100
2000
1900
1800
1700
Note: General information considered applicable for motorway design purposes.
The above values may need to be adjusted for site-specific conditions which will impact motorway capacity,
including road characteristics and vehicle mix. The ramp merges are usually the critical capacity sections
which determine a motorway’s maximum operational capacity.
The UK Design Manual for Roads and Bridges, Volume 6, Section 2, Part 1, TD 22/6 (The
Highways Agency 2006) provides guidelines for the selection of entry ramp layouts for motorways of varying
mainline and merging flows. The guidelines include a chart for the selection of an appropriate entry ramp
layout. In regard to traffic flows (Chapter 3) it indicates:
For the purpose of designing grade-separated junctions and interchanges, the maximum flow per lane for
motorways must be taken as 1800 vehicles per hour (vph). These flows do not represent the maximum hourly
throughputs but flows greater than these will usually be associated with decreasing levels of service and
safety.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
12
Document Information
Title:
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2:
Traffic Theory (2015)
Department:
Network Standards
Directorate:
Policy and Programs
Approved by:
Jeremy Burdan
Manager – Network Standards
Date of Approval:
October 2015
Amendment Record
Edition / Revision
Pages(s)
Issue Date
Amendment Description
Austroads Guide to
Traffic Management,
Part 2: Traffic Theory
(2015)
All
October 2015
First Edition
Previous versions of this document are available on request by contacting the VicRoads – Network
Standards team.
For enquiries regarding this supplement, please contact the VicRoads – Network Standards team
via [email protected] or 9854 2417.
VicRoads Supplement to Austroads Guide to Traffic Management, Part 2 (2015) – Edition 1
October 2015
13