Design and Construction Planning of the

Macleay River and Floodplain Bridge

(Australia’s Longest Bridge)

Giammaria Gentile1 and Joseph Canceri2

1
Abigroup Contractors Pty Ltd , Australia
2
Roads &Traffic Authority of NSW

Abstract The Macleay River and Floodplain Bridge forms part of the overall
Kempsey to Eungai Pacific Highway Upgrade Project and when completed will be
the longest bridge in Australia, with an overall length of 3.2km, extending over the
Frogmore Floodplain and the Macleay River. The bridge consists of a single
structure, 21.6m wide, providing two carriageways separated by a median barrier
with two 3.5m wide traffic lanes in each direction, plus an outer shoulder width of
2.5m and inner shoulder width of 1.0m. This paper will cover the following
aspects of the project: (a) Description of the site and the final structural and
geotechnical design; (b) The development by Abigroup and its designers (Arup
and Benaim) of the design of the bridge, from the tendered design through to the
detailed design; (c) Key aspects of the design including geotechnical, flood
modelling, scour and drainage design; and (d) The challenges for the construction
including the installation of 328 steel driven piles and 64 bored piles and the
logistics involved in the manufacture, delivery and installation of 941 Super-T
girders. The project is currently in the detailed design phase, with construction
having commenced in July 2011, with a completion target of December 2012.

Introduction

This paper explores the design and construction planning of the Macleay River
and Floodplain Bridge (MRFB). Abigroup was the successful tenderer to the
Principal, the Roads and Traffic Authority of NSW (RTA), for the design and
construction of the MRFB. Although part of the Kempsey Bypass Upgrade, the
bridge over the Macleay River and Floodplain forms a separate Design and
Construct project. The contract was awarded to Abigroup in December 2010 and

V. Ponnampalam, H. Madrio and E. Ancich 258

Sustainable Bridges: The Thread of Society
AP-G90/11_011© ABC 2011
Design and Construction Planning of the Macleay River and Floodplain Bridge 259
(Australia’s Longest Bridge)

permanent works commenced in July 2011. The bridge is being constructed

concurrently with the Kempsey Bypass works (by others) to the south.
With an overall length of 3.2km, the bridge when constructed will be the longest
in Australia, extending over the Frogmore Floodplain and crossing the Macleay
River east of the town of Frederickton. Comprising a single structure
approximately 21.6m wide, the bridge will provide two 3.5m wide traffic lanes in
each direction, plus an outer shoulder width of 2.5m and inner shoulder width of
1.0m.

Parties in the MRFB project include Abigroup Contractors, Benaim, Arup and
HBO+EMTB on the design and construct team. The RTA provided design
reviews to ensure that the bridge design conformed with the RTA’s Scope of
Works and Technical Criteria (SWTC). As the Principal, the RTA developed the
route and concept design and the RTA’s advisors include APP/Hyder Consulting
as the Project Verifier and SKM as the Proof Engineer.

Site Description

The project is situated in the local government area of Kempsey and lies on the
floodplain of the Macleay River Valley. It is approximately 400 kilometres north
of Sydney and 500 kilometres south of Brisbane. The site is located
approximately 9km from the nearest point of coastline between Hat Head and
Crescent Head. The structure is therefore classified as being in a “near-coastal”
region (1km to 50km from coastline).

The landscape is characterised by the floodplain that has been largely cleared of
natural vegetation and is used predominantly for farming, particularly grazing and
contains isolated and scattered mature trees.

The raised profile of the highway alignment and the bridge structure will provide
motorists with expansive views over flat agricultural land and low lying pastures.
The Macleay River and Floodplain Bridge crossing the floodplain will introduce
an impressive new structure into an otherwise flat, natural, agricultural landscape.
The bridge structure has been sensitively designed to sit comfortably within this
context and its high visibility is moderated by clean, horizontal, structural lines
and revegetated abutment landforms.
260 Giammaria Gentile and Joseph Canceri

Bridge Description

The design for the Macleay River and Floodplain Bridge consists of precast
concrete Super-T girders composite with a 21.6m wide in-situ deck supported on
tapered concrete headstocks, twin concrete columns with rounded ends and
concrete pile caps and piles. The 94 spans have a total length of 3.2km and this
will make it as the longest bridge in Australia

The Super-T girders are all 1500mm deep. Between Abutment A and Pier 27 the
girder span is 33.865m. Between Pier 28 and Abutment B it is 34.100m, with the
exception of the span over the Navigation Channel in the river between Piers 85
and 86, where the span is 35.000m, to accommodate the 20m wide navigation
requirements. In general there are 10 girders across the width of the bridge. At the
Navigation Channel between Piers 85 and 86 there are 11 girders, to achieve the
greater span.

The spans are simply supported at each pier with link slabs connecting individual
spans between sawtooth and finger-plate movement joints. These are typically
located every five spans for the section across the floodplain, while from Pier 82
to Abutment B there are no intermediate joints for the 410m river crossing section.

The piers consist of portal frame type piers with cantilevering headstocks. The
pier columns are blade shaped with rounded ends. The pier columns for Piers 1 to
81 across the floodplain are 1.000m x 3.000m and the columns for Piers 82 to 93,
at the Macleay River are 1.300m x 3.000m. Pier heights across the floodplain
typically vary between 3m and 5m, depending on the ground profile.

Fig. 1. View of Macleay River and Floodplain Bridge crossing the northern bank of the Macleay
River, looking east (Artist Impression).

The pier headstocks are rectangular in plan, have vertical sides at the top and then
taper down to the same width as the pier columns that support them. The
cantilevering headstocks which support the edge Super-T girders taper down to
cleanly meet the top of the blade columns of the piers. All headstocks are 21.750m
long. Over the floodplain (Piers 1 to 81) the headstocks are 2.200m wide, tapering
Design and Construction Planning of the Macleay River and Floodplain Bridge 261
(Australia’s Longest Bridge)

down to 1.000m and over the river (Piers 82 to 93) they are 2.500m wide tapering
down to 1.300m.

The foundation at Abutment A comprises four steel driven tubular piles with a
composite reinforced concrete plug for the top 15m of the piles which is
connected to the abutment headstock.
The foundation from Pier 1 to Pier 81 comprises two steel driven tubular piles
connected directly into each of the two pier columns. The foundation from Pier 82
to Pier 93 consists of four or six bored piles and a single pilecap supporting the
two columns at each pier location.

The foundation at Abutment B comprises four reinforced concrete bored piles

connected to the abutment headstock.

For the edge parapets, precast concrete elements with double steel rails have been
provided which will have a consistent, high-quality finish and colour. The outside
faces of the parapets overhang and hide the edges and underside of the bridge deck
slab. A 1.1m high concrete barrier is provided in the median to comply with the
medium performance requirements.

Fig. 2. View of Macleay River and Floodplain Bridge crossing South West Rocks Road, looking
east (Artist Impression).

A spill-through type abutment forms the termination of the bridge at either end as
it meets the proposed earth embankments.

The road alignment has been designed for 110km/hr traffic, with a horizontal
alignment that is straight for the most southern 27 spans and then has a constant
radius of 4400m for the remaining of the bridge. The vertical alignment, dictated
by the tie-ins points at the abutments and the clearances over the roads and the
river, is less than 1% throughout the entire length, so achieving deck drainage is a
challenge.
262 Giammaria Gentile and Joseph Canceri

Key Challenges

While this contract does not include works on the approaches, the intention of the
RTA to complete the bridge within 104 weeks from the execution of the contract,
together with the size of the project, has been the main complexity for the design
and construction team.
The project site is located in the Macleay River Floodplain, which is subject to
frequent inundation with the ground water level close to the surface. This has
driven the design to maximise the use of prefabricated elements such as steel piles,
precast parapets and concrete Super-T girders to minimise site work.

Due to the length of the bridge, it has been important to optimise the design and
construction methodology, simplifying the typical details, enhancing the speed of
construction, and reducing the risk profile.

Attention to significant detail included the precast parapet length, construction

tolerances during piling installation and drainage connections to the girders. Issues
such as reinforcement, pier and headstock formwork, plus selection of the
bearings and joints, yet were deemed to be critical for a successful delivery of the
project. A single not-properly-thought-through detail would be repeated over and
over, slowing down construction.

The geological profile is quite variable along the length of the bridge, with the
depth of rock and the thickness of the upper strata varying quite considerably in
the longitudinal and transverse direction. This is reflected in the selection of two
different piling methodologies in the floodplain and in the river section.

Superstructure

The key elements of the superstructure are the 1500mm deep pretensioned precast
Super-T girders used for all the 94 spans. The ready availability of precast
concrete Super-T girders from the nearby yard of Australian Precast Solutions Pty
Ltd (APS) – a fully-owned subsidiary of Abigroup – provides significant
advantages in terms of certainty of delivery and quality of the product. In order to
deliver the 941 support girders within less than one and a half years, the precast
manufacturing yard implemented an expansion that allows it to produce up to six
Super-T girders in a single day.

Casting production and the erection speeds using cranes have been streamlined by
the construction team so as to minimise double handling and storage. Being
Design and Construction Planning of the Macleay River and Floodplain Bridge 263
(Australia’s Longest Bridge)

located in a floodplain area, on site storage would increase the risk of downtime
during flood events.

The Super-T girders act compositely with a 180mm minimum thickness cast in-
situ reinforced concrete deck slab. The girders will be erected by a 250-tonne
crawler crane at an average of 10 girders per day.

Deck Drainage

The deck drainage design has represented one of the most challenging elements,
given the very flat grade of the longitudinal profile. The original tender design was
for up to 13 scuppers per span discharging into longitudinal pipes connected to
downpipes at most of the pier locations.

Instead, the final design comprises, instead, a single grate at every span collecting
the water from the deck and then discharging into a longitudinal pipe over the
Macleay River and Bull Creek, or – without a downpipe – directly to grass swales
located underneath the outlet. The grass swales have been designed to meet the
water quality treatment levels set in consultation with the RTA.

The deck drainage, made of a single grate per span – instead of several scuppers –
presents a significant advantage in terms of maintenance with fewer elements to
inspect and clean, all without compromising the drainage performance.

The longitudinal location of the grates have been designed to minimise water flow
across the expansion joint, avoid pier staining and reduce the stress on the
structural element (in particular the webs where a penetration occurs).

In two locations – near the south bank of the Macleay River and Bull Creek – the
water is collected by a pipe system and discharged into two spill basins with a
capacity of 40,000L.

Substructure

Site Investigations

The regional geology is dominated by alluvial and estuarine sediments overlying a

weathered bedrock profile.
264 Giammaria Gentile and Joseph Canceri

Extensive ground investigations were carried out before the tender and during the
detailed phase of the design. Abigroup elected to go beyond the Scope of Works
and Technical Criteria (SWTC) by performing additional tests that would deliver
more detailed information critical for the design.
Given the extension of the ground investigation and the lab tests required, the
team selected tests and technology that would maximise the productivity. The
intention of the additional tests was to:
• Fill in gaps in the existing data and provide a more detailed model;
• Provide larger and higher quality samples to minimize sampling errors; and
• Measure in-situ strength and stiffness which can be directly interpreted and
inputted into the design.
One of the key ground investigation tests performed has been the Seismic
Dilatometer Marchetti Test (SDMT). It is the combination of the standard Flat
Dilatometer (DMT) with a seismic module. Such module is a probe outfitted with
two sensors, spaced at 0.5m, for measuring the shear wave velocity Vs, used to
determine the small strain shear modulus G0.

As the SDMT provides low strain G0 and the working strain M, the choice of the
proper G-Gamma curve is more accurate as two points can be used. It also
provides at each depth two independent estimates of the liquefaction resistance.

The SDMT determines in a quick, precise and simple way important parameters
used in geotechnical design, with minimal soil disturbance. On site the blade has
been advanced in the soil by pushing the rods with a rig.

Fig. 3. SDMT Equipment

Design and Construction Planning of the Macleay River and Floodplain Bridge 265
(Australia’s Longest Bridge)

Scour Assessment

Scour is the removal of material from the bed and banks of a channel by the action
of water and is a natural phenomenon caused by flow of water over an erodible
boundary. The two flood events that have been analysed are the 2000ARI and
100ARI.

General Scour or natural scour is the term describing typical bed degradation and
lateral channel migration irrespective of any of structural elements (e.g. piers and
abutments). Local Scour is caused by the bridge piers or abutments as they interact
with the water flow. Tidal Scour can occur in a river bed due to currents induced
by frequent astronomical tidal variations. Unlike fluvial scour, tidal scour can
occur in both directions at times of low fluvial flow.

Global Scour

The findings from the geomorphologic assessment suggest that long-term general
scour under normal flow conditions (degradation or lateral bank migration) is not
anticipated - in fact the system is generally aggrading.

A review of historic maps and aerial photographs show that this reach of the
Macleay River has remained stable in relation to lateral channel migration over the
last 105 years.

General Scour was calculated using regime equations for the 2,000 year ARI. The
calculation reports up to 3.5m of bed degradation for the main channel and no
general scour for the floodplain. An active flood channel is indentified on the
southern bank of the river, potentially affecting piers P79 – P82. It is possible that
during the 2,000 year ARI flood event that this channel could deepen from scour
and therefore a 3.5m scour has been adopted in this area.

Local Scour

During tender, and up to the 15% design stage, the design team used a theoretical
approach to determine the local scour depth.

The assessment of local scour is complex and is dependent on a number of

parameters. Laboratory data is used to assess the influence of each parameter and
subsequently derived a simple equation as follows:
• ds = KybKIKdKsKOKt (Equation 1)
Where:
266 Giammaria Gentile and Joseph Canceri

• ds = estimated local scour depth (m);

• Kyb = flow depth – foundation size factor (m);
• KI = flow intensity factor;
• Kd = sediment size factor;
• Ks = foundation shape factor;
• KO = foundation alignment factor; and
• Kt = time factor
This approach could be unrealistic and overly conservative when the shape of the
foundations – piles and pilecaps – doesn’t fall within regular, commonly used
shapes. To refine the assessment further, the team engaged a world-renowned
specialist, Prof. Bruce Melville of the University of Auckland (who is an active
researcher with an international reputation in the field of fluvial sediment
transport), with the intent of carrying out physical modelling in his laboratory. His
expertise encompasses most aspects of water resources engineering, including
hydraulic, river, environmental and hydro-electric engineering.

The physical models were constructed in a flume and developed using a 1:53 scale
for the bridge pier and pile groups. The model was run with a velocity of between
0.325m/s and 0.35m/s. These velocities were deliberately chosen to be at about the
threshold velocity for sediment movement on the approach flow bed, because that
condition leads to the maximum scour.

The sediment used was fine-grained sand and therefore gave slightly conservative
results when compared to the fine and slightly cohesive material anticipated in the
top soil layers for both the river channel and floodplain. The models were run with
this constant velocity for between 24 hours and five days. The results showed that
it took approximately five days for the maximum scour depths to be generated and
where shorter tests were run the depths were adjusted to reflect the test length.

The scour depths were measured by draining the water from the flume in 20mm
increments (which represents approximately 1m of water depth) and plotting out
the bed level with string to show the bed contours and the effects of the scour.
Each of the contours in the following figures represents approximately one metre
of scour depth.
Design and Construction Planning of the Macleay River and Floodplain Bridge 267
(Australia’s Longest Bridge)

Fig. 4. Local Scour Results For Piers 82, 83, 91, 92 And 93 With Debris.

The effects of debris was simulated by running the model for an additional 24
hour period, with a large block of polystyrene fixed at the water surface to
simulate a debris raft of 21m by 8.5m by 2.65m for the river channel and a raft of
18.5m by 4.24m by 2.65m for the floodplain and river bank. Both raft sizes are
particularly large and were developed to assist in the prediction of conservative
scour depth.

Fig. 5. Local Scour Results For Piers 84 To Pier 90 With Debris.

Tidal Scour

The assessment considered scour from tidal currents only, with and without the
presence of the proposed bridge piers, assuming no residual fluvial flow (‘dry’
conditions). Tidal velocities of around 0.2m/s were calculated, which are very low
in comparison with fluvial currents during extreme flood events (around 10%).

General scour potential (without the proposed bridge) has been estimated using
the Lacey (1930, 1933) and Blench (1969) prediction methods that suggest general
268 Giammaria Gentile and Joseph Canceri

aggradation would occur from tidal currents at this location of the river. Local
scour potential (with the proposed bridge in place) has been assessed using the
Austroads (1994) method for predicting scour in the presence of bridge piers.
Calculations suggest that local scour would be limited to less than about one metre
in depth along the upstream and downstream projections of the piers.

Floodplain Foundations

The tender design (resubmitted at the 15% submission) was for four 1200mm
diameter steel piles at every pier location, in a single row alignment. The
connection between the top of the 1200mm diameter piles and the pilecap was to
be achieved by welding the reinforcement to the inside of the tube and casting a
700mm concrete plug in the pile, to embed the full length of the welded
connection.

The design then progressed to the 85% stage and the revised foundation evolved
to 750mm and 825mm driven steel piles, connected directly to the pier columns
with a reinforced concrete plug (12-15m long), but without the use of pilecaps.

Fig. 6. Revised Foundation, Without The Use of Pilecaps.

In the floodplain there will be 328 driven steel tube piles of 16mm thickness, 750
– 825mm diameter and varying from 20 – 47m long, driven to nominal refusal
into “sound and unyielding rock” in accordance with the SWTC. As mentioned
previously, there will be two pier columns per headstock and two piles per pier
column, with these columns varying in height from approximately 4 – 6m. The
upper 12 – 15m of the piles will be filled with reinforced concrete for connection
Design and Construction Planning of the Macleay River and Floodplain Bridge 269
(Australia’s Longest Bridge)

and load transfer from the columns. Abigroup is self-performing the piling works
in the floodplain.
As the floodplain’s water table is very close to ground level, during the design
development phase it was determined that the pilecaps from P1 to P81 could be
eliminated by connecting the piles directly to the reinforced concrete columns.
This decision was made to speed up construction, as excavations for pile caps
would have had to be pumped out continuously and disposal of the excavated acid
sulphate soils would have added further time and costs.

While there were efficiencies to be gained from the deletion of pile caps, this will
require additional precision on site, with much tighter construction tolerances for
the 750mm and 825mm diameter piles to align within the one metre wide columns
than would conventionally be the case with pile caps. To achieve the very tight
tolerances for the location of the driven steel piles a specially designed pile frame
was fabricated as shown in the following figure.

Fig. 7. Crane and Piling Frame.

The vertical geotechnical capacity of the piles is achieved by a combination of end

bearing and shaft friction. The lateral capacity of the piles is governed by the
structural element as they behave as long piles.

River Foundations

The river piles are much larger because of the less favourable geology and
massive scour effects of a 1:100 year flood. Following the results of the site
investigation and a more refined finite element modelling, the tender design has
270 Giammaria Gentile and Joseph Canceri

been developed to the final stage that comprises four 1800mm bored piles with
pilecaps for the piers on the river banks and six 1800mm diameter bored concrete
piles with pilecaps for the piers in the water. Installation of the bored piles for the
river section will be performed from temporary rock platforms, across the river. It
will be done with a combination of bentonite and temporary steel casings driven
down into the river bed, followed by the construction of the rock socket. After
cleaning of the pile base the reinforcement cages will be installed. Concrete
placement will be achieved by tremmie pouring. Temporary casing extraction can
commence once sound concrete has been achieved to the desired level.

Fig. 8. River Foundations.

Acknowledgments
The Authors wish to acknowledge the Roads and Traffic Authority of New South
Wales for permission to publish this paper.The Authors also wish to acknowledge
the contribution from Arup, Benaim, and HBO+EMTB.

References

[1] Abigroup, Arup, Benaim. Design Report for Macleay River and Floodplain Bridge for
Kempsey Bypass
[2] Abigroup, HBO+EMTB. Urban Design and Landscape Management Plan for Macleay
River and Floodplain Bridge for Kempsey Bypass
[3] Abigroup, Arup, Auckland University. Scour Assessment Report for Macleay River
and Floodplain Bridge for Kempsey Bypass
[4] Marchetti S, Flat Dilatometer and Seismic Dilatometer. www.marchetti-dmt.it