NCD - Natura Channel Design

6945 Nat Channel/Guide 4/5/01 2:43 PM Page 2
GUIDELINES
November 2003
Natural Channel Design
Dedicated to a better Brisbane

Document Control
Issue A December 2000 Original

Issue B November 2003 Modification to long riffle rock size formula on page 49
Natural Channel Design

Guidelines
November 2003
Brisbane City Council

GPO Box 1434
Brisbane
Queensland
Document was produced in co-operation with Grant Witheridge, Catchment & Creeks Pty Ltd.
ISBN 187609 141X

Natural Channel Design Guidelines

Contents
Executive Summary ..........................................................1 Step 9...............................................................................

Review riparian and floodplain vegetation..................41
Section 1 .........................................................Introduction Step 10.............................................................................
Design of channel bed and low flow channel .............42
Purpose of guidelines ............................................................3
Step 10-1(a)......................................................................
Use of guidelines ...................................................................5 Bed design for channel with little or no
How to the guidelines assist me............................................6 bed vegetation ............................................................42
Natural Channel Design (NCD)..............................................7 Step 10-1(b) .....................................................................
Bed design for channels with bed vegetation
Background............................................................................9
determine bed form ....................................................43
Symbols ...............................................................................10
Step 10-2(b) .....................................................................
Determine the length of the low flow channel............43
Section 2 .................................................Planning Aspects Step 10-3(b) .....................................................................
Introduction .........................................................................12 Determine the number of riffles ..................................44
The need for channel works ................................................13 Step 10-4(b) .....................................................................
Determine typical riffle dimensions .............................45
Selecting the extent of works ..............................................14
Step 10-5(b) .....................................................................
Potential impacts of works...................................................14 Determine typical pool dimensions.............................47
Planning issues.....................................................................14 Step 10-6(b) .....................................................................
Plan form..............................................................................17 Determine pool length ................................................48
Corridor widths ....................................................................18 Step 10-7(b) .....................................................................
Site constraints.....................................................................19 Determine riffle rock size .............................................49
Flood relief works ................................................................22 Step 10-8(b) .....................................................................

Design low flow channel..............................................50
Step 11.............................................................................
Section 3 ..............................................Design Procedures Incorporate channel features .......................................54
Introduction .........................................................................24
Data collection.....................................................................25 Section 4 .......................................................Revegetation
Design steps ........................................................................25 Introduction .........................................................................55
Symbols ...............................................................................26 Interaction with Natural Channel Design ............................56
Step 1............................................................................... Significants of vegetation type ............................................56
Select the reach of watercourse to be rehabilitated ...27
Channel vegetation .............................................................58
Step 2...............................................................................
Determine channel fall.................................................28 Floodplain vegetation..........................................................59
Step 3............................................................................... Planting patterns and hydraulic constraints.........................60

Determine bankfull flow rate .......................................29 Group plantings ...................................................................61
Step 4............................................................................... Planting around major hydraulic structures .........................62
Select the type of channel ...........................................29 Species for revegetation in Brisbane...................................64
Step 5............................................................................... Plants used for erosion control ............................................65
Determine channel width and channel depth .............31
Woody vegetation ...............................................................68
Step 6...............................................................................
Channel bank slope .....................................................33
Step 7............................................................................... Section 5 ..................................Evaluation and Monitoring
Determine typical channel meander radius.................35 Introduction .........................................................................70
Step 8............................................................................... Planning the evaluation .......................................................70
Select trial channel location.........................................36
Typical rehabilitation activities .............................................77
Summary ..............................................................................81
Contents
Contents
Riffles..................................................................................115
Section 6 ........................................................Bibliography Riparian vegetation............................................................117
Bibliography.........................................................................82 Rocks..................................................................................118
Rock ribs ............................................................................119
Appendix A.............................................Channel Features Sediment............................................................................119
Introduction .........................................................................85 Shading and cover.............................................................120
Aquatic wildlife (ecosystems)...............................................89 Shelter................................................................................121
Aquatic movement corridors ...............................................90 Sinuosity.............................................................................121
Aquatic plants ......................................................................91 Snags .................................................................................122
Bankfull flow rate .................................................................91 Steep, eroded bare banks .................................................123
Banks....................................................................................91 Stepping stones.................................................................123
Base flow..............................................................................91 Stormwater outlets ............................................................124
Berms ...................................................................................93 Substrate............................................................................125
Boulders...............................................................................93 Terrestrial movement corridors ..........................................125
Bridges.................................................................................94 Undercut banks..................................................................126
Buffer zones .........................................................................94 Water quality......................................................................126
Causeways ...........................................................................95 Water temperature.............................................................126
Channel banks .....................................................................95 Waterway crossings............................................................126
Channel beds.......................................................................99 Weirs ..................................................................................127
Channel meanders...............................................................99 Wildlife corridors................................................................127
Chutes................................................................................100 Woody debris ....................................................................127
Compound channels..........................................................100
Culverts ..............................................................................101 Appendix B ........................................Hydraulic Structures
Deep pools ........................................................................103 Introduction .......................................................................128
Deflectors...........................................................................103 Migration of aquatic wildlife ..............................................128
Drop structures ..................................................................106 Arch structures ............................................................128,133
Ephemeral creeks ..............................................................106 Bridges........................................................................129,133
Flows ..................................................................................107 Chutes.........................................................................129,133
Grouted rock......................................................................108 Culverts .......................................................................130,133
Habitat ...............................................................................108 Drop structures ..................................................................132
Leaf litter............................................................................109 Weirs ..................................................................................132
Log weirs............................................................................109 Migration of terrestrial wildlife...........................................132
Logs ...................................................................................109
Low flow channels..............................................................111 Appendix C ......................................Mannings Roughness
Overhanging vegetation....................................................111 Introduction .......................................................................138
Pipe crossings ....................................................................111 Strickler equation...............................................................138
Plan form geometry ...........................................................111 Cowan method ..................................................................138
Pools ..................................................................................111 Limerinos equation for sandy channels .............................139
Pool-riffle systems ..............................................................112 Flow depth considerations ................................................139
Recessed banks .................................................................114 Mannings n photos ...........................................................140
Executive Summary
Natural Channel Design (NCD) is important for all waterways, but especially important
where the waterway corridor links with a bushland reserve or forms an important aspect of
an aquatic or terrestrial movement corridor.
The principle of NCD is based on providing the required hydraulic conveyance of a

drainage channel and floodway while maximising its potential environmental values. This
holistic approach combines the disciplines of hydraulic engineering, fluvial
geomorphology and in-stream and riparian ecology. NCD also encompasses non-
scientific/engineering principles such as community requirements.
Unfortunately, we cannot have a totally natural creek in an unnatural catchment, but by

incorporating the principles of NCD we can rehabilitate a creek to look and behave in a
manner similar to a natural system.
NCD principles are based on four levels of investigation and design, firstly the regional
level of catchment management, secondly the local plan form of the watercourse, thirdly
the channel cross-section, and finally the design of in-channel features. Ultimately, it is the
in-channel features that provide the life blood of Natural Channel Design; however,
without the appropriate consideration of the other three levels of investigation and
design, the cost and effort placed into NCD are usually wasted.
It is important for planners and designers to recognise that the principles of NCD cannot
work in isolation from the rest of the catchment. NCD is not just the ecological or
hydraulic design of a channel. It is an urban drainage philosophy that must integrate a
wide range of principles and ideals into a functional, affordable and aesthetic outcome.
Natural Channel Design not only has to integrate well with the catchment hydrology and
ecology, but must also integrate well with the local community and the values the
community places on their waterways.
Chapter 2 reviews many of the planning aspects of NCD and how to integrate the channel
with external features such as accessibility, aesthetics, inter-catchment bushland corridor
links and flood control. Site constraints are also reviewed, including, steep banks, existing
trees and public assets, stormwater outlets, soil properties and in-stream sediment flow.
Chapter 3 presents a channel design procedure. This chapter concentrates on the

hydraulic aspects of Natural Channel Design. The formulas presented for sizing the
various features of a channel should be considered as a first trial. In the end, the channel
1
Executive Summary
Executive Summary
design must reflect the presented design principles, but must also reflect the random
variability of these channel dimensions as found in most natural channels.
Chapter 4 discusses the revegetation of channels and floodplains. Guidelines are also
presented on the selection of plant species for floodsensitive areas of Brisbane.
Chapter 5 provides step by step proceedures to evaluate NCD Projects. Monitoring

followed by Evaluation (Assessment) is a key to the success and on-going improvement of
the NCD principles and methods.
Appendix A of the document discusses the various in-channel features that are likely to be
associated with a NCD. Some of the most important or common features include, aquatic
movement corridors, in-stream boulders, channel meanders, deep pools, habitat areas,
logs, overhanging vegetation, pool and riffle systems, shade, snags and terrestrial
movement corridors.
Other Appendices deal with selection of hydraulic roughness and the design of
environmentally sensitive hydraulic structures.
As can be seen from the above description, these Guidelines are technically based and
have been developed for use by professional design teams. The Guidelines concentrate
on planning and design aspects. If the reader requires information on the basic
philosophy of NCD and why NCD is important for urban waterways, then reference should
be made to the various documents provided within the bibliography.
Finally, creeks should be regarded as corridors of opportunity that should not be wasted.
These Guidelines aim to maximise the benefits that our waterways can provide to the
human and wildlife inhabitants of our urban areas.
2
1.Introduction
Background Design of typical conveyance systems such as engineered channels and the rehabilitation
techniques applied to natural waterways have been the subjects of extensive research in
the last decade for their improvement and consistency with the ecological sustainability
and protecting environmental values.
One of the key conveyance systems that forms an integral component of many urban
development proposals is the construction of grass-lined, concrete invert trapezoidal
channel systems. These channel systems are designed to cater for either the runoff
generated from the development site or to compensate for the conveyance requirement
that might have been lost due to piping or modification of the natural watercourse. These
channel systems not only destroy the ecosystem, but also are contributing factors to
transport of significant volumes of sediments. Hydraulically, they may be very efficient, but in
terms of sediment retention for improved water quality and ecological health these channels
are ineffective. As part of its commitment to continually improve the ecological values of its
waterways, in 1996, Brisbane City Council prepared a guideline titled, Hydraulic Geometry of
Brisbane Streams Guidelines for Natural Channel Design. The 1996 Guideline has been
reviewed during the last four years for its suitability and application to urban areas, particularly
Brisbane. The results of this review, along with relevant planning and design tools including
many other features of natural conveyance systems, are included in the current Guideline.
The Natural Channel Design (NCD) Guidelines are aimed at practitioners to design and
construct major drainage channels and also to rehabilitate degraded waterways to
incorporate, improve or conserve natural and/or desirable social, ecological and drainage
features.
Brisbane City Councils Brisbane City Council recognised the need for the use of this Guideline to achieve better
Strategic Link environmental outcomes for the citys existing and constructed waterways. The Guidelines
are consistent with the principles of ecologically sustainable development (ESD). The City
Plan (October 2000) Stormwater Management Code and Waterway Code emphasise the
use of this Guideline. Apart from new developments, the Guideline is also to be used in
all Council and community projects that involve waterway restoration, erosion control
works and habitat enhancement. Figure 1 shows the relationship between NCD and the
desired environmental outcomes through relevant planning and legislative links.
Objectives The primary focus of the Guidelines is not to re-create an existing conveyance system into
a natural-looking waterway. In an urban environment most of the waterways are heavily
modified and attempts to re-create/transform them into an undisturbed/pristine state are
expensive and time-consuming with limited success. The objective of the Guideline is to
assist practitioners through following steps in undertaking waterway rehabilitation works:
3
Introduction
Choosing the appropriate site and deciding the extent of works

Setting the environmental and social goals based on desired environmental values.
Planning the project
Undertaking step-by-step design using the appropriate tools to address all aspects of
natural features and desired goals
Considering critical factors during and after construction
Evaluating and monitoring.
Using the Guidelines, practitioners are expected to construct and rehabilitate creeks and
drainage channels to achieve natural and/or desirable social and ecological outcomes
while minimising the impacts of flooding by fulfilling the drainage requirements. The
Guidelines aim to balance all aspects of engineering and environmental issues. The
Guidelines also promote ecologically sustainable development that incorporates naturally
functioning waterway systems.
The NCD principles and techniques outlined in the Guideline are primarily applicable for
urban creeks and constructed drainage channels. Many of the concepts cannot be applied
directly to river systems without appropriate modifications. For guidelines on the
rehabilitation of the river systems, refer to Land and Water Resources Research and
Development Corporations (LWRRDC) document A Rehabilitation Manual for Australian
Streams, 2000.
Mission statement:
To minimise erosion, flooding and maintenance while maximising

ecological and aesthetic values of waterways
4
How do the guidelines assist me

SITUATION TYPICAL SPECIFIC
APPLICATIONS OUTCOMES
Flood mitigation Self-sustaining systems

Waterway restoration which meet Councils
COUNCIL
commitment to
PROJECTS Erosion control works
protecting the built and
Habitat enhancement natural environment.
Guidelines provide a
reference to assist
COMMUNITY Habitat enhancement experienced individuals
in the community to
PROJECTS Waterway Restoration
produce sustainable
ecological and drainage
outcomes.
DEVELOPMENT DEVELOPMENTS THAT: Guidelines are a referenced document

UNDER under the CITY PLAN and assist developers
CITY PLAN Propose works within a in providing acceptable solutions to the
waterway corridor defined following performance criteria:
under City Plan WATERWAY CODE
CODE OR IMPACT ASSESSMENT
Propose drainage works General criteria
REQUIRING CONSIDERATION OF
involving construction of *P1 The natural functions of Waterways as
THE WATERWAYS AND
new channels landscape and environmental
STORMWATER MANAGEMENT
CODES propose restoration of corridors must be preserved.
existing channels or *P4 Where a site includes degraded land
WATERWAYS CODE waterways as part of identified for rehabilitation in a
Purpose .protect and enhance meeting development Catchment Management Plan, SMP
the water flow, water quality, standards WMP or a rehabilitation plan
ecology and open space and
approved by Council, it must be
recreation and amenity values of
implemented and maintained at the
the Citys Waterways and Brisbane
River and their corridors, in an landowners expense.
ecologically. sustainable way.. STORMWATER MGT CODE
*P1 Maximise use of natural channel
STORMWATER MGT CODE design principles and water sensitive
Purpose .ensure that the design of urban design principles
channel works as part of development *P8 Any channel works that are part of
maximises the use of natural channel the development or major drainage
design principles where possible to system or flood mitigation works must
establish (for new channels) or maintain and/or enhance the
enhance (for existing waterways) environmental values of the waterway
waterway corridors. corridor or drainage corridor.
HIGH LEVEL OUTCOMES

CONVEYANCE SYSTEM CONTROL TO MINIMISE STORMWATER POLLUTION AND ENHANCED WATERWAY VALUES
UNDER BRISBANE CITY COUNCILS URBAN STORMWATER MANAGEMENT STRATEGY
CONTRIBUTE TO ECOLOGICALLY SUSTAINABLE DEVELOPMENT UNDER THE INTEGRATED PLANNING ACT
MEET CORPORATE AND INDIVIDUAL DUTY OF CARE UNDER THE ENVIRONMENTAL PROTECTION ACT
FIGURE 1
*P = Performance Critera. SMP = Stormwater Management Plan. WMP = Waterway Management Plan

5
Introduction
Use of guidelines Natural Channel Design (NCD) is important for all waterways, but especially important
where the waterway provides a link with bushland reserves or forms an important part of
an aquatic or terrestrial movement corridor. A city-wide network of natural channels
whether natural in formation, or constructed to appear and operate as natural channelsis
essential for the protection of both terrestrial and aquatic biodiversity.
There are no universally agreed-upon design guidelines/techniques that can be applied to

all situations. Design relationships presented in these guidelines are essentially empirical
functions based on survey data. Thus, the relationships presented here should be used
with extreme caution when applied outside the original survey area.
However, many of the relationships presented relate strongly to universally accepted

hydraulic principles and therefore should be applicable around the world, while others will
only apply to the immediate Brisbane or similar urban areas from where the data was
obtained.
Therefore, these guidelines are generally suggested for use by professionals experienced
in open channel hydraulics and creek engineering.
The guidelines are not intended to provide sufficient information to allow any person to
design creek rehabilitation works. As a minimum, the design team should consist of an
ecologist, hydraulics engineer, local vegetation expert, and in some cases, a soils expert.
An ecologist is required to effectively incorporate the channel features described in

Appendix A. An engineer is required to integrate the channel into floodplain and to
review the channel sizing calculations presented in Chapter 3. The vegetation expert is
required to select vegetation that best satisfies the habitat requirements of the ecologist
and the flood and velocity control requirements presented by the engineer (Chapter 4).
A soils expert is often required because many watercourses flow though areas of unstable
soils. In Brisbane, the creeks often have a layer of stable clayey soil overlying highly
erosive sandy soils. As a result, channel rehabilitation works can initiate severe erosion
problems.
From a hydraulic point of view, NCD concepts mainly deals with two aspects of a channel:
(i) channel geometry; and

(ii) the low flow pool-riffle system.
The channel geometry equations can be used in a wide variety of situations. However,
pool-riffle systems can only be used when there is a regular low flow and there are only
small quantities of coarse-sediment flowing in the watercourse (i.e. the relationships are
generally not suitable for sandy creeks and rivers).
It is also noted that outside the Brisbane region, a creek system may require different low
flow systems depending on the local hydrology, sediment load, canopy cover and known
aquatic species.
6
The concepts presented in these guidelines can generally be used in the following
situations:
The enhancement of open, grassed, overland flow paths during the development of
rural lands into urban areas.
The determination of minimum corridor widths during catchment and development

planning.
The design of flood mitigation works that involve relocation, widening or deepening of
a watercourse.
The planning and/or management of floodplain vegetation for the control of flooding.
The rehabilitation or reconstruction of a channelised or otherwise degraded

watercourse.
Photo 1. Rehabilitated flood channel in Denver, Colorado
Natural Channel The basic principles of NCD are to maintain the hydraulic conveyance requirements of
Design (NCD) engineered or affected channels, while improving environmental values. This holistic
approach combines the disciplines of hydraulic engineering, fluvial geomorphology and
in-stream and riparian ecology. NCD also encompasses non-scientific/engineering
principles such as community requirements. (Brisbane City Council, 1996)
The value of a healthy ecosystem to the social and physical well-being of humans and
their environment should not be underestimated. Therefore, it is important for a drainage
system to not only continue to provide protection from flood and erosion hazards, but to
also address social and environmental concerns such as:
7
Introduction
Preservation of an ecosystems integrity by conserving or promoting diverse

communities and species and the processes that support them.
Development of a self-sustaining system to minimise long-term maintenance costs and

any adverse downstream impacts.
Design and/or incorporation of attractive natural features to improve and integrate the
landscape of the watercourse.
The principles of NCD do not aim to produce an oasis of nature in the middle of a city,
but to retain, restore and/or rehabilitate the natural features of a watercourse that will be
compatible with the greater environment in which it is located.
Photo 2. Habitat features added to a concrete drainage channel, Brisbane
Design approach NCD principles can be applied to a watercourse in one of two ways:
(i) the Field of Dreams Method; and

(ii) the Integration Method.
The Field of Dreams Method is based on the principle: "Build it and they will come" (as
in the Kevin Cosner movie "Field of Dreams"). In this method, desirable channel features
are built into a watercourse for the ecosystem that either historically existed in the area, or
that the community currently desires. Upon building the channel, you wait and see what
animals inhabit the watercourse and how it sustains floods and other urban influences.
Generally this method is not desirable, but in some circumstances it may be the only
option. At worst, this approach should result in the formation of a channel that could be
8
modified in the future to a more appropriate system as further knowledge of the local
environment is obtained.
The Integrated Method is based on establishing ecological features that are likely to be
both sustainable and ecologically desirable for the environmental conditions that exist or
will exist within the channel in both the short and long-term. Some of the environmental
conditions that need to be considered include: catchment geology, hydrology, corridor
links, water quality and vegetation.
Whichever method is used, it is important not to concentrate on pure aesthetics, but to

build something that nature would have built given the modified geology and hydrology
of the catchment. Generally it is not advisable to try to entice platypus, frogs and other
creatures into environments they would not normally inhabit.
Unfortunately, we cannot have a natural creek in an unnatural catchment, but by

incorporating the principles of NCD we can have a creek that looks and behaves very
similar to a natural system.
Regime Theory "The relations between width, depth and discharge are collectively known as regime
theory. Regime essentially means equilibrium, where there is no net erosion or deposition,
and sediment in equals sediment out. Regime theory and hydraulic geometry are very
similar concepts. Hydraulic geometry is a broader concept and describes all aspects of
river form, of which width and depth (described by regime equations) are a part."
(Brisbane City Council, 1996)
Background In 1996 Brisbane City Council released the document "Urban Stream Rehabilitation
Principles and Guidelines". This document provided basic ecological and engineering
considerations, but did not incorporate the necessary regime equations applicable for the
Brisbane region.
Later in 1996 the first series of regime relationships for Brisbanes waterways were
published in Councils new guideline: "Hydraulic Geometry of Brisbane Streams
Guidelines for Natural Channel Design". This document presented techniques and
methods that had originally been developed for other regions of Australia and around the
world.
Following release of the 1996 guidelines Council embarked on a process of surveying a

number of selected waterway reaches and comparing the observed channel and bed
geometry with the adopted regime relationships at the time.
Large format aerial photos where used to locate examples of pool-riffle systems that had
not been artificially created nor modified. Four urbanised sites where surveyed in
Brisbane City as well as two natural systems within a catchment west of the city.
The design equations provided within these Guidelines have been developed from the
analysis of this survey data.
9
Introduction
Symbols
D [m] Channel depth from bankfull water level to channel invert
DN [m] Nominal rock diameter of which N% of the rock is smaller, i.e. 10% of the rock is smaller
than the D10 rock size. Rock size is measured as the diameter of a sphere with an
equivalent volume to the individual rock
Dp [m] Maximum pool depth
d/s [-] Downstream
Fr [m] Riffle fall measured from riffle crest to the low flow pool level within the downstream pool
Ft [m] Total channel fall across the proposed rehabilitation area equals upstream channel invert
level minus downstream channel invert level
L1 [m] The valley length measured along a smooth straight or curved path (not including the
channel meanders) from the upstream end of the channel reach to the downstream end of
the channel reach
L2 [m] Length of a meandering watercourse channel measured along the channel centreline
L3 [m] Length of the low flow channel measured over a given reach of the watercourse
Lm [m] Straight line distance between successive channel meanders, or meander inflection points
Lp [m] Length of a low flow pool measured at the low flow water level
Lr [m] Riffle length measured from the riffle crest to the upstream end of the downstream low
flow pool
N [-] Number of riffles within a reach
n [-] Mannings roughness value
q [m3/s/m] Discharge per unit width for a nominated design event
Q [m3/s] Discharge
Qf [m3/s] Bankfull discharge
Qfds [m3/s] Bankfull discharge immediately downstream of the works
Qfus [m3/s] Bankfull discharge immediately upstream of the works
Q [N] [m3/s] The peak design discharge at a given location that is expected to occur on average once
every N years, i.e. the Q100 is the expected peak discharge during the 1 in 100 year
design storm
10
Symbols
Rc [m] Channel bend radius measured to the centre of the channel, or the low flow channel
meander radius measured to the centre of the low flow channel
Ro [m] Bend radius measured to the outer bank of the channel at the elevation of bankfull flow
S [m/m] Slope
So [m/m] Average channel slope over a selected reach
Sr [m/m] Slope of downstream face of riffle
Si [-] Sinuosity = L2/L1 or L3/L2
u/s [-] Upstream
V [m/s] Average flow velocity
W [m] Channel top width measured at the height of the lower bank
Wp [m] Maximum pool width measured at low flow water level
Wr [m] Riffle bed width
[m] Average meander displacement, or half the amplitude of the meandering channel
centreline or low flow channel centreline
[deg] The angle measured between two consecutive inflection points
11
Planning Aspects
2. Planning Aspects
Introduction Even the best looking, most stable, constructed watercourse can be considered a failure if
during the planning stage an alternative solution to the problem was not identified or
considered. It is during the planning stage that options must be identified, and more
importantly, the need for the proposed works must be openly questioned.
This chapter introduces many of the issues and aspects of Natural Channel Design that
should be investigated and possibly resolved prior to commencement of the detailed
design phase.
Before setting the planning agenda on the individual items/features, the planning for the
overall project must be addressed. The following are key steps that need to be
considered:
(i) Collect existing data (ecological, aesthetic, flooding and water quality) and assess
community interest, issues and concerns.
(ii) Analyse the data and identify key constraints and potential opportunities.
(iii) Set the environmental and social values (or outcomes) in consultation with the
community and various stakeholders based on the analysis of the available
information/data, for example:
establishing a particular fauna corridor;
improving water quality;
revitalise the waterway to support colonisation of a fauna species;
revegetate the waterway for shading, weed control and other ecological support.
(iv) Select the section and extent of works.
(v) Based on the available data, set the indicators or select the key natural/engineering
features that will deliver the outcomes.
(vi) Initiate design and prepare a consultation feedback plan to ensure that the project
can deliver with its intended outcomes.
(vii) Allow for investigation of new options and revising the targets where necessary.
12
In the planning stage, consideration needs to be given to a wide range of issues and
concerns. When designing a fully reconstructed watercourse, consideration should be
given to the following:
Hydraulic requirements, including: the control of flood levels, channel velocities, and
the frequency of inundation of multi use floodplains (i.e. parks and ovals), and the
management of floodways.
Ecological requirements, including: riparian vegetation, shading, wildlife corridor

movement, variable bed geometry and particle sizes, the provision of bed, bank and
overbank habitat areas, and the establishment of a pool-riffle system where
appropriate.
Aesthetic requirements, including: natural appearance, diversity of plant species,

meandering channel and variably channel geometry (many of these also being
essential for ecological purposes).
Community issues, including: all the issues listed above as well as recreational and
educational aspects, cultural and historic issues, community consultation, ownership
and participation in the project.
The need for In planning channel works, some of the first questions to be asked and answered include:
channel works (i) What is the purpose of the works?
(ii) What are the alternatives?
(iii) What are the issues and concerns?
(iv) What are the problems that need to be solved?
(v) Is the reasoning behind the proposed works ecologically-based, or aesthetically-
based?
It is important not to confuse the desire to have a natural looking creek with the need for
ecological rehabilitation of a creek. For example, simply planting trees along a creek may
not be the most effective way of improving the local ecology, while establishing a fish-
friendly channel may not satisfy community interests.
When rehabilitating or reconstructing a watercourse there are various levels of work,

including:
(i) full reconstruction;

(ii) watercourse relocation;
(iii) reforming the channel cross section in its current location;
(iv) rehabilitating only the channel bed;
(v) adding minor features to the existing channel;
(vi) erosion control of the bed and/or banks;
(vii) revegetation.
It is important to clearly identify the level of work that is actually required in order to
satisfy the needs of the project. There is little point in fully reconstructing a creek if all it
needs is improved bed conditions and the re-establishment of riparian vegetation.
13
Planning Aspects
Sometimes it is best to let nature take its course, and sometimes it is better to lend a
helping hand. Bed erosion is a problem that usually requires a helping hand.
Do NOT try to reconstruct the original natural watercourse in a disturbed unnatural

catchment. The size and shape of a watercourse are directly related to its catchment land
use. If the catchment changes, then it is highly likely that the watercourse shape and
alignment will also change.
Selecting the extent Try to limit the total bed fall of the rehabilitated reach to avoid high flow velocity problems
of works during the revegetation stage. By doing this, flow velocities through the rehabilitated
reach will be reduced because it lies within the backwater region of the downstream
channel roughness.
In the absence of any other data or model results; if bankfull flow velocities are expected
to be high, then the length of the proposed rehabilitation works should not exceed a
channel fall equal to half the channel depth.
It is essential for the new works to be properly integrated into the existing upstream and
downstream channel to avoid transition instabilities. This applies especially to the
upstream and downstream extremes of the channel works.
Potential impacts of Consideration needs to be given to the possible impacts of the proposed works on the
works upstream and downstream channel reaches. Relocating or reconstructing a watercourse
may affect channel erosion in the following ways:
(i) bed erosion in the upstream channel due to increased flow velocities;
(ii) upstream bank erosion due to an increase in bankfull flow rate;
(iii) bank erosion in the downstream reach due to higher flow velocities leaving the new
channel;
(iv) bank erosion in the downstream channel due to the formation of a new channel
bend immediately upstream of the old channel.
Channel works, especially channel relocation, can significantly impact upon the instream
and riparian wildlife, both upstream and downstream of the works. This applies especially
during the breeding season of species such as the powerful owl, platypus and native fish.
Due consideration should be given to all potential impacts.
Planning issues A short discussion is provided below on various issues that may need to be addressed
during the planning phase.
Accessibility Watercourses generally need to be accessible to the public. However, accessibility

requirements must to take into consideration the need to protect critical habitat areas
from public interference, as well as the requirements for channel maintenance and fire
control. Accessibility requirements are discussed in Councils Environmental Best Practice
Manual for Parks.
14
Aesthetics The size, location and type of trees that surround a creek can have a dominant influence
over the aesthetics of the watercourse.
Often there are conflicts between the desire to improve the visual appearance of a
floodplain and watercourse, and the need to rehabilitate or conserve the ecological values
of the watercourse.
Adjacent land usage A number of conflicts can exist between the requirements of certain land uses and the
requirements of waterways. Examples are given below:
Golf courses golf courses can work well with creeks if they are planned around the
watercourse, but if the fairways cross the creek, then there is a conflict between the need
for a riparian zone and the need to keep trees off fairways.
Parkland waterways usually integrate well into passive recreational bushland, but issues
relating to safety, visibility and vegetation can exist.
Safety can be an issue in relation to steep or unstable creek banks, and possible drowning
in natural or constructed pools. Visibility issues can relate to the need to have clear vision
of children playing throughout the park. Vegetation issues often relate to conflicts
between the need for open grassed areas and the creeks requirements for riparian
vegetation.
Residential areas in residential areas issues exist relating to safety, flooding and
mosquitoes. Safety issues usually relate to steep eroding creek banks and vermin.
Conflicts can also exist between the need to control flooding and the need for riparian
vegetation. Mosquito problems are often more noticeable immediately adjacent to
waterways.
Community consultation Can be a very important aspect of Natural Channel Design. NCDs usually have a long
vegetation development period. During the first few years the watercourse can become
overgrown with weeds until the canopy trees fully develop. This short-term appearance
can concern some people, therefore, it is important that the community is adequately
informed about both the expected long-term and short-term appearance of the channel.
As another example, in some circumstances, trees planted in areas close to the edge of an
unstable bank are fully expected to be undermined by erosion before they reach maturity.
These trees are planted because in the short-term they can provide a valuable role in the
rehabilitation of the watercourse. Such trees are referred to as sacrificial trees. In these
cases, if the sacrificial trees are planted by the community, then the community group
needs to be informed of the role of the tree so they are not concerned when the trees
begin to fall into the creek.
Conservation If creeks are expected to change as a result of the urbanisation of a catchment, then it is
often difficult to decide how much effort should be taken to protect the natural features of
a watercourse. These issues need to be addressed on a case-by-case basis.
15
Planning Aspects
Corridor linkage (human, Urban terrestrial movement corridors should be sized in accordance with Councils
aquatic and terrestrial) Ecological Assessment Guidelines, 1998.
Discontinuities within the corridor should be avoided. Discontinuities can be formed by

roads or buildings, or by the removal of essential riparian vegetation.
Desirable water quality To protect the ecological values of urban creeks, stormwater drainage and sediment
control within the catchment should reflect the basic aim of achieving high quality low-
flow water.
Ecological functionality The keys to ecological functionality and sustainability are: habitat diversity, water quality,
and sustainability environmental flows, shading and corridor movement.
Efficient drainage A watercourse by its nature is a drain. Traditionally, drainage formed the highest priority
in channel design. In NCD drainage issues should still be a priority. The difference is that
the drainage aspects need to be appropriately integrated with the economic, ecological
and aesthetic aspects of the channel.
Overbank floodplain areas need to drain freely into the channel to avoid overbank
ponding and mosquito breading problems.
Expected sediment flow within If the principles of Natural Channel Design are to be adopted, then there must be a the
watercourse commitment to adequately control sediment runoff from the catchment.
A sediment trap is often required immediately upstream of any creek rehabilitation works.
However, a sediment trap is usually not required on river projects or in sandy creeks with
movable beds.
Fauna monitoring Prior to commencing the detailed design of major channel works, a study should be
conducted of the existing fauna in and around the waterway. The extent of this study
should reflect the size of the project, its potential impact and the community values
associated with the fauna likely to exist in the area.
Where possible, guidelines and/or trigger points should be established that will initiate
changes to the channel design or construction methods if the monitoring conducted
during the construction phase indicates an adverse, unexpected or unacceptable impact
on the fauna.
Fire control The buffer width around a watercourse needs to allow for fire control and in some
circumstances the provision of fire access tracks.
Flood control Conflict often exists between the desire to have fully vegetated creeks and the need for
flood control. Further discussion is provided on this issue later in this chapter and in
Chapter 4.
Low maintenance Vegetation densities and Mannings n roughness values should be selected such that long-
term maintenance requirements of the waterway are minimised.
16
Open space management The channel design needs to be integrated with Councils Open Space Planning Strategy,
Local Plans and Park Master Plans.
Physical stability Channel instability is a natural component of most waterways; however, in urban areas it is
usually desirable to minimise channel instabilities. Therefore, steep, high, unstable banks
should generally be avoided.
However, steep unstable banks can constitute a significant habitat resource for some
wildlife species such as platypus, rainbow bee-eater, pardalote and fairy martin.
Protection of cultural values Recognised cultural values must be considered in the rehabilitation and reconstruction of
waterways.
Recreation The recreational aspects of waterways need to be integrated into the overall open space
strategy for the region.
Safety Safety issues that need to be reviewed in the planning phase include: the use of drop
structures and energy dissipaters, high flow velocity water and the existence of steep or
high channel banks.
Serviceability Maintenance access may be required along a watercourse depending on the fire risk, the
need for flood control and the requirements for debris and sediment removal.
Plan form Issues that should be considered when determining the desirable plan form of a
watercourse are listed below and discussed in detail in the following chapters.
(i) historical or natural location of the channel;

(ii) existence of steep, high channel banks or adjacent land forms;
(iii) flood flow;
(iv) bankfull velocity;
(v) bend radius.
Historical or natural location The historical or natural location of a watercourse can often be determined from
of creek historical aerial photographs, usually available at the local authority (e.g. Brisbane City
Councils Plan Custodian) or the Department of Natural Resources. Unfortunately, the
natural location of a watercourse may not be appropriate for an urbanised catchment
because changes in channel width caused by the changes in hydrology usually result in
change to the radius of the channel meanders.
However, even if the historical location of the watercourse is not totally valid for the
current catchment conditions, it can provide valuable information about the nature of the
watercourse. If channel straightening had occurred in the past, then the historical location
of the creek can indicate where potentially unstable backfill had been placed within the
old meandering channel.
Steep, high channel banks Where there are steep, high banks or other land forms adjacent to the existing channel,
any lateral movement of the channel towards these banks could mean extensive
earthworks and costs, or the formation of steep, potentially unstable banks.
17
Planning Aspects
This can often be a problem when rehabilitating channelised watercourses that were
constructed during the 70's and 80's when the primary goal was to maximise the
developable land. In these drainage channels the low flow channel is often located at the
base of steep 1 in 4 banks.
Introducing a meander into the plan form of a watercourse or into the low flow channel
can often only occur if a steep, heavily reinforced, bank is constructed on the outside of
the meander. These steep banks can be expensive to construct, but can also add
considerable diversity and aesthetics to the watercourse.
Flood flow If the channel and associated riparian vegetation meanders across the floodplain, then it is
likely to interfere with the passage of flood waters and thus increase upstream flood levels.
In such circumstances it may be desirable to decrease the density of the vegetation at
locations where flood waters cross from one side of the channel to the other. However,
this may decrease the value of the riparian zone and the value of the waterway as a
movement corridor.
It is important to select a meandering path that achieves the desired channel slope as well
as allowing for the required flood gradient.
Bankfull velocity If meanders are introduced to a straightened and channelised creek, and if it is desirable
not to have heavily armoured banks, then there should be sufficient meandering within the
watercourse to increase the effective channel roughness and thus reduce the bankfull
velocity to an acceptable level.
The desirable bankfull velocity will depend on canopy cover, channel vegetation and
channel material.
Bend radius Waterway corridors are often linked with bushland corridor and hilltop corridor links
Corridor widths (Figure 2.1). It is important that the terrestrial movement corridor requirements are clearly
established in regional plans and that these plans are referred to in the planning phase of
watercourse works.
Urban terrestrial movement corridors should be sized in accordance with Councils

Ecological Assessment Guidelines, 1998.

When developing a compromise on corridor width, remember that the middle ground
may not be good enough. For example, if the engineer requires a 70m wide corridor for
flood control, and a developer requires 20m for economics, and the ecologists requires
90m for an adequate movement corridor, then a compromise of (70 + 20 + 90)/3 = 60
may at first appear to be fair, but may not achieve the minimum requirements of a
functioning waterway.
18
ry
nda
bou
ent
tchm idor
Ca ay corr
erw
wat
n-
iparia
R
Bushland
reserve
ry
da
un
bo
t
en
el
m
nn
ha
tch
c
ay
Ca
terw
Wa
Bushland link
and terrestrial
Hilltop
corridor
bushland
Figure 2.1 Terrestrial, aquatic and bushland corridor links
Site constraints Possible site constraints include the following:
topography of area
encroachment of existing buildings/structures
waterway crossings/bridges/culverts/weirs/pipes
adjacent land use
land ownership
existing management plans for catchment/waterway
bed or bank rock
valued vegetation/trees
existing, highly-valued habitat features
stormwater outlets
soil conditions
19
Planning Aspects
Banks Existing bank-stabilising trees can influence the channel location, width and bank slope.
However, most trees will become unstable if their root system is undermined by bank
erosion. Therefore, if a watercourse channel wishes to expand or move, then most trees
will only delay the process, they will not prevent it in the long-term.
Trees If trees are removed from a steep or high bank, then consideration must be given to the
long-term stability of the bank. It may take five years before the root system within the
bank begins to deteriorate resulting in bank slips.
If the trees or shrub located on or near the watercourse bank cannot be replaced - say due
to flood control reasons - then the bank may need to be benched to compensate for the
long-term removal of essential root reinforcing within the bank.
Existing structures Existing buildings and structures (bikeways, bridges, culverts, etc) can affect the location,
width and/or desirable bank slope of a channel. Some structures can also dictate the
invert level of the channel.
Service lines (pipe crossings) can also influence the desired bed level and bank width.
Land ownership Land ownership can have a controlling influence on the location of a reconstructed
watercourse.
State Government requirements Existing State Government legislation can result in the State having the final say on any
proposed watercourse rehabilitation or relocation.
Rock Bedrock often defines the channel invert level, while bank rock usually defines its location
and width. The location of rock can sometimes be estimated from historic borehole
surveys preformed for the design of bridge foundations.
Stormwater outlets Existing stormwater outlets can influence the direction and location of channel meanders
as well as the channel invert level. However, it is often desirable to recess and/or
reconstruct stormwater outlets to better integrate the outlets into the new environment
and to avoid future erosion problems at the outlet.
When a channel is relocated, it is important to remember that all existing stormwater

outlets will need to be extended to the new channel location. Extending the existing
stormwater outlets may result in a significant lowering of the invert level of the outlet.
Recessing the outlets into the bank may avoid some of these problems and expenses.
Soil properties The local soil properties can affect the side slope of the channel. Heavy clay soils may
allow relatively steep bank slopes, while sandy soils are likely to require much flatter
slopes.
Sediment flow A natural or unnatural high sediment flow in a watercourse can significantly influence the
feasibility of introducing a pool-riffle system to a channel bed.
20
Bed conditions Desirable bed conditions in a reconstructed watercourse usually depends on the
following factors:
(i) catchment area;

(ii) catchment soil type (infiltration capacity);
(iii) canopy cover.
Flood relief works Flood relief can usually be achieved using one or more of the following options:
(i) snag management;

(ii) bank modifications (multi-stage channels);
(iii) flood relief bypass channels;
(iv) total reconstruction of the watercourse.
Snag management Discussion on snag management is contained in Councils Erosion Treatment for Urban
Creeks and Councils Environmental Best Practice Manual and maintenance manuals.
Bank modifications Many of the essential ecological features of a watercourse are contained within the
low flow channel. An aim of using bank modifications to achieve flood relief is to
minimise the impact on the channel bed.
Modifications to the channel banks are preferably made to one side only. An
advantage of this technique is that it can relieve hydraulic stress on the channel.
Figure 2.2 shows three different channel cross sections for decreasing flood capacity,
but increasing ecological benefit.
Further discussion is contained in Chapter 5 of Urban Stream Rehabilitation

Principles and Guidelines (1996).
21
Planning Aspects
Increased channel capacity
Natural level
Main stream channel

Modified level
Densely
vegetated Scattered vegetation
Main stream channel

Modified level
Densely vegetated
Main stream channel

Modified level
Figure 2.2 Possible flood relief modification to a waterway channel
22
Bypass channels Care needs to be taken when designing flood relief bypass channels to avoid adverse
effects caused by an increase in upstream flow velocities. Another erosion problem can
occur at the downstream end of the bypass channel when the bypass flood water re-
enters the main channel.
Further discussion is contained in Chapter 5 of Urban Stream Rehabilitation Principles

and Guidelines (1996).
Figure 2.3 shows a typical layout of a bypass channel.
A
Bypass
channel
High stage flows spill into

Retain existing creek adjacent bypass channel
A A
Figure 2.3 Flood relief bypass channel
23
Design Procedures
3. Design Procedures
Introduction This chapter presents a procedure for the design of natural channels. The chapter
primarily focuses on those physical properties of the channel required for hydraulic
stability. The design procedure is presented only as a guide and in most circumstances
final channel dimensions should be appropriately varied along the channel to form a
natural rather than constructed appearance.
Note: This chapter should only need to be reviewed if the proposed stream rehabilitation
works require the relocation, reconstruction or alteration of the main channel. In any case,
wherever possible, all appropriate channel features (Appendix A) should be incorporated
into the existing or reconstructed channel.
Depending on the type of creek works proposed, there can be a number of design steps
that should and should not be performed. In some cases the channel location may be
preset, in other cases the purpose of the creek works may only be to improve the
ecological values of the channel bed or banks. In any case, the design philosophy should
be to review only those design steps relevant to the given circumstances.
The design procedure presented below is based on the following design philosophy:
(i) The channels geometry, bed and floodplain are influenced by different design flow
rates, those being: bankfull flow, low flows and flood flows.
(ii) The channel width, depth, meander radius and allowable channel slope are
determined by the adopted bankfull flow rate and the allowable bankfull flow
velocity at various stages of revegetation.
(iii) The bed design is determined by the adopted channel slope and low flow conditions.
(iv) The overbank geometry and vegetation density is governed, in part, by high level
flood flows.
Remember: There is more than one way to solve a hydraulic problem, and there is more
than one way to solve a ecological problem. It is important to be creative, thoughtful and
flexible.
Also, aesthetics is often not the primary controlling factor. Good aesthetics does not
necessarily result in good ecological values.
24
Data collection Where possible, obtain a long section survey of the channel bed. This will assist in the
selection of channel gradient and in locating pools and riffles.
Soil tests should be conducted to determine if the soil is dispersive (Emerson Aggregate
Test) and what type of soil it is, e.g. sandy, sandy loam, or clay.
Design steps The following steps are used in this design procedure.
Step 1 Select the reach of watercourse to be rehabilitated
Step 2 Determine channel fall
Step 3 Determine bankfull flow rate
Step 4 Select the type of channel
Step 5 Determine channel width and channel depth
Step 6 Channel bank slope
Step 7 Determine typical channel meander radius
Step 8 Select trial channel location
Step 9 Review riparian and floodplain vegetation
Step 10 Design of channel bed and low flow channel
Step 10-1(a) Bed design for channel with little or no bed vegetation - determine bed form
Step 10-1(b) Bed design for channels with bed vegetation determine bed form
Step 10-2(b) Determine the length of the low flow channel
Step 10-3(b) Determine the number of riffles
Step 10-4(b) Determine typical riffle dimensions
Step 10-5(b) Determine typical pool dimensions
Step 10-6(b) Determine pool length
Step 10-7(b) Determine riffle rock size
Step 10-8(b) Design low flow channels
Step 11 Incorporate channel features
25
Design Procedures
Symbols
D [m] Channel depth from bankfull water level to channel invert
DN [m] Nominal rock diameter of which N% of the rock is smaller, i.e. 10% of the rock is smaller
than the D10 rock size. Rock size is measured as the diameter of a sphere with an
equivalent volume to the individual rock
Dp [m] Maximum pool depth
d/s [-] Downstream
Fr [m] Riffle fall measured from riffle crest to the low flow pool level within the downstream pool
Ft [m] Total channel fall across the proposed rehabilitation area equals upstream channel invert
level minus downstream channel invert level
L1 [m] The valley length measured along a smooth straight or curved path (not including the
channel meanders) from the upstream end of the channel reach to the downstream end of
the channel reach
L2 [m] Length of a meandering watercourse channel measured along the channel centreline
L3 [m] Length of the low flow channel measured over a given reach of the watercourse
Lm [m] Straight line distance between successive channel meanders, or meander inflection points
Lp [m] Length of a low flow pool measured at the low flow water level
Lr [m] Riffle length measured from the riffle crest to the upstream end of the downstream low
flow pool
N [-] Number of riffles within a reach
n [-] Mannings roughness value
q [m3/s/m] Discharge per unit width for a nominated design event
Q [m3/s] Discharge
Qf [m3/s] Bankfull discharge
Qfds [m3/s] Bankfull discharge immediately downstream of the works
Qfus [m3/s] Bankfull discharge immediately upstream of the works
Q [N] [m3/s] The peak design discharge at a given location that is expected to occur on average once
every N years, i.e. the Q100 is the expected peak discharge during the 1 in 100 year
design storm
26
Symbols
Rc [m] Channel bend radius measured to the centre of the channel, or the low flow channel
meander radius measured to the centre of the low flow channel
Ro [m] Bend radius measured to the outer bank of the channel at the elevation of bankfull flow
S [m/m] Slope
So [m/m] Average channel slope over a selected reach
Sr [m/m] Slope of downstream face of riffle
Si [-] Sinuosity = L2/L1 or L3/L2
u/s [-] Upstream
V [m/s] Average flow velocity
W [m] Channel top width measured at the height of the lower bank
Wp [m] Maximum pool width measured at low flow water level
Wr [m] Riffle bed width
[m] Average meander displacement, or half the amplitude of the meandering channel
centreline or low flow channel centreline
[deg] The angle measured between two consecutive inflection points
Step 1. Select the reach of watercourse to be rehabilitated
Waterways are normally rehabilitated in stages. Often the length, or reach, of a

watercourse to be rehabilitated is determined by external factors beyond the control of
the designer; however, where practicable, the length of rehabilitation should be restricted
to a hydraulically and financially manageable reach length.
A hydraulically manageable reach length is one where the occurrence of a 1 in 2 year

flood during the revegetation phase does not result in unacceptable bank erosion or
damage to newly established plants. A financially manageable reach length relates to the
initial establishment cost as well as the short-term maintenance costs (i.e. weed control).
The 1 in 2 year (or Q2) flow is important because it is considered that if most new plants
are able to survive the first 2 years without suffering damage from a flood event, then they
are likely to reach maturity. In a channel where the bankfull flow rate was around, say, the
Q10 event, it would be unlikely for a Q10 event to occur within the first two years.
Discussion on selecting the maximum reach length is provided in Technical Note 3.1.
27
Design Procedures
TECHNICAL NOTE 3.1

Often the governing factor in natural channel design is the allowable bankfull velocity during the early revegetation
phase when the soil is disturbed and susceptible to erosion, and the effective channel roughness is low. Typically,
during this post construction phase the Mannings n roughness of the channel is around n = 0.04 depending on the
irregularity of the channel, the degree of meandering, and the size of the planted seedlings.
If bankfull flow is less than say Q2, then it is likely that a bankfull flow will occur during the early growth stage of the
plants. Therefore, if the bankfull flow velocities are high, say greater than 1 m/s (depending on soil type), then it is
likely that significant damage will occur to the plants before they reach maturity.
To avoid such problems, it may be desirable to rely on existing high channel roughness downstream of the
rehabilitation works to reduce bankfull flow velocities throughout the work area. However, this will only be successful if
the area of the proposed works is within the backwater region of the downstream channel. This can be confirmed
with the use of a hydraulic backwater model, otherwise restrictions may need to be placed on the maximum allowable
channel fall across a proposed section of creek works.
In the absence of any other data or model results, if bankfull flow velocities are expected to be high, then the length
of the proposed rehabilitation works should not extend over a total channel fall equal to half the channel depth (D).
Under these conditions, it is likely that the channel works will occur within the backwater (ie low velocity) region of the
downstream, vegetated channel.
Step 2. Determine channel fall (Ft)
It is important to walk the full length of any proposed channel works prior to commencing
the design or even planning the works. Take note of any hydraulic controls, such as bed
rock, pipe crossings, culverts or drop structures (natural or constructed).
Any fixed bed features, such as bedrock, will affect the allowable fall of the channel bed
and thus the profile of the channel bed as shown in Figure 3.1.
Measure the elevation of any fixed bed structures and the total channel fall (Ft) across the
proposed rehabilitation length (L2).
Ft
L2
Figure 3.1 Channel long section
28
Step 3. Determine bankfull flow rate (Qf)
Estimate bankfull flow rate upstream (Qfus) and downstream (Qfds) of the proposed works.
In the unlikely situation where the bankfull flow rate is greater than the estimated Q100
discharge, then adopt the Q100 discharge.
Calculate the Q1, Q1.5, Q2 and Q10 flow rates.
As a first trial, nominate the bankfull flow rate (Qf) based on the most appropriate of the
following recommendations:
(i) Average of Qfus and Qfds (if the length of the rehabilitated channel is small and there is
a desire (aesthetically or otherwise) for the new channel to be of a similar size to the
upstream and downstream channels.
(ii) A calculated bankfull flow rate (Qf) based on the monitoring of similar, stable streams
and catchment land uses in the local region.
(iii) Q1.5, if no specific regime equations have been developed for the local area, noting
that in the Brisbane area, bankfull flow rates vary from less than Q1 to greater than
Q10.
(iv) A bankfull flow value based on local site constraints.
In some circumstances the channel capacity may be determined by the requirements for
flood control. In such cases a hydraulic model and study will be required for the complete
floodplain. There is significant interaction between the channel and its floodplain and this
should not be forgotten throughout the full design procedure.
In long reaches it may be necessary to vary the bankfull discharge down the reach.
Note: at Step 8 it may be necessary to reduce the nominated bankfull flow rate in order to
obtain a stable channel (eg to reduce the bankfull velocity).
Step 4. Select the type of channel
The desirable style of drainage channel can vary from a grass-lined overland flow path for
very small catchments to a fully established river channel for large catchments. Desirable
bed conditions in a reconstructed watercourse usually depend on the following factors:
catchment area
catchment soil type (infiltration capacity)
canopy cover
Table 3.1 provides typical bed form conditions that are likely to occur in the Brisbane
(Clayey Soil) region. The desirability of these bed form conditions of course depends on
local hydraulic and ecological factors.
29
Design Procedures
Table 3.1 Typical (desirable) bed conditions for the Brisbane region
Catchment area Channel slope Desirable bed condition
Less than 30 ha Open canopy Grassed bed

urbanised catchment Closed canopy Rocky and/or vegetated bed
Urbanised catchment between Steep channels Rock lined bed
between 30 and 100 ha Grades > 1 in 20
Mild slopes Pool/riffle system, however, low flows
Grades between may not be sufficient to maintain dry weather flows
1 in 20 and 1 in 100 and adequate pool water quality
Flat slopes Open earth or vegetated channel bed
Grades < 1 in 500
Sandy channels Sandy beds, some pools and riffles
Greater than 100 ha mainly Steep channels Rock lined bed
urbanised catchment Grades > 1 in 20
Mild slopes Pool/riffle system
Grades between
1 in 20 and 1 in 100
Flat slopes Open earth or vegetated channel bed
Grades < 1 in 500
Sandy channels Sandy, irregular bed with pools
Less then 50 ha bushland catchment Open canopy Grassed bed
Closed canopy Rocky and/or vegetated, bed
Greater than 50 ha rural or Steep slopes > 1 in 150 Rock lined
bushland catchments Mild slopes < 1 in 150 Pool/riffle system
Flat slopes < 1 in 500 Open channel
Sandy channels Sandy bed, some pools and riffles
30
Step 5. Determine channel width (W) and channel depth (D)
The channel width (W) is defined as the channels top width measured at the elevation of
its lowest bank (Figure 3.2). The channel depth (D) is defined as the elevation difference
between bankfull water level and the channel invert.
When short sections of a watercourse are to be reconstructed or rehabilitated, it is often

desirable to maintain a channel depth similar to the existing channel depth to reduce
earthworks and to allow the channel bed to have a constant fall to the existing
downstream channel invert.
For the reconstruction of gravel bed channels, the desired channel width and depth may
be determined using one of two methods: the Hey and Thorne equations, or the Brisbane
equations. For the reconstruction of sandy bed channels, the Simons and Albertson
equation should be used.
Figure 3.2 Channel and floodplain cross section
TECHNICAL NOTE 3.2

The Hey and Thorne equations presented below for gravel bed waterways suggest that the channel width is
significantly greater for an open canopy channel (e.g. grassed channel) than for a closed canopy channel with a dense
stand of bank vegetation. This may be the case for streams with little or no bed vegetation, such as rivers and streams
fed by snow melts, but for creeks in the Brisbane region, observations have shown that closed canopy creeks are often
wider than open canopy creek systems. This is because open canopy creeks generally contain more groundcover
vegetation (usually grasses and weeds) and the presence of this groundcover greatly reduces erosion.
If bed vegetation is not significant within the given type of watercourse, then the Hey and Thorne equations should be
used to estimate bank width and depth. If bed vegetation is a common feature of the type of watercourse you are
designing, then the Brisbane-based equations are recommended. However, the final judgement on how channel
width and depth are determined, is left to the designers experience and judgement.
It is always useful to compare the results of the two sets of equations with existing channels in the region.
31
Design Procedures
(i) Hey and Thorne equations for gravel bed streams
W = 4.33 x (Qf ) 0.5

for vegetation type 1 grassed riparian zone, no trees and
shrubs, Mannings n typically 0.03 to 0.035
W = 3.33 x (Qf ) 0.5

for vegetation type 2 scattered trees and shrubs, dense grass
and weeds, Mannings n typically 0.04 to 0.06
W = 2.73 x (Qf ) 0.5

for vegetation type 3 light to medium stand of trees and
W = 2.34 x (Qf ) 0.5

for vegetation type 4 medium to dense stand of trees and
(ii) Brisbanes clay-based creek systems with significant sand and gravel bed deposits
Qf 100 m3/s - typical bank width, W = 2.41 (Qf ) 0.5

and typical depth,
D = 0.75 (Qf ) 0.5
Qf < 100 m3/s - typical bank width, W = 4.37 (Qf ) 0.373

and typical depth,
D = 1.07 (Qf ) 0.224
The typical range of bank widths: 4.33 (Qf ) 0.5

> W > 1.78 (Qf ) 0.5 for all flow rates.
The typical range of bank depths: 0.598 W 0.6

> D > 0.295 W 0.6
(iii) Simons and Albertson equations for sand bed waterways
The equations may be applied to all sand bed streams in the Brisbane area. It should be
noted that in some cases, this may be outside the range of data from which the equations
were derived.
(a) For sand bank (sandy loam)
Bed width (m) = 5.72.Qf 0.512
Mean depth (m) = 0.504.Qf 0.361
(b) For cohesive banks (some clay content or internal strength)
Bed width (m) = 4.29.Qf 0.512
Mean depth (m) = 0.59.Qf 0.361
Surface width (m) = 1.1 x Bed width + 2. This should yield bank batters in the range
2H:1V for cohesive banks and 3H:1V for sandy banks.
(c) For sand and cohesive banks

Meander arc length (m) = 6.31.W
32
Step 6. Channel bank slope
The desirable channel bank slope on reconstructed channels may be determined from
one of the following methods:
(i) Obtained from Table 3.2.

(ii) Obtained from geotechnical investigations.
(iii) A bank slope similar to existing stable upstream and/or downstream bank slopes.
In short reaches it is often necessary to blend in with the bank slope immediately
upstream and downstream of the proposed channel works. However, in most cases
trying to re-establish a steep bank slope similar to existing conditions will result in
poor revegetation of the banks.
Introducing a wide variation of bank slopes can be an effective tool in making a

constructed or modified watercourse appear as natural as possible. It is usually highly
desirable to avoid a regular trapezoidal channel shape down the reach of a
watercourse.
Steep banks of the outside of bends may be natural, but they are also highly unstable.
Avoid placing steep, unprotected banks in areas where watercourse stability is desired.
Berms (Figure 3.3) may need to be introduce to steep banks that are higher than 3 metres,
or where erosion is expected along the toe of the bank. A typical berm width is one
metre.
It is noted that bank erosion is frequently caused by deepening of the channel (bed
erosion). Therefore, there is often no point in treating the banks unless the bed is
stabilised first.
Where possible, when designing a modified watercourse, restrict the gradients on the
'over-bank' or 'high-bank' areas to 6(H):1(V) with a maximum grade of 5(H):1(V). Generally,
for safety reasons, grassed bank grades of 4(H):1(V) and steeper are not recommended
adjacent to a watercourse where:
public access exists;
the watercourse is considered a safety hazard during dry weather conditions or during
regular flood events (up to the 5 year ARI); and
satisfactory safety barriers such as fences, shrubs, bushland or flat landings do not exist
between the embankment and floodway.
33
Design Procedures
Berm or terrace
Figure 3.3 Berm located on steep, high bank
Table 3.2 Suggested bank slopes [1]
(V):(H) Bank description
1:2 Good, erosion resistant clay or clay-loam soils with a healthy, deep-rooted bank vegetation
formed by suitable riparian groundcover species, shrubs and trees
1:3 Sandy-loam soil with groundcover vegetation or a closed canopy channel with shaded banks and
sparse bank vegetation
1:4 Sandy soils
1:6 Mowable, grassed banks (undesirable in Natural Channel Design)
Note [1] The bank slopes presented in this table should be considered as a general guide only. Final bank slope
should be based on bank stability and revegetation requirements.
34
TECHNICAL NOTE 3.3

Establishing a berm on a watercourse bank can provide the following benefits:
(i) If bank erosion occurs at the toe of the bank and this erosion results in soil slumping on the bank, then the
berm can reduce the area of the bank over which this slumping occurs.
(ii) Bank berms can be used to increase the stability of steep banks.
(iii) Berms can be used to delay the effects of soil erosion around the root system of establishing trees.
(iv) Wide berms can be used to provide pedestrian access to the watercourse banks for recreational and
educational purposes.
Step 7. Determine typical channel meander radius
If a channel is to be reconstructed or relocated (ie when replacing an existing concrete

drain with a natural channel) it is often desirable, if not necessary, to introduce some
channel meanders. Meandering the channel can improve its aesthetics, increase habitat
and channel diversity (through large-scale hydraulic turbulence) and can increase the
effective channel length. Increasing the channel length is one of the most effective ways
of reducing the bankfull flow velocity.
The meander radius to the centre of the channel (Figure 3.4) is generally greater than
three times the normal channel width.
Select a desirable meander radius (to centre of channel) such that: Rc 3W
Alternatively, the meander radius to outer bank is given by: Ro 3.5W.
If a sharper meander radius needs to be constructed, then it may be necessary to rock-line

the outer bank of the meander.
Channel meander
W
Rc
Ro
l
ne
an
y ch
a
t erw
Wa
Figure 3.4 Channel meander
35
Design Procedures
Step 8. Select an initial channel location
Often stream rehabilitation projects commence with a desired channel layout already
known. However, as the design procedure progresses and hydraulic checks are done, it
may become necessary to adjust the channel location until a stable channel geometry and
flow condition is obtained.
Some of the issues that should be considered when determining the desirable plan form
of a watercourse are listed below:
(i) historical or natural location of the channel;

(ii) existence of steep, high channel banks or adjacent land forms;
(iii) flood flow;
(iv) bankfull velocity;
(v) bend radius;
(vi) location of essential or valued floodplain features (eg existing football fields, existing
bushland or mature trees);
(vii) current channel location.
Discussion on each of these issues is provided in Chapter 2, Planning Aspects.
The channel normally follows the invert of the valley floor so that the floodplain can drain
freely into the channel. Ponding on the floodplain can cause maintenance problems (eg
mowing) and mosquitoes; however, in some locations, ponding may be desirable to
promote frog breeding. Expert advice should be sought.
The channel location is important because it establishes the effective channel slope, and it
is this slope that determines bankfull velocity. In many cases the ideal channel slope and
dimensions cannot be achieved due to site constraints such as the proximity of existing
structures, the desirable meander radius or the required minimum bed width.
The location of the channel is often controlled by existing site constraints and urban
structures such as bridges, culverts, stormwater outlets, rare or otherwise valued trees,
including canopy trees, and other public assets. These items should be clearly identified
on field or survey plans. A channel system should never be designed without performing
a site visit and walking the length of the valley.
In cases where the location of the channel is defined by site constraints, consideration
should be given to the formation of minor channel meanders or variations in channel bank
slope to provide some degree of irregularity to the channel.
Two procedures are recommended for the establishment of a trial channel location:
(a) the allowable velocity method; and

(b) the allowable channel slope method.
Each of these methods is detailed over.
36
(a) Allowable velocity method
(i) Estimate channel roughness for both short-term (post-construction, say n0.04) and
long-term conditions (fully vegetated).
(ii) Determine the allowable bankfull velocity for both post-construction condition (low
vegetation roughness Table 3.7) and long-term channel conditions (high vegetation
roughness Table 3.3).
(iii) Calculate the desired maximum channel slope (S) using the Mannings Equation or a
hydraulic backwater model for the following conditions:
(a) initial channel roughness during Q2 or bankfull flow whichever is lower; and
(b) long-term channel roughness during Q100 or bankfull flow whichever is lower
(in this case the 1 in 100 year flood event is used because it is the standard
design event in Brisbane for major channels).
Table 3.3 Recommended maximum velocity for various Mannings roughness to avoid
significant vegetation damage[1]
Average Mannings n roughness Recommended maximum velocity during bankfull

flow and 1 in 50 year flood event [2]
n = 0.03 2.0 m/s
n = 0.06 1.7 m/s
n = 0.09 1.5 m/s
n = 0.15 1.0 m/s
Notes:
[1] As the vegetation density increases, and thus the channel roughness increases, it
becomes harder for flood flows to pass around or through the vegetation without causing
significant damage. A Mannings roughness of 0.03 represents a typical, deep water,
grass channel where high flow velocities of around 2 m/s are expected to cause only
minor damage. However, a Mannings roughness of 0.15 represents typical natural
bushland for the Brisbane region. If a flow velocity of 2 m/s was allowed to flow through
such bushland, significant damage to the vegetation would be expected.
[2] The 1 in 50 year flood event is chosen because it represents an extreme flood event
during which some vegetation damage would be expected, but it is desirable to minimise
this damage where possible. It is considered that a 1 in 20 year flood event would in
most cases be too small to act as the design event for vegetation damage. The 1 in 100
year flood can be chosen, but it may be considered unrealistic to have minimal vegetation
damage during such an extreme event.
37
Design Procedures
(iv) Determine the channel length (L2): where L2 = (Ft /S) Refer to Figure 3.5.
(v) Measure the valley length (L1)
(vi) Determine the required channel sinuosity: Si = (L2 /L1)
(vii) Determine the average meander displacement () using Table 3.4 and Figure 3.6.
(viii) Sketch the trial channel centreline on the valley plan.
(ix) If the required channel length is excessive, then revise the channel location using
Method (b) below.
TECHNICAL NOTE 3.4

In the above design method it is assumed that bankfull velocity is controlled by the average channel slope (i.e. the
pool/riffle system is totally drowned out). Such a flow condition can be assumed to exist if the channel depth (D) is at
least twice the average riffle fall (Fr). In some cases it may be desirable to sketch an undistorted long section of the
channel bed to visualise what possible impact that the pools and riffles may have on bankfull flow velocity. If the riffle
height is significant, then the pools and riffles should be included in the hydraulic backwater model and the cross
section spacing should be appropriately reduced.
L2
Figure 3.5 Channel reach length
38
Table 3.4 Meander dimensions
Meander Sinuosity /Rc Lm/Rc Meander Sinuosity /Rc Lm/Rc

angle () (L2/L1) angle ( ) (L2/L1)
10 1.0013 0.0038 0.174 80 1.086 0.234 1.286
20 1.005 0.015 0.347 90 1.111 0.293 1.414
30 1.012 0.034 0.518 100 1.139 0.357 1.532
40 1.021 0.060 0.684 120 1.209 0.500 1.732
50 1.032 0.094 0.845 140 1.300 0.658 1.879
60 1.047 0.134 1.000 160 1.418 0.826 1.970
70 1.065 0.181 1.147 180 1.571 1.000 2.00
Lm
Rc
Wa
ter w el
ay c h a n n
Figure 3.6 Channel meander
(b) Allowable channel slope method
(i) Select a trial channel location based on typical channel meander patterns upstream
and downstream of the proposed works (assuming the u/s and d/s channel reaches
are in the form of a natural channel design).
(ii) Estimate the channel roughness for both short-term (post-construction) and long-term
conditions.
(iii) Given the measured channel fall (Ft) and channel length (L2), determine the channel
slope (S). It is noted that the channel length is measured along the centreline of the
channel, not along the low flow channel.
39
Design Procedures
(iv) Calculate the short-term (post-construction) bankfull or Q2 flow velocity (whichever is

the lower discharge) and long-term bankfull or Q100 flow velocity (whichever is the
lower discharge).
Note: In the short term we are interested only in flood events less than the Q2 because it
is desirable to minimise vegetation damage during the early growing phase. In the long
term we need to consider the Q100 event because there may be factors other than
vegetation damage that we need to consider, such as public safety. If there are no other
stated velocity control issues other than vegetation damage, then the Q50 event should
be checked rather than the Q100.
(v) Compare the above channel velocities with the desirable flow velocities for the given
channel conditions. If channel velocities are excessive, then modify the trial channel
location or bed slope using one or more of the following options:
(a) Increase the channel length.
(b) Increase the channel roughness - usually by increasing the post-construction

channel roughness by planting more plants during the post-construction period
or by planting more mature plants.
(c) Decrease the channel gradient by installing a rock chute or a series of rock
chutes (vertical or near-vertical drop structures should be avoided in locations
where fish migration is important).
It is noted that installing a pool-riffle system will not decrease the effective
channel gradient; however, it may slightly increase the channel roughness.
(d) Decrease the bankfull discharge, i.e decrease the channel width and depth.
(e) Maintain the selected bankfull discharge, but decrease the channel depth and
consequently increase its width.
When locating a trial channel location, consideration should be given to the following points:
(i) The typical spacing of inflection points (i.e average spacing of meanders, Lm) is in the
range 5W<Lm<7W.
(ii) When the valley length (L1) and the desired channel length (L2) are known, the average
meander spacing (Lm) and amplitude can be determined from Table 3.4, where
amplitude = 2.
(iii) The valley topography and the encroachment of existing and/or future buildings,
roads or other public or private assets can affect the location of the waterway channel.
Where practicable, public assets such as roads and buildings should not be located within
15 m of the channel bank, or within the zone defined by a 1(V):3(H) gradient from the toe
of the bank, unless the bank is formed from stable rock (Figure 3.7).
40
Note that if a channel meander approaches a steep bank, then any future movement of
the channel towards the bank may result in significant erosion, bank instabilities and
sediment loss to the watercourse.
15m
Figure 3.7 Minimum placement of structures from channel bank
(iv) Significant benefit can be achieved by starting rehabilitation in the upstream reaches of
a watercourse thereby using the existing downstream fully vegetated or weed-infested
creek to reduce bankfull velocities within the rehabilitated reach. Unfortunately, this
usually does not apply to projects involving the reconstruction of concrete lined
drainage channels where it may be more desirable to start at the downstream end of
the watercourse.
Step 9. Review riparian and floodplain vegetation
When an acceptable trial channel alignment and cross section has been obtained, check
that riparian vegetation can be introduced along the edges of the watercourse without
causing unacceptable flood levels during the peak design flood event, usually the 1 in 100
year flood.
Chapter 4 provides information on the design of riparian and floodplain vegetation.
If the proposed channel meanders across the floodplain, then the establishment of riparian
vegetation along the banks of the channel may significantly affect the hydraulic capacity of
the floodplain. If this is the case, then the options are:
(i) Decrease the density of riparian vegetation in selected areas to allow overbank flood
flows to pass down the valley with few restrictions (refer to Chapter 4).
(ii) Relocate the channel along one side of the valley (where channel velocities allow) to
reduce the potential interference on overbank flood flows.
41
Rock size may be determined using the guidelines presented in Step 10-7(b), below.
Step 10-1 (b) Determine bed form
(b) Watercourses containing some form of bed vegetation
The bed form may be determined using one of the following methods:
(i) A bed form contained in a similar nearby watercourse, or within the existing
watercourse.
(ii) A bed form suitable for the terrestrial and/or aquatic wildlife known to inhabit or
migrate through the local area.
(iii) A bed form obtained form the guidelines presented in Table 1.
TECHNICAL NOTE 3.5

It is generally not desirable to establish a particular bed form for encouraging, non-native species to an area (i.e.
establishing a pool/riffle system within an area that does not normally contain pools and riffles). However, it is noted
that when urbanisation or other forms of land development significantly change the catchment hydrology, it is possible
for the bed form of a watercourse to also change resulting in pools and riffles being formed in areas where they
previously didnt exist.
Step 10-2 (b) Determine the length of the low flow channel
The low flow channel typically meanders within the bed of the channel, occasionally
bouncing off the banks, especially around bends in the channel. The sinuosity of the low
flow channel within the bed can be totally different from the sinuosity of the main channel
within the valley.
Rc
hannel centre l
amc ine
tre
S
Lm
Figure 3.8 Low flow channel meander
43
Step 10-4 (b) Determine typical riffle dimensions
Within the Brisbane region, the riffle slope was found to be highly variable, but typically
between 1 in 30 and 1 in 100. Constructed riffles usually have an upstream slope of 1 in 4
and a downstream slope of 1 in 20 or flatter.
In locations where fish migration is expected, the riffle slope should not be steeper than 1
in 30. In the upper reaches of small catchments where pool-riffle systems are being
introduced to areas that they did not naturally exist, then a riffle slope of 1 in 20 may be
acceptable.
To allow fish migration, the longer the riffle, the flatter it must be because fish can only
sustain burst speed for short periods. This often makes the design of riffles very difficult.
It is for this reason that a riffle fall of greater than 0.5 metres is considered rare.
Riffle length (Lr) is determined by dividing the nominated riffle fall (Fr) by the riffle slope (Sr).
The riffle width (Wr) is typically 1 to 3 metres. Where possible, the width of the riffle
should be determined from the width of the existing low flow channel upstream and/or
downstream of the proposed works.
If a pool-riffle system is constructed within a wide channel, it is likely that only a small part
of the riffles fully constructed width will be exposed at any one time. The rest of the riffle
will be covered with sediment and bed vegetation (i.e. the pool riffle system forms a
subset of the channel bed, refer to Appendix A, Figure A17).
Because it is difficult to control where the low flow channel will pass over the riffle crest, it
will be necessary for the rock riffle to be constructed across the full width of the channel
bed. However, if the low flow channel is well defined, and the channel bed is stable, then
the riffle may only need to be recessed approximately one metre into the banks of the
existing low flow channel.
It is essential for the riffle width, length, fall and slope should vary slightly from riffle to
riffle to obtain some degree of habitat and aesthetic diversity.
TECHNICAL NOTE 3.7

Riffles are generally straight. Only mild sloping riffles, say flatter than 1 in 30, are likely to have a slight curve.
Changes in the direction of the low flow channel typically only occur within the pools or at the riffle crest.
The riffle crest should be aligned perpendicular to the riffle chute, and riffle chute should be aimed directly into the
centre of the downstream pool. This will produce even turbulence and good energy loss in the pool, thus avoiding
damaging whirlpool effects.
45
Design Procedures
Riffle in a narrow low flow channel
0.2 - 0.4 m
(typical)
Part of riffle covered in

vegetation
Part of riffle exposed
to trickle flows
Riffle in a wide channel
Rough, open void surface

with some emergent rocks
in wide riffles
Flow 4
1 30
1
Pool
Pool
Riffle profile
Figure 3.9 Typical riffle geometry in creeks
46
Step 10-5 (b) Determine typical pool dimensions
Pool width (Wp) can be selected from the range shown in Figure 3.11 for a given riffle fall
(Fr).
Select a pool depth (Dp) such that the depth is equal to, or
greater than, the riffle fall: Dp Fr
The maximum pool depth normally occurs within the middle

third of the pool length (Figure 3.10).
Riffle
Fr Dp
Pool
Lr
Lp
Figure 3.10 Pool-riffle system
Following initial selection of Dp and Wp, check that:
Dp (0.221 Wp) - 0.28
or alternatively
Wp 1.27 + 4.52 Dp
If the above conditions are not satisfied, then select a new pool width based on the above
equation. If the required pool width is greater than the proposed channel bed width (Wb),
then the options are:
(i) Increase the channel bed width.

(ii) Increase the bank slopes adjacent to the pools, possibly with the use of rock.
(iii) Line the pools in rock to avoid bank erosion.
(iv) Adopt a pool width equal to the maximum bed width and then select a maximum riffle
fall from Figure 4.11, then return to Step 10-3.
47
Pool length (Lp) is usually between 2 to 4 times the pool width (Wp). Typical pool lengths in
Brisbane creeks were found to be between 5 and 50 metres for a 20 km2 catchment area.
The pool width, depth and length should vary from pool to pool (following the above
rules) to obtain an acceptable degree of habitat and aesthetic diversity. Pool length and
riffle spacing may need to be adjusted to locate a pool on the outside of significant
channel meanders.
TECHNICAL NOTE 3.8

On mild slopes, the pool length can become excessive. If this is the case, then it may be necessary to form a small
low flow channel between the end of the pool and the downstream riffle. The design of low flow channels is
discussed below.
Step 10-7 (b) Determine riffle rock size
In sand and clay-based watercourses where the water depth is significantly greater than
the anticipated riffle rock size and the riffle rock is required to be stable under normal flow
conditions (i.e the rocks do not move during bankfull flow), the riffle rock size may be
determined from the following formulas.
Steep creeks with short riffles D50 = 2 Fr (Sr) 0.50
Long riffles D50 = 8.45 y (Sr) 0.75

or
D50 = 1.42 (q) 0.67

(Sr) 0.50
In existing gravel bed streams, the riffle rock size should be determined from surveys of
existing riffle systems within the watercourse. In such cases it is likely that the rocks will
not be stable during bankfull flows. However, it is also likely that any rock removed from
the riffle during a flood event will be replaced by a similar rock migrating down the stream
during that event.
Chutes with a fall greater than one metre should be designed using appropriate design
guidelines such as the Denver charts or Brisbane City Councils creek erosion guidelines.
In all cases D10 10 y Sr
where:
D50 = nominal rock diameter of which 50% of the rock is smaller, [m]
D10 = nominal rock diameter of which 10% of the rock is smaller, [m]
Fr = riffle fall, [m]
Sr = riffle slope, [m/m]
y = depth of flow down riffle for the design discharge, [m]
q = discharge per unit width down the riffle for the design discharge, [m3/s/m]
49
Design Procedures
TECHNICAL NOTE 3.9

The above equations are based on a Mannings n roughness for rocks chute given by the formula:
n = 0.0482 (D50) 1/6
And, an allowable D50 rock size determined by the formula:
D50 = 0.04 V 2
Where V = average flow velocity down the riffle/chute for the design discharge, [m/s].
The above Mannings n equation is appropriate only when the water depth is greater than the rock size, which may not
be the case in some constructed riffles. Thus some errors may be expected with this design procedure. It is therefore
good practice to treat these equations with caution. In any case, if concerns are raised, then the best alternative is to
observe the typical rock size in existing creek systems of similar size and hydraulic gradient.
Step 10-8 (b) Design of low flow channel
In flat graded areas, a low flow channel may be required between a pool the downstream
riffle to avoid the formation of excessively long pools. The maximum size of pools within a
watercourse channel is in part governed by the dry weather flow rate. If the pool is too
large, then it may have insufficient through flow to maintain a desirable water quality and
control mosquito breeding.
This is especially true in urban areas where base flow water quality is likely to be inferior to
a natural stream. It is partly for this reason that the pool size is closely related to the
dimensions of the immediate upstream riffle.
As the overall channel grade decreases, the distance between riffles is likely to increase.
Now if the size of the downstream pool is limited by the quantity of base flow, then a low
flow channel must be constructed between the pool and the immediate downstream riffle.
This low flow channel should be designed to maintain non-erosive velocities during
channel full conditions. In this case, channel full conditions refers to the size of the low
flow channel, not the size of the main channel.
Where possible, low flow channels should be sized in accordance with existing conditions,
either at the site, or upstream or downstream of the site. In the absence of site specific
flow data, the trickle flow rate may be determined from Figure 3.12 (Brisbanes annual
rainfall approx. 1200 mm). The low flow channel may be designed to carry a total flow
rate of 1 to 10 times the estimated trickle flow rate.
50
Typical flow velocities for an open earth (n = 0.04), low flow channel (0.3 m depth, 1.0 m
width) are presented below.
Table 3.6 Typical low flow channel velocities

Low flow channel slope Channel velocity (m/s)
1 in 500 (0.002) 0.37
1 in 400 (0.0025) 0.41
1 in 300 (0.0033) 0.47
1 in 200 (0.005) 0.58
1 in 100 (0.010) 0.82
The low flow channel may need to be rock lined if flow velocities are expected to exceed
the allowable velocity for the given soil conditions.
Allowable flow velocities for different soil types are presented in Table 3.7.
51
Design Procedures
Table 3.7 Maximum allowable flow velocities for open soil (non-vegetated)
low flow channels
Soil description Allowable flow velocity (m/s)
Extremely erodible soils 0.3
Highly erodible soils (black earth, fine surface texture soils) 0.5
Moderately erodible soils 0.6
Low erodible soils (krasnozems, red earth) 0.7
Sandy soils (Mannings n = 0.04) 0.45
Fine colloidal sand (n = 0.02) 0.45
Sandy loam, non-colloidal (n = 0.02) 0.5
Alluvial silts or silt loam, non-colloidal (n = 0.02) 0.6
Fine gravel or firm loam (n = 0.02) 0.7
Graded loam to cobble, non-colloidal (n = 0.03) 1.1
Alluvial silts, colloidal (n = 0.025) 1.1
Stiff clay, very colloidal (n = 0.025) 1.1
Coarse gravel, non-colloidal (n = 0.025) 1.2
Graded silts to cobbles when colloidal (n = 0.03) 1.2
Loose rock, nominal size around 200 mm 1.5
Cohesive soils Lean clayey soils Heavy clayey soils
Loose 0.34 0.46
Fairly compact 0.6 0.7
Compact 0.9 1.0
Very compact 1.2 1.5
52
10000
8000
6000
4000
2000
1000
800
600
400
m
m
ll -
200 nfa
ai
lr
100 nua
80 An
0
Discharge (l/sec)
0 0
90 80 60
0
60
00
0
70
00
0
50
10
50 40
11
00
40
12
00
13
30
00
14
00
15
20
10
8
6
5
4
3
2
1.0
0.1
10000
1000
300
500
600
400
4000
6000
100
200
2000
20
60
80
40
50
10
30
5
6
8
Area (ha)
Figure 3.12 Graphical method for determining trickle flow discharge for area and annual
rainfall (Source: Aveyard)
53
Design Procedures
Step 11 Incorporate channel features
The design procedure presented in this chapter focus primarily on developing a stable
channel geometry. However, it is important to realise that the dimensions developed
within this procedure are average dimensions only. The final layout of the channel must
incorporate a random nature into the plan form and cross section so as to develop a
natural appearance and more importantly, habitat diversity.
An important aspect of habitat diversity is the incorporation of channel features as

described in Appendix A. At this point in the design procedure it is essential that the
various issues raise in Appendix A are reviewed and where appropriate, incorporated into
the final channel design.
Where practicable, the incorporation of channel features should be done in partnership

with local experts and interest groups experienced in local ecology, aesthetics, hydraulics
and community issues. Such groups could include the local frog society, Landcare and
catchment care groups, environment and habitat officers from local authorities and
fisheries groups.
54
4. Revegetation
Introduction Riparian vegetation can contribute many benefits to a watercourse, including erosion
control, bank stability, buffer zones, a food source, the control of light and heat, the
provision of shade and shelter, the management of unwanted aquatic plants and the
provision of essential habitat.
Riparian vegetation plays an important role in stabilising the stream banks and preventing
bank erosion. Bank vegetation decreases water velocities near the bank and dampens
turbulence by suppressing eddies. However, to be effective, the vegetation must extend
to at least the low water level, otherwise flow will undercut the root zone.
Grasses and sedges are effective at both low and high velocities, being capable of
withstanding much higher flow velocities than woody species such as trees. Plant roots
also increase the shear strength of the bank soils.
Riparian and floodplain vegetation acts as an effective buffer between developed lands
and their associated watercourses. Riparian vegetation functions as a source of leaves and
small and large woody debris. However, in urbanised areas of Brisbane, street tree
plantings include deciduous trees such as Jacaranda (Jacaranda mimosaefolia) and
Golden rain tree (Koelreuteria elegans). Widespread planting of deciduous species and
the reticulated stormwater systems in urban areas consequentially result in a significant
increase in the volume of leaf litter being transported into Brisbane's watercourses.
Shading produced by trees assists in the control of heat and light and can also be used as
a management strategy to control the growth of aquatic plants. However, it is noted that
it may take five to ten years for a canopy to be developed over a creek. Obviously this
depends on the tree species and the width of the creek. During this canopy development
time extra maintenance such as weed control may be required.
It is important that local politicians, interest groups and the local community are all aware
that during this often long canopy development phase the revegetated channel may look
messy and weed infested. Natural Channel Design should always be seen as a long-term
management technique, not a short-term fix.
The provision of habitat by riparian vegetation is a key benefit for aquatic and terrestrial
organisms. The vegetation may be of periodic importance as a refuge habitat during
occasions of environmental adversity as well as function as corridors for wildlife movement
between forest remnants (Arthington and Catterall 1990).
55
Revegetation
Interaction with Natural In Natural Channel Design, the long-term stability of the channel is primarily related to the
Channel Design suitability of the channel geometry to the given hydrological conditions. In major
watercourses, such as river systems, the channel and overbank vegetation may only play a
minor role in the long-term stability of the channel. However, in minor creeks and streams,
vegetation can significantly influence both the short-term and long-term stability of a
watercourse.
Natural Channel Design is primarily concerned with two aspects of vegetation; the
revegetation phase immediately following the channel construction, and the long-term
maintenance of channel and floodplain vegetation. The key to the appropriate integration
of a natural channel into an existing developed or urbanised valley is the development of
a watercourse that has long-term stability with minimal maintenance requirements, i.e. self
maintaining.
Watercourse maintenance usually relates to the following factors:

(i) weed control
(ii) flood control
(iii) fire control
(iv) habitat management and conservation
Occasionally, vegetation maintenance is required for human safety and pest control reasons.
Relating vegetation Many difficulties exist when trying to relate the desired planting densities to the hydraulic
density Mannings engineering roughness ( termed Mannings n or Mannings roughness). To assist in this matter,
roughness reference may be made to Appendix C Mannings Roughness, in particular, Table C.5
Significance of There are basically five types of vegetation that can be used in and along a watercourse.
vegetation type Each of these forms of vegetation have different characteristics that affect soil erosion and
water flow in different ways. To design and maintain the vegetation along a watercourse it
is important to understand the features of each form of vegetation.
The basic vegetation types are listed below in Table 4.1
Table 4.1 Vegetation types and characteristics
Vegetation type Erosion control Bank stability Hydraulic effects
Aquatic plants Provide good stability to Can assist bank stability Usually cause little flow
the low flow channel by protecting the toe resistance if the water depth
of the bank is greater than the plant
height, i.e. plant height
These plants can become is less than the bank height.
inflexible as plant density
increases. This can cause Thick stands of reeds can
channel flow to be deflected effectively block a channel
into the channel bank and aggravate upstream
causing bank erosion. flood levels.
56
Table 4.1 Vegetation types and characteristics (cont)
Vegetation type Erosion control Bank stability Hydraulic effects
Ground covers The most effective form Usually ineffective in Generally have little effect
of soil erosion control. the provision of bank on flood levels.
These plants control only stability
soil scour (erosion of the Some plants, such as
surface layer), not the mass These plants usually have Lomandra, can grow to a
movement of soil resulting a shallow root system and height of around 1 metre,
from bank failures. thus can only provide stability and thus may choke small
to the surface soil layer. channels.
To be effective, ground cover They can help to stabilise
plants should be flexible and the bank during the early
continuous. Isolated, clumped stages of revegetation
plants can aggravate soil
erosion.
Plants with a matted or
fibrous (hairy) root system
are the best, especially in
sandy soils.
Shrubs and Can provide effective erosion These plants can significantly These plants have the
woody weeds control if the branches increase bank strength greatest potential to affect
prevent high velocity water depending on the height the hydraulics of the
from contacting the soil. of the bank and the depth watercourse and increase
of the root system upstream flood levels.
Soil erosion can occur around
the edge of isolated plants Unlikely to prevent Avoid the planting of
caused by flood waters undermining of the bank shrubs in areas where
accelerating around the unless the shrubs are located flood control is important.
plant. close to the toe of the bank.
Single trunk trees Usually provide little Trees provide the main Grouped trees can
protection against soil form of bank reinforcement. significantly affect flood
erosion. levels if their spacing is less
They are needed to than 5 times their trunk
Some plants have root stabilise the bank, especially diameter
systems that survive when when toe erosion occurs and
exposed to air. The root when the bank becomes Generally well-spaced
system of these plants saturated during a trees with branches above
can control toe erosion. flood event. the flood level provide little
hydraulic interference.
Multi-trunk trees As for single trunk trees As for single trunk trees Grouped trees can significantly
affect flood levels.
Well-spaced trees with

branches above the flood level
can still provide significant
hydraulic interference.
57
Revegetation
Channel vegetation There are limits to the role that vegetation alone can play in controlling erosion. Although
vegetated watercourses in the natural environment appear to be stable and experience
extremely low erosion rates, it should be noted that these conditions of stability have
evolved over many years.
The long-term objective of vegetation as an erosion control measure is the establishment

of a ground cover that will be self-maintaining and thus be able to provide long-term
(sustainable) erosion control. Ideally, plants should be native to the area, must be good
soil binders, crowd out weeds, and form a good ground cover. Unfortunately, in urban
areas, some of the most successful erosion control plants are weeds!
Channel vegetation can be divided into four categories (Figure 4.1):

(i) in-stream or aquatic vegetation
(ii) toe vegetation
(iii) middle bank vegetation
(iv) upper bank vegetation
Medium sized plants with good

root systems and larger canopies
which shade the stream
Low-growing, multi-trunked Larger trees with deep

plants with matted roots to bind root systems.
the toe. Best species for
erosion control.
Stream
Figure 4.1 Categories of channel vegetation
Species such as reeds and sedges, with a dense network of fibrous or matted roots, are
more efficient in the control of soil erosion than those with a sparse network of woody roots.
If trees are removed from a steep or high bank, then appropriate consideration needs be
given to the long-term stability of the bank. It may take around five years before the old
58
tree root system within the bank begins to deteriorate. As the tree roots lose their
strength, sheer stresses in the bank can fracture the roots resulting in sudden bank failures
(land slips).
If, due to flood control reasons, the trees or shrub that were located on or near the
watercourse bank cannot be replaced, then the bank may need to be benched or batter at
a flatter grade to compensate for the long-term removal of the essential root
reinforcement of the bank.
Floodplain vegetation The impact of trees on the hydraulic roughness of a floodplain depends on the flow
velocity; the shape and size of the trunk and canopy (if below flood level); and the
number, arrangement, and spacing of trees. When the flow velocity is high, an
obstruction such as a tree exerts a sphere of influence that is much larger than the width
of the obstruction because the obstruction affects the flow pattern for considerable
distance on each side.
The sphere of influence for flow velocities that generally occur in channels that have
gentle to moderately steep slopes is about three to five times the width of the
obstruction. Therefore, if the trees are spaced more than five times there truck diameter,
then they gradually begin to act as independent obstructions. At a spacing of around ten
times the trunk diameter the trees may be considered as totally independent obstructions.
The impact of trees on flood levels depends on the depth of water. Riparian trees located
in relatively deep water along the edge of the channel can have a much greater impact on
flooding than trees planted along the outer edges of the floodplain.
A summary of the hydraulic requirements for trees in floodplains is provided in Table 4.2.
59
Revegetation
Table 4.2 Vegetation in floodplains
Vegetation type Floodplain planting requirements
Ground cover plants Suitable in most areas of a floodplain.
Shrubs and woody weeds Avoid in flood control areas.
May be used in backwater areas and in areas of tree grouping where flood flows are
mainly designed to flow around the vegetation not through it.
Trees In flood control areas use smooth, single trunk trees with branches above the flood level.
Tree spacing less that 5 times the trunk diameter:

Considered as group plantings. Likely to have high restrictions to flood flow. Trees
should be planted in rows parallel with the flow direction to minimise hydraulic affects.
Tree spacing between 5 and 10 time the trunk diameter:

Questionable benefit of planting trees in rows unless flood control is critical. Soil erosion
may occur around individual trees if located in high velocity floodplains. Erosion may be
controlled with selected planting of sedges around the base of the tree.
Tree spacing greater than 10 times the truck diameter:

No hydraulic benefit obtained by planting in rows. Soil erosion likely to occur around
individual trees if located in high velocity floodplains. Erosion may be controlled with
selected planting of sedges around the base of the tree.
Grouped trees Tree spacing less than 5 times the truck diameter, or trees at wider spacing surrounded by
shrubs.
In flood control areas, grouped trees are only suitable in backwater areas, or adjacent to
open, grassed floodplains where floodwaters can readily bypass the trees.
Planting patterns and In urban areas of Brisbane there may be conflicts between ecologically preferred
hydraulic constraints re-vegetation and flooding issues. This has lead to restrictions being placed on the
revegetation of some floodplains.
It is generally shrub species and multi-branch trees that have the greatest influence on
channel roughness and that may constrain the ecologically desirable planting pattern. The
following discussion illustrates some general principles to address these issues.
When making estimations of hydraulic roughness, it is necessary to include consideration

for regrowth of vegetation on cleared channel sides and overbank areas because this can
significantly increase the resistance factor within one or two growing seasons. Understorey
vegetation grows readily after removal of any shade-producing canopy.
A row of trees roughly aligned with the current can offer much less resistance to flow than
perpendicular blocks of vegetation yet retained much of the shade and visual harmony of
an uncleared bank. Further reduction in flow resistance may be obtained by pruning limbs
that protrude below normal flood height.
60
The desired planting pattern would closely relate to the natural vegetation community
structure for the area, with tall trees, an understorey of small trees and shrubs and a sparse
herb layer. Where the canopy does not over-hang the stream bank, reeds, rushes and
sedges could be planted.
In areas where hydrological constraints do not allow rehabilitation with dense vegetation,
alternative planting patterns are planned.
Brisbane City Council (1995) details that the spatial patterns of species distribution follows
two strategies. One strategy is planting parallel to the stream bank. This conforms with
plant-water relations and produces least impact on water flow. The second strategy aims
for planting in clumps with sparse connections between the clumps. Clumping vegetation
has the effect of increasing the number of habitats for wildlife and allows migration
between clumps. The location of these clumps will need to take account of the probable
high flow paths to minimise their hydraulic impact.
Plans for rehabilitation, while using these two strategies as fundamental to the plan, also
should consider the following points:
(i) Maximise the availability of habitat by considering the topography of the stream bank,
floodplain and surrounding bushland, e.g. Melaleuca wetlands could be planted on
lowlands as well as point bars or where the channel bank slopes gently into the
channel bed and would be subject to frequent inundation.
(ii) The stream morphology may constrain planting, e.g. in a meander, it may be
necessary to stabilise one bank with sedges and shade the opposite bank.
(iii) Problems due to ponding and appropriate selection of species for such areas.
(iv) Proximity to bridges. Trees with root systems which are susceptible to being undercut
such as Eucalypts, should not be planted in such locations.
(v) Proximity to stormwater drains. Discharge of stormwater requires a consideration of

water velocity and water quality. Use of a discharge structure and wetland filter may
be required.
(vi) Access for maintenance equipment.
Group plantings Group planting may be considered to be trees with a spacing less than 5 times the truck
diameter or trees at a wider spacing but planted amongst shrubs and other understorey
plants.
Grouped trees can provide significantly more ecological benefit than the equivalent
number of widely spaced trees located across the floodplain. If the watercourse channel
flows approximately parallel with the flood flow, then where hydraulically allowable,
grouped trees should be located within the riparian zone as well as the high bank area as
shown in Figure 4.2.
61
Revegetation
If the watercourse channel meanders across the floodplain, then care should be taken to
avoid grouped plantings in areas were flood waters pass from one side of the channel to
the other, as shown in Figures 4.3 and 4.4.
Floodplain vegetation should be selected and landscaped so as to require low

maintenance. This would include group placement and sufficient density of shrubs and
trees to avoid mowing. When grasslands are provided that require mowing, they should
be placed in large enough areas and interconnected to permit easy mowing.
Planting around major In critical flood control areas, vegetation that may interfere with flood waters should not
hydraulic structures be located within the following areas:
(i) upstream of a bridge or culvert within a radius equal to the total bridge or culvert
opening width;
(ii) downstream of a bridge or culvert within the zone defined by a 1 in 4 expansion of the
outlet jet and for a distance equal to three times the flood water depth;
(iii) between the bridge or culvert opening and the bypass floodplain;
(iv) any areas judged necessary by hydraulic modelling.
Access/bikeway
Section A-A
Dense buffers Modifed
at top of flood plain channel
A
A
Access/bikeway
Figure 4.2 Floodplain vegetation along a straight channel

62
Figure 4.3 Floodplain vegetation along a channel meandering
Figure 4.4 Floodplain vegetation densities along a meandering channel

63
Revegetation
Species for revegetation Given the previously identified constraints associated with riparian vegetation and the
in Brisbane effect on channel roughness and consequently flood frequency, the following section
provides an analysis of those characteristics of vegetation that are required as to have a
minimal impact on the effect of the roughness coefficient and subsequently the flooding
frequency.
The species described in the following section were assessed as having the potential for
use in waterway channel and floodplain rehabilitation projects in Brisbane. Assessment
was based on:
(i) the influence of vegetation on channel roughness and hydrology;
(ii) ability to withstand and rapidly recover from inundation by floodwaters and battering
by bed load and gravel;
(iii) ecological validity;
(iv) form of root system;
(v) ease of propagation by seed or division, and growth rates;
(vi) growth habit;
(vii) physiological adaption to flooding and to soil type;
(viii) appropriateness to various locations and problems within Brisbane; and
(ix) dominance of the species within the riparian zone.
Along undisturbed streams there is a diversity of riparian species present ranging from
groundcovers and shrubs to trees, all of which have differing habits and root structures
and contribute in various ways to stream stability. Riparian floristics also vary at
catchment-wide and local scales.
Whilst the majority of the species detailed are both common and widespread in Brisbane,
previous land use practices have severely restricted their current distribution. However,
species distribution needs to be considered before undertaking any rehabilitation work so
that those species best adapted to a particular site and indigenous to that part of the
catchment are used. Locally collected seed should be used, where practicable, to
maintain genetic integrity.
The following species are divided into principal locations within the riparian zone,
obviously species occurrence and vegetation community composition is dependant on
numerous more parameters than just proximity to mean water level within a given
waterway. For example, a floodplain area may be predominately composed of Melaleuca
woodland community with patches of standing water surrounded by Carex and Juncus
species or alternatively a closed forest notophyll-type community.
64
Species of vegetation which have been determined as suitable for planting on the bank
toe (Figure 4.1) within Brisbane's Waterways are detailed in Tables 4.3 & 4.4.
The species of vegetation which have been determined as suitable for planting on the
middle bank area (Figure 4.1) are detailed in Tables 4.5 & 4.6.
The vegetation suitable for planting on level ground at the top of bank and the upper
section of the bank (Figure 4.1) are detailed in Tables 4.7 & 4.8.
The vegetation suitable for planting on lowlands where flooding is expected and water
may be retained are detailed in Tables 4.9 & 4.10.
Whilst the key objective of waterway rehabilitation projects is to establish a mixture of

vegetation forms i.e. ground covers, climbers, shrubs and trees to give maximum
structural diversity, this is not always achievable in flood sensitive locations. In general,
planting schemes should be kept simple with a view to establishing ground cover/herb
layer and a dense tree canopy.
Table 4.3 Emergent/Herb Layer plants suitable for planting on bank toe
Carex appressa Philydrum lanuginosum
Crinum pedunculatum Themeda triandra
Echinochloa telmatophila Triglochin procera
Juncus usitatus Triglochin striatum
Lomandra longifolia
Table 4.4 Trees suitable for planting on bank toe
Acmena smithii Leptospermum polygalifolium
Callistemon viminalis Waterhousia floribunda
Table 4.5 Herb Layer plants suitable for planting on middle bank area
Crinum pedunculatum Themeda triandra
Lomandra longifolia
65
Revegetation
Table 4.6 Trees suitable for planting on the middle bank area
Acmena smithii Ficus fraseri
Alphitonia excelsa Ficus macrophylla
Aphananthe philippinensis Flindersia australis
Araucaria cunninghamii Flindersia bennettiana
Argyrodendron trifoliolatum Flindersia schottiana
Castanospermum australe Grevillea robusta
Casuarina cunninghamiana Jagera pseudorhus
Cryptocarya glaucescens Mallotus philippensis
Cryptocarya triplinervis Melaleuca bracteata
Dissiliaria baloghioides Melaleuca quinquenervia
Elaeocarpus obovatus Melia azedarach
Eucaltptus microcorys Polyscias elegans
Eucalyptus siderophloia Toona australis
Eucalyptus tereticornis Waterhousia floribunda
Table 4.7 Herb Layer plants suitable for planting on level ground at the
top of bank and upper section of bank
Cymbopogan refractus Lomandra longifolia
Dianella caerulea Themeda triandra
Table 4.8 Trees suitable for planting on level ground at the

top of bank and upper section of bank
Alphitonia excelsa Eucalyptus tereticornis
Aphananthe philippinensis Flindersia australis,
Araucaria cunninghamii Flindersia bennettiana
Argyrodendron trifoliolatum Flindersia schottiana
Castanospermum australe Grevillea robusta
66
Table 4.8 Trees plants suitable for planting on level ground at the
top of bank and upper section of bank (cont)
Casuarina cunninghamiana Hymenosporum flavum
Corymbia citriodora Jagera pseudorhus
Corymbia intermedia Lophostemon confertus
Corymbia tessellaris Lophostemon suaveolens
Cryptocarya glaucescens Mallotus philippensis
Cryptocarya triplinervis Melaleuca quinquenervia
Dissiliaria baloghioides Melia azedarach
Elaeocarpus grandis Polyscias elegans
Elaeocarpus obovatus Toona australis
Eucalyptus microcorys Waterhousia floribunda
Eucalyptus siderophloia
Table 4.9 Herb Layer plants suitable for planting on lowlands,

where water may be retained
Carex appressa Themeda triandra
Crinum pedunculatum Triglochin procera
Cyperus difformis Triglochin striatum
Echinochloa telmatophila Philydrum lanuginosum
Juncus usitatus Scirpus mucronatus
Lomandra longifolia
Table 4.10 Trees suitable for planting on lowlands, where water may be retained
Callistemon salignus Lophostemon suaveolans
Eucalyptus propinqua Melaleuca bracteata
Eucalyptus tereticornis Melaleuca quinquenervia
Leptospermum polygalifolium
67
Revegetation
Plants used for Some herbaceous species within Brisbane can provide erosion protection by forming a
erosion control mat-like root system that physically covers creek banks. Individual species are suitable for
differing locations where the plants have access to groundwater. Table 4.11 details those
herbaceous species that have been recorded in Brisbane waterways with these
characteristics.
Table 4.11 Herbaceous species with attributes for bank protection
Carex appressa Lomandra longifolia
Cyperus difformis Phragmites australis
Juncus usitatus
The matrush Lomandra longifolia is a very effective stabiliser. It grows in clumps and has a
dense branching rhizome system which acts as a soil binder and promotes soil stability.
The matrush is hardy, both in direct sunlight and shade and can be planted to prevent soil
erosion at the mean water line on the steepest parts of banks and on mid slopes. It is
generally planted at one metre centres, or at half metre centres in critical locations. It
regenerates prolifically in moist sites. Rushes (Juncus usitatus) and sedges (Carex
appressa, Cyperus difformis) are the most useful species in streams which change height
quickly and which do not have a wide range of continuous flows, such that most of the
marginal vegetation is not submerged for long.
Woody vegetation A common misconception is that trees prevent creek erosion by binding the soil particles
together, when it is well known that tree roots are easily exposed if subject to medium to
high velocity flows.
Trees typically only stabilise a creek by providing structural strength to the banks.
However, when masses of weather-resistant tree roots are exposed these roots can prevent
high velocity flows from reaching the underlying earth bank. In these cases the trees and
shrubs do provide significant erosion protection for the creek banks.
Successful bank stabilising tree species have root systems that can withstand exposure
without drying out and capable of forming a dense mat over the creek bank as a physical
erosion barrier. The roots must also be long enough to pass below the level of active
bank erosion. Table 4.12 details tree species which have been recorded as exhibiting
these characteristics which and which are endemic to the Brisbane local area.
68
Table 4.12 Tree species with attributes for bank protection
Acmena smithii Elaeocarpus obovatus
Aphananthe philippinensis Ficus fraseri
Callistemon viminalis Ficus macrophylla
Castanospermum australe Leptospermum polygalifolium
Casuarina cunninghamiana Waterhousia floribunda
Cryptocarya triplinervis
Trees greatly influence the stability of creeks not subject to meandering or changing
catchment conditions, but they have only limited long-term influences on creeks
responding to changing catchment conditions caused by urbanisation.
Shrubs and trees are valuable for erosion protection especially when they can be planted
densely. When planted sparsely, they can result in enhanced erosion due to eddying
effects. Thin trunked species such as River Lilly Pilly (Acmena smithii), Weeping
Bottlebrush (Callistemon viminalis), or similar species will reduce the eddying effect. A
recommended density would be of the order of one tree/shrub every one to four square
metres (Raine and Gardiner, 1995). Although trees and shrubs can sometimes be used for
protection of actively eroding sites regularly exposed to flow (Table 4.12), they are
generally more valuable for strengthening the bank and protecting it from collapse.
Some species such as River Oak (Casuarina cunninghamiana) have strong fibrous root
systems and can provide excellent bank stabilisation. They produce a dense mulch from
needle drop and are generally hardy but may be restricted by extremely wet conditions.
The toxic effect of the needle mulch may however suppress ground cover.
Callistemons and Melaleucas can be used on the lower slopes of banks from just above
mean water level. Hard Quandong (Elaeocarpus obovatus) and Tallowwood (Eucalyptus
microcorys), once established along a waterway regenerate rapidly and provide the
foundation for highly stable tree cover higher up the bank. River Bottlebrush such as
Callistemon viminalis and C. salignus can grow when flooded. C. viminalis is a riparian
zone species, often found with the root zone partly submerged. C. salignus grows on
higher ground and tolerates flooding. The bottlebrush root system is quite dense and is
effective in protecting the bank. Melaleuca quinquenervia and Eucalypt species are not
suitable for planting on actively eroding gully sites as their root systems may be
undermined. Undercutting may be prevented if the bank is stabilised with herbaceous
groundcover type vegetation such as Lomandra longifolia.
69
Evaluation and Monitoring
5. Evaluation and Monitoring
Introduction The functionality of rehabilitated or constructed channels using Natural Channel Design
principles is not as predictable as conventional channel systems. Risk of failure in the
rehabilitated channels can be high, so appropriate evaluation of the success or failure of
the NCD should be an integral part of the whole project.
NCDs attempt to achieve specific ecological and social outcomes. The evolution of this
design technique is absolutely dependent on how we monitor and evaluate the expected
outcomes at various stages during and after construction. We need to learn from our
successes and failures. Natural systems are so diverse and complex that often it is hard to
explain what caused the success; however, failure can be relatively easier to explain
(LWRRDC publication, A Rehabilitation Manual for Australian Streams, Vol. 1, March 2000)
Evaluation is often avoided in stream rehabilitation projects simply because of the

complex nature of the natural system and its very slow to response to changes leading
long and costly evaluation processes. Evaluation needs firm commitment from the
stakeholders who must appreciate its need and be able to comprehend the benefits of
evaluation.
Often the term Evaluation is confused with Monitoring. Monitoring comprises collection
of information (eg data, samples, visual aspects) whereas evaluation means assessment of
the monitored information to decide whether the project is a success or a failure. A well-
planned evaluation task can even rescue a rehabilitation (or NCD) project from ultimate
failure. That means evaluation at various stages can provide useful data for the designers
and planners to derive alternative approaches and take necessary actions to minimise
unexpected incidents. A good evaluation plan consists of detailed information on:
What to measure?
How to measure?
When to measure?
How often and how long do we need to measure?
What recording and analysis techniques do we need to use?
What are the benefits and learnings at various stages?
What is the likely cost of evaluation?
(Note: the concept and details of this topic have been derived from the LWRRDC
publication A Rehabilitation Manual for Australian Streams, Vol. 2, March 2000).
Planning the Evaluation Table 5.1 provides a sequence of evaluation tasks.
70
Table 5.1 Planning the evaluation of an NCD or rehabilitation project

(LWRRDC publication A Rehabilitation Manual for Australian Streams, Vol. 2, March 2000)
Task 1
What are the project objectives?
Task 2
What type of evaluation do we need?
Task 3
How confident do we need to be?
Task 4
What level of evaluation do we need?
Task 5
What are the project objectives?
Task 6
How frequently should we monitor?
Task 7
How long do we need to monitor?
Task 8
Who is responsible for the evaluation?
Task 9
What recording technique will we use?
Task 10
How are we going to analyse the information?
Task 11
How much will this cost?
71
Task 1: Defining project It is important to be clear on what needs evaluation. This is directly linked with the
objectives objectives that are set for the NCD (or rehabilitation) project. Objectives should be as
specific and measurable as possible. For example instead of stating improve fish habitat
consider a measurable target, eg a particular species of fish population within a specified
reach to be increased by 50% in 5 years. This provides a clear idea of what to monitor and
whether the monitoring data shows any trend (evaluation) to reach the objectives.
Objectives need a reality check based on evidence, existing data or local knowledge.
Monitoring of existing conditions (physical, chemical, ecological etc) is extremely
important. A good evaluation plan should therefore include monitoring of data even prior
to setting clear targets. This ensures that the objectives set have a reasonable chance of
success.
This guideline has been developed for urban waterways and creek systems. Designers
and planners must take into consideration the fact that urban waterways and creeks have
almost no chance of being rehabilitated (or reconstructed) to their pristine (or pre-
urbanisation) state. Even if we spent millions of dollars to transform the system into
pristine look-like conditions, the channel will still take its own form by adjusting to the
heavily modified urban hydrologic and hydraulic regime.
Rehabilitation and NCD exercises are attempts to do the best we can to establish a
sustainable ecosystem for our urban waterways in the long term. There is no point in
trying to enhance or establish an ecological feature when there is clear absence of such a
feature within or near the target section of the stream.
Task 2: Deciding the type The LWRRDC Publication A Rehabilitation Manual for Australian Streams Vol. 2 March
or evaluation 2000, refers to five output, or outcome driven evaluation types. These are:
(1) Project Execution (output driven): This is simply to monitor whether the works have
been completed according to the plan and design.
(2) Project Survival (output driven): Whether the project has survived the design floods
for example. This also identifies the obvious causes of failure, seepage destroyed the
structure because geotextile was not used or the pools are drying because the
system doesnt have a constant trickle or base flow.
(3) Aesthetic (outcome driven): Stream rehabilitation works are commonly related to
aesthetics, ie whether the rehabilitated streams look good. Photographic evidence of
the rehabilitated system over time is a good way to monitor aesthetics.
(4) Physical and chemical (outcome driven): The ecological condition of the stream is
often improved by the physical and chemical changes ie the hydraulic habitat, scour
holes, bed and bank formation, water quality etc.
(5) Ecological changes (outcome-driven): Evaluation of ecological changes is not as easy

as other types mentioned above and often involves much longer timeframes and
higher costs. Ecological improvement is the vision driving most of the rehabilitation
work. Evaluations most commonly measure types, abundance, diversity, etc.
72
Task 3: Building confidence The confidence-building exercise should focus on:
How confident the designers and planners are that the planned project will work:
No one can be completely confident when it comes to rehabilitation and NCD projects;
its all about the level of confidence. Rigorous evaluation is needed to determine if the
project was a success (or failure). Designers and planners need to consider their desired
level of confidence, the associated costs and available resources in determining if the
project is affordable.
Who are we trying to convince with the intended evaluation:

The evaluation needs to address the objectives to a level necessary to convince the
appropriate stakeholders, eg:
Media
Politicians
General public
Funding agencies
Scientific community
Yourself as a designer or planner
Different stakeholders will be persuaded in different ways, so your evaluation strategy may
vary according to your audiences needs. A number of examples of who to convince
about which element of a project are provided in the LWRRDC publication A
Rehabilitation Manual for Australian Streams, Vol. 2: March 2000, P.191.
Task 4: Determining the Basic sampling designs are used in A Rehabilitation Manual for Australian Streams, Vol 2,
level of evaluation March 2000, pp.192,195 for detailed physical or ecological evaluation, level of confidence
and effort. These sampling vary significantly in cost from project to project. The following
activities lead to various levels of confidence:
1. Observation after rehabilitation (anecdotal) very low to low level of confidence*

(sustainability issue in the long run)
2. Sampling after rehabilitation low to moderate level of confidence*, need to undertake

the sampling over a longer time** to ensure sustainability of the outcome
3. Sampling before and after rehabilitation works high to very high level of confidence*
* Confidence to assure success or failure

** To be discussed in the How long to monitor and evaluate item
Task 5: Deciding what to The main purpose of evaluation is to check whether the project met the objectives set in
measure the first place, hence the measurements should relate to the project objectives. For
example, if an objective is to increase or bring back the frog population, a survey of the
number of frogs or frog species could be conducted (preferably before and after). Some
common measurement techniques for the five different types of evaluation mentioned
earlier are outlined below:
73
Execution outputs
Need to measure whether the project has been built as per the design and construction
plan. For example whether the pools and riffles were constructed with appropriate
spacing, depth, height etc and whether the specific vegetation species have been planted
at the designed spacing and locations. In other words this is an inventory check of the
constructed features.
Survival outputs
This is an extension of the execution to see if the various elements of the project have
survived certain conditions. The aim of this task is to ensure that the structural elements
are functioning as designed.
Aesthetic outcomes
Photographic records can be used to assess the aesthetics before and after rehabilitation
works. Photographic records of similar sites can also be useful in comparing the success
or failure of the project. It is important for photographs to be taken from the same
location each time.
Physical and chemical outcomes

The success of rehabilitation projects lies with the long-term sustainability of the
ecological diversity sought by the project. The success of the project is dependent on the
attainment of the physical and chemical changes as designed. It is assumed that if the
physical changes take place, then the ecological change will follow. This is not necessarily
the case. Evaluation helps to determine if ecological improvements have occurred.
Physical changes can be measured by:
- Channel morphology (average depth, channel slope, variation of depth along the
stream taking into account pools and riffles, shape of the channel, channel widths etc),
- Sedimentation and erosion characteristics (monitoring short-term erosion problems
after construction),
- Influence due to structural features (eg woody debris, snags etc),
- Hydraulic parameters (eg velocity, flow depth)
Chemical changes can be measured by:
- water quality parameters (eg various levels of nutrients, turbidity, dissolved oxygen,
temperature, etc)
Ecological outcomes
The main aspect of this task is identification of the flora or fauna listed in the project
objective.
The presence, appearance or reappearance of the ecological habitats must satisfy an

appropriate taxonomic level. This task can demand considerable experience, depending
on what is being measured. For example, plant species can be easily identified, whereas
the macroinvertebrates may not be easily detectable. Fact Sheets produced for
74
Waterwatch (Department of Natural Resources, Qld) or "Streamwatch" (eg CSIRO and

Sydney Water) may be helpful in identifying common species (eg dragonfly, leech, beetle,
caddis-fly larva, flatworm etc).
Fish species can be captured by careful use of nets. Platypuses, birds and frogs can be
surveyed by careful observations at appropriate times of the day. Brisbane City Councils
Ecological Assessment Guidelines, 1998 provide some general directions on undertaking
flora and fauna surveys. It is important to note that abundance and diversity are two key
elements in measuring ecological outcomes. Look for the behaviour of the animals, eg are
they present around the new structures?
A range of tools (eg AusRivas, models based on empirical data and complex statistical
analysis) are available to analyse data sets.
Task 6: Deciding how Sampling time and period may be decided by an event-based strategy or some
frequently to monitor predetermined sampling frequency. Event-based sampling is particularly important for
structural works, eg to determine whether the rehabilitated site has survived a flood that
exceeded the design frequency or the pools are scoured much deeper than the anticipated
depth. Event-based monitoring is also important for water quality indicators, when the aim is
to assess the impacts on the biological habitats of storm-based nutrient levels within the
rehabilitated reach of the stream.
Ambient water quality data can be collected at a predetermined frequency. Predetermined

sampling frequencies are not linked to the flood events but may be based on the design
stages and the expected appearance or reappearance or growth of biological features. Two
important questions need to be answered before deciding the predetermined frequency.
Firstly, what time of the year is of importance? (the answer to this question may depend on
how the fauna and flora respond to the rehabilitation works during summer or winter).
Secondly, do we need to monitor every year? (eg different macroinvertebrates are often
observed in different seasons). If monitoring vegetation, spring may be the best time
because plants are easier to identify when flowering.
Table 5.2 shows some common rehabilitation activities, measurable objectives, and suggested
frequencies and duration of sampling. Table 5.3 shows a summary of evaluation projects, their
key measures and the frequency and duration of sampling. Both of these tables are just
examples to provide a feel for the evaluation tasks involved in stream rehabilitation and NCD
projects and these are extracted from LWRRDC publication, A Rehabilitation Manual for the
Australian Streams, Vol. 2 March 2000.
Task 7: Deciding how long The Evaluation Plan should outline the target duration for monitoring the desired
to monitor outcomes/outputs. The key question is how long will it be before we can expect some
sustained response in the variables we are measuring? Tables 5.2 and 5.3 provide some
ideas on duration of sampling. The following are considered important in setting duration:
Type of evaluation
Different types of evaluation will require different durations. For example execution of the
project as per the design can be checked as soon as the construction is completed or
75
even during construction in progress. To evaluate the aesthetics, time should be allowed
for the trees to grow. Physical changes may take a long time, particularly if design floods
need to be monitored to assess the impact on structures. Biological outcomes may take
even longer because they are dependent on the physical changes.
Lag between the works and response (expected outcomes)

The rate at which the physical and biological systems respond to rehabilitation will depend
partly on the flow regime. This means the physical recovery may be delayed until a flood
of sufficient magnitude has occurred, and biological recovery may be slow until the
minimum flow requirement of the species in question is met. Recovery may be slow at the
start but will gain momentum with time.
Some biological responses may not occur until a considerable period after the
rehabilitation work has been completed, then however with the right environment the
population may grow at a significant rate.
The success of the stream rehabilitation and NCD works can only be assessed through
effective long-term evaluation which includes sound baseline information.
Sustainability of the response

Fluctuations in responses are commonly observed in rehabilitation works. Long-term
sustainability is dependent on a number of factors, eg abundance and growth rate of
biological species in the new habitat (eg pool-riffle system) after construction. The success
may be short-lived due to ongoing sedimentation and nutrient deposition. Ample time
should be allowed to monitor trend and assess sustainability. The duration of the
monitoring will vary between projects.
Task 8: Responsibility Monitoring can be undertaken by a range of stakeholders including the community
groups, eg Waterwatch program. The sustainability of rehabilitated waterways and
constructed natural systems is dependent on all stakeholders taking the ownership of the
project. Therefore, it would be ideal to get everyone involved in monitoring. This is cost-
effective and can generate considerable interest in the community in providing support
and maintenance of projects in the longer term.
It is important to note that the people who designed and constructed the project should
be closely involved in evaluation. During the design stage designers should take into
consideration community involvement in monitoring and data collection. The analysis of
data gathered during the monitoring stage should be undertaken by experts and the
results related to the projects objectives. The experts undertaking the analysis should be
experienced enough to identify key elements of the project that is not functioning as
designed and be able to develop strategies to address any issues.
Evaluating the number of fish larvae in a stream is a highly specialised task. Catching the
larvae can be tricky, but amateurs can learn to use the equipment needed to sample a range
of habitat types. It is important to get the community involved where appropriate, and
provide necessary training to collect samples and other data. Identifying larvae and analysing
results is a task for the experts.
76
Table 5.2
Typical rehabilitation activities, the sorts of measures which may be used to evaluate the
activity, and an adequate frequency and duration of sampling to determine the response.
Rehabilitation Objective Measure of How frequently do How long do
activity Response you sample? you sample for?
Before After
Rehabilitation Rehabilitation
Riparian Closed canopy Canopy cover Once every two 1 year 10 years
vegetation (tropics) Green Survival or three years
density of Species present
surviving trees Presence of
Self-regenerating seedlings
strand
Re-snagging More fish, more Survey before and Seasonal 1 (eg. 2 years 3 to 5 years
diverse after, and control Spring and autumn)
macroinvertibrate reach
Rock riffles More diverse fish, Surveys before and Seasonal 1 (eg. 2 years 3 to 5 years
more diverse after, and control Spring and autumn)
macroinvertebrates reach
Small weirs Create pool-riffle Survey thalweg Survey physical 2 years 5 years
sequence; increase cross-sections, flow habitat before and
fish and diversity, depth, bed after works, then
macroinvertebrate material, fish and after 2 year flood
diversity macroinvertebrates survey biota
Bypass of fish Increase in fish Fish passing up fish Seasonal1 2 years 3 years
barrier population above barrier or survey (when fish are
barrier of population up migrating past
and downstream the barrier)
Grade control Stabilise bed so Survey thalweg After 5 year flood 1 year 10 years
structure no further and cross-sections and at 10 years
incision occurs
Erosion Reduced erosion Works survive Floods greater than 2 years 10 years
control works rate to that of a erosion pin 1 year return
template reach measurement interval
Reinstate cuttoff Reduced erosion Flow velocity Floods greater 2 years 5 years
bends and low velocity erosion pin than 1 year
reduced bank measurement return
erosion interval
Sand extraction Fall in bed return Cross-section Annual or at five 5-10 years
of bed surveys year flood (depends on
complexity longitudinal extraction rate,
surveys bed size of stream
material and supply rate)
composition
Note [1] It is important to sample at the same time of year.

Ref: LWRRDC A Rehabilitation Manual for Australian Streams Vol.2; March 2000
77
Once the data has been analysed and conclusions drawn, the recommendations must be
communicated to the community in simple and transparent terms so the community has an
appreciation of the results. This may even inspire them to put more effort in data gathering
exercises as they see the contribution they make to the sustainability of the waterway.
Table 5.2 left shows typical rehabilitation activities, the types of measures which may be used
to evaluate the activity, and suggested frequency and duration of sampling to determine
responses.
Table 5.3
Duration and frequency of sampling from some rehabilitation projects
STUDY MEASURE FREQUENCY DURATION

BEFORE REHAB AFTER REHAB
Koehn (1987) Fish surveys Once before, once 2 months 3 years
after rehabilitation
Newbury and Trout eggs Annual None 6 years

Gaboury (1993)
Mallen-Cooper Fish numbers passing Annual 2 years 2 years

et al. (1995)
Shields et al. Bed and bank stability cover of Variable None Variable. Up to
(1995c) vegetation 10 yrs
monitoring,
with up to 8 yrs
casual
observations
after that
Shields et al. Fish species composition and abundance Twice yearly 2 years 1 year
(1995c) physical habitat (pool area, heterogeneity,
riparian vegetation, shade, woody debris)
Frissell and Condition of stream structures After a 2 10 None Once

Nawa (1992) year flood
House and Stability of structures channel morphology Annual Shortly before 2 years
Boehne (1985) fish utilisation of habitat juvenile fish rehabilitation
density and biomass
House (1996) Habitat diversity juvenile fish populations Habitat was measured 1 year 11 years
spawning sites gravel quality in year 1, 3 and 5
other measures taken
annually 1 year 11 years
Hunt (1976) Number and size distribution of trout Annually 3 years 7 years
biomass
Ref: LWRRDC A Rehabilitation Manual for Australian Streams Vol.2; March 2000
78
Task 9: Recording The ability to easily store and retrieve data is a key issue to be considered when planning
techniques for evaluation. This is particularly important when a number of stakeholders and
community groups are involved in the monitoring exercise. Communication between all
parties and sharing information are also important. Many evaluation exercises are never
completed because the data gathered was not continuously recorded, or recorded in
different formats in different locations, or some of the data is lost. There are a few key
rules when it comes to data collection and reporting:
Always prepare your own proforma recording sheet with a space available for every
piece of information you require. Discuss the proforma with all the parties involved in the
monitoring exercise so that, if possible, everyone uses the same proforma or close to it.
Dont forget simple data like date and time!
Every space in the recording sheet must be filled out even with N/A. A blank space may
imply wasn't sure, or its obvious. In fact it probably is not obvious to someone who was
not there, or even yourself in 12 months time when the information is being collated.
Do not rely on your memory. Even the most obvious data must be documented,
because you might forget or even leave the project.
Task 10: Analysing the Two types of analysis of data are suggested in the LWRRDC publication, A Rehabilitation
information Manual for Australian Streams, Vol. 2, March 2000. These are:
Eyeballing the data

Statistical tests
Eyeballing the data

Examining the data by eye can be fast and effective, particularly for small data sets, and to
check if there have been some major improvements recorded as a result of the
rehabilitation. This step may involve some minor mathematical manipulation of the data,
such as calculating average values. For example it may be useful to compare the average
number of biological communities before and after the construction.
Counting the total number of communities to check if it is x times higher than before is a
good way of noting the inclining or declining trend and pattern of change in the desired
habitat. Another example of the eyeballing exercise may be to compare the profiles of
the reach before and after the riffles were constructed, with the profile straight after a 1 in
2 year flood. The results may indicate shallower riffles and deeper pools between the
riffles.
Eyeballing the results is not always easy as the data sets can be large and complex, and
require rigorous mathematical modelling using statistical procedures. For example a
relatively healthy stream may have 50 different species of macroinvertebrates.
Statistical Tests
Statistical tests can be complex, time-consuming and expensive. Careful collection of a
79
considerable volume of data is usually required. Statistical tests are accurate, take into
account a range of variables, and offer objectivity within a desired scale. For example if
the design water depth in a pool for the survival of a number of biological habitats is 250
mm, we can determine statistically whether the depth of water in the pool was greater
than 250 mm 90% of the time during the monitoring period.
Statistical testing requires us to decide how confident we want to be before we accept

that a project has succeeded. For example a 75% chance that a rehabilitation project
using a specific technique was successful may be acceptable to the stream manager. This
confidence level will then be used in the statistical analysis to check if the desired result
falls within this confidence limit. It is important to decide at the start of evaluation
planning process the method to use in analysing the monitored data sets. This decision
will influence the design of the monitoring program.
The decision on whether to use statistical analysis or eyeballing largely depends on what
is to be monitored and whether detailed analysis is required. If the response to the
rehabilitation is expected to be obvious and only a simple analysis is desired, then
eyeballing may be sufficient. If a large and complex rehabilitated waterway is to be
monitored for habitat diversity or macro invertebrate populations, then a statistical analysis
may be required.
Task 11: Cost Cost varies with the level of evaluation, the analytical approach and number of individuals
involved in the monitoring and final evaluation process. Evaluation costs include:
Buying or hiring of equipment

Labour (may be a significant component when expert assistance is required)
Laboratory analysis (water quality testing, particle size analysis, etc)
Evaluation of the data and regular reporting
Execution, survival and aesthetics can be monitored relatively cheaply, whereas monitoring
involving sensitive equipment can be very expensive, eg an electro-fishing backpack unit
can cost $8,000 to $10,000 and a full selection of nets and other gear for fish sampling
can cost about $4,000. Projects that require samples to be taken and analysed by
professional staff can be expensive too. At the time of project planning, it is essential to
build these costs into the project cost.
It is critical for the success of evaluation that the funds to undertake the tasks be allocated
and guaranteed over the full duration of the project. The rehabilitation project may fail if
adequate and timely evaluation is not undertaken.
Evaluation can be expensive over a longer period depending on the desired outcomes.
There have been instances of the cost of evaluation being similar to the rehabilitation cost.
The extent and type of evaluation needs commitment from the stream managers so that
funds are allocated at the appropriate times during the project.
80
Summary Evaluation is the key to success of any NCD and rehabilitation project. All such projects
should be evaluated, though the extent of evaluation will vary from project to project.
The evaluation may simply be detailed photographic records and mapping of vegetation
types, or on more complex projects, include extensive data collection on a range of
parameters, which will indicate whether the project objectives have been achieved.
All stream rehabilitation projects should be seen as experiments. Learnings from one
project should be used on other projects. Progress can only be made when researchers
and project managers openly learn from failures and build on successes.
The two key questions to ask in designing an evaluation approach are: how confident do
we want to be that we have identified a response, and who are we trying to convince with
the result?
Evaluate only to the level needed. When developing an evaluation plan:
Select measurable characteristics that directly relate to the project objectives (careful
selection of an appropriate spatial scale will be critical)
Establish the desired level of these characteristics by taking measurements in an

appropriate reference stream (or streams)
Determine the timing and duration of measurements needed at the project and control
sites both before and after works (this may include sampling of different flow levels)
Evaluation projects should be designed by experts, who have a full understanding of the
project outcomes and experience in tracking and rectifying problems as they appear.
81
Bibliography
Bibliography
Arcement, G. J. and Schneider, V. R. (1989) Guide for Selecting Mannings Roughness
Coefficients for Natural Channels and Flood Plains. U.S. Geological Survey Water-Supply Paper
2339 United States Government Printing Office, Denver.
Arthington, A. H. and Catterall, C. P. (1990) Brisbane Waterways Strategy Plan! Ecological

Issues in Waterway Management. Centre for Catchment and In-stream Research Division of
Australian Environmental Studies, Griffith University.
Aveyard, J. M. Design Manual for Soil Conservation Works Technical Handbook No. 5. NSW
Soil Conservation Service.
Beschta, R. L. (1991) Stream habitat management for fish in the northwestern United States:
The role of riparian vegetation. Fisheries Bioengineering Symposium, ed. J. Colt and R.J.
White. pp. 53-58. American Fisheries Society Symposium 10.
Brisbane City Council (1996) Urban Stream Rehabilitation Principles and Guidelines.
Brisbane City Council, Queensland.
Brisbane City Council (1998) Ecological Assessment Guidelines, 1998. Brisbane City Council,
Queensland.
Brisbane City Council (1996) Hydraulic Geometry of Brisbane Streams Guidelines for Natural
Channel Design, Part 1 Guidelines. Prepared by ID&A, Brisbane City Council, Queensland.
Brisbane City Council (1997) Erosion Treatment for Urban Creeks Guidelines for Selecting
Remedial Works. Prepared by Rust PPK and Griffith University for Brisbane City Council,
Queensland.
Cotterell, E. (1998) Fish Passage in Streams Fisheries Guidelines for design of stream
crossings. Fish Habitat Guideline FHG 001, Fisheries Group, Queensland Department of
Primary Industries, Brisbane. ISSN 1441-1652.
Cowan, W. L. (1956) Estimating hydraulic roughness coefficients. Agricultural Engineering Vol.

37, No. 7, pp. 473475.
Hicks, D. M. and Mason, P. D. (1991) Roughness Characteristics of New Zealand Rivers. Water
Resources Survey, DSIR Marine and Freshwater, Wellington, New Zealand. ISBN 0-477-02608-7.
Keller, E. A. (1978) Pools, Riffles and Channellisation. Environmental Geology 2(2), 119-127.
Lewis, G. and Williams, G. (1984) Rivers and Wildlife Handbook A Guide to Practices which
further the Conservation of Wildlife on Rivers. Royal Society for the Protection of Birds, and
Royal Society for Nature Conservation: London.
82
Limerinos, J. T. (1970) Determination of the Manning coefficient from measured bed

roughness in natural channels. U.S. Geological Survey Water-Supply Paper 1898B, 47 p.
Newbury, R.W. and Gaboury, M.N. (1994) Stream Analysis and Fish Habitat Design A field
manual. 2nd printing, Co-published by Newbury Hydraulics Ltd. and Manitoba Natural
Resources, Gibsons, B.C. Canada.
Raine, A. W. and Gardiner, J. N. (1995). Rivercare!Practical Methods for Ecologically

Sustainable Management of Rivers and Riparian Vegetation. Occasional Paper Series, No.
03/95. Land and Water Resources Research and Development Corporation: Canberra and
NSW Department of Land and Water Conservation.
Rutherfurd, I.D., Jerie, K. and Marsh N. (2000). A Rehabilitation Manual For Australian
Streams - Volumes 1 & 2. Land and Water resources Reseach and Development Corporation
Shrubsole, D. (1994) Natural Channel Design: Perspectives and Practice. Proceedings of the
first International Conference on Guidelines for Natural Channel Systems held in Niagara
Falls, Ontario, Canada. Canadian Water Resources Association. ISBN 0-9694535-7-4.
Swales, S. and OHara, K. (1980) Instream habitat improvement devices and their use in
freshwater fisheries management. Journal of Environmental Management 10, 167-170.
The Binnie Group, Lidstone and Anderson, and Ian Drummond and Assoc. (1987) River
Training Manual, Department of Transport, Papua New Guinea.
Van Groenewoud, H. (1977) Interim recommendation for the use of buffer strips for the
protection of small streams in the Maritimes. Info. Report II-X-74, Canadian Forestry Service,
Department of Fisheries and the Environment, Fredericton, NB.
83
Channel Features
Bank full flow
Posts
2
3
Rails
1 1m
3 (typical)
Figure A.10 Retard fence (developed from The Binnie Group, et.al., 1987)
Drop structures Traditional vertical and near-vertical drop structures and weirs usually cause serious
impediments to fish passage. Some drop structures can even cause breaks to terrestrial
movement corridors.
To avoid disruptions to aquatic and terrestrial corridors, several small drop structures
should be built rather than one large structure. In areas where fish movement is required,
rock chutes or grouted boulder chutes should be used (refer to Chutes).
In the Brisbane region, a number of gabion drop structures have either suffered damage
to the wire baskets or have been invaded by undesirable weed species such as vines.
Vines invasion can be a problem in natural channels because the vines can damage native
trees. Long-term gabion structures are generally only recommended for use in
watercourses if suitable vegetation can be integrated into the design.
Drop structures can also be a public safety hazard. In areas where fish migration is not an
issue, the maximum drop height should be one metre for safety reasons.
Ephemeral creeks Ephemeral creeks are creek systems that are usually dry and only experience flows for
short periods following storm events.
Pool and riffle systems can still exist in these systems, but the pools are often dry
(Photo A.6). However, some deep pools may retain water all year round.
In urban areas, some naturally ephemeral creeks may begin to experience long periods of
low flow caused by runoff from lawn and garden watering. This water may enter the
106
APPENDIX A
watercourse directly through the stormwater drainage system, or may through increased
groundwater flows.
Before designing or modifying an ephemeral watercourse, visit, investigate and study

existing ephemeral systems in the local area. In particular, take note of the channel
features and the typical surface area and depth of the pools.
Photo A.6 A dry ephemeral creek west of Brisbane
Flows The channels geometry, bed and floodplain are influenced by different design flow rates,
those being: bankfull flow, trickle flow and flood flow.
In a constructed watercourse the channel width, depth, meander radius and allowable
channel slope are determined by the adopted bankfull flow rate and the allowable
bankfull flow velocity at various stages of revegetation.
The bed design is determined by the adopted channel slope and low flow conditions.
While the overbank geometry and vegetation density is governed, in part, by high level
flood flows.
Traditional drainage channel design has concentrated on the specified maximum design
discharge. However, this design methodology is inappropriate when dealing with a
Natural Channel Design for the following reasons:
(i) Not all functions of a watercourse are directly related to discharge.
(ii) Critical aspects of various functions of a watercourse may occur at different

discharges - for example, flooding, erosion and public safety can each reach a critical
level at a different discharge.
107
Channel Features
(iii) In practical terms, there is always a greater possible discharge and a higher possible
flood level.
(iv) Unlike many aspects of engineering, nature has the ability to repair itself with time -
the sustainable level of damage generally being inversely proportional to the
frequency of damage.
It would be expected that a modified watercourse that was designed for non-scouring
velocities during the 100 year ARI event, would in fact be subject to siltation problems
during the more regular flood events. On the other hand, a natural watercourse would
generally be subject to significant vegetation damage during a similar 100 year event with
only minor siltation during regular flood events.
Therefore, if designers are to entertain the idea of conserving natural creeks and rivers, or
constructing vegetated waterways and drainage paths, then the concept of sustainable
damage without failure will need to enter the design philosophy.
A possible design philosophy for an engineered watercourses is provided below:
2 year ARI event Very minor in-bank damage. Non-scouring in-bank flow velocities.
Flows often contained within the banks of the watercourse with no
additional safety hazard to readily accessible public areas.
10 year ARI event Some in-bank vegetation damage. Minor movement of bed
material. Scour holes downstream of tree trunks with minor loss of
trees within the over-bank area, also moderate shrub damage. No
additional safety hazard within public thoroughfare areas.
20-50 year ARI event Significant shrub damage. Moderate tree loss. Isolated bank
damage. Moderate damage to over-bank grassed areas.
100 year ARI event Significant but repairable vegetation damage. No major water
storage embankment failures. Moderate bank erosion. No
significant re-alignment of watercourse. No flood damage to
habitable floor levels of surrounding dwellings.
Grouted rock Grouted rock is used in locations of high flow velocity or shear stress. Given the problems
often experienced in loose rock chutes, grouted boulder chutes are occasionally used in
preference to loose rock. However, grouted rock does not provide the necessary voids
and cavities desired by aquatic wildlife for shelter.
Limited experience with grouted rock chutes in Brisbane has shown that sediment and
vegetation will eventually collect around the rock (refer to Photo 2).
Habitat Habitat is where wildlife lives. Aquatic habitats usually require high quality water during
periods of low flow, and shading to control water temperature.
Habitat also requires shelter - a location to hide from humans and predators, and shelter
from high velocity flood flows.
108
APPENDIX A
Leaf litter In a natural system, much of the stormwater runoff that enters a stream, enters in the form
a sheet flow that passes through the riparian zone. This riparian zone filters out significant
amounts of organic matter from the stormwater along with many other pollutants. As a
result, much of the organic matter that enters a watercourse comes directly from
overhanging vegetation and not from the surrounding catchment.
However, in urban areas stormwater runoff is usually concentrated into some type of
drainage channel, usually directly linked to the street drainage system. Thus most of the
organic matter that falls onto the road surface enters the watercourse unfiltered by the
riparian zone. This very high, unnatural supply of organic matter from street drainage can
alter the food balance resulting in the production of eutropic (over fed) conditions.
Therefore, the deliberate introduction of significant volumes leaf litter into waterways is
questionable in urban areas where stormwater runoff already contains excessive amounts
of organic matter. However, the rate of decay of the leaf matter can also be an important
factor. Many native leaves decay slower than the non native leaves that fall from trees
used to decorate street landscapes. In conclusion, some level of riparian leaf litter supply
will be essential in urban areas, but this is likely to be significantly less than in an
equivalent natural stream. Further discussion on this topic is presented in Chapter 4.
Log weirs Log weirs (Photo A7) are uncommon because there a few naturally available timbers that
can survive long periods in wet conditions without rotting. The use of rock riffles is
possibly a better alternative.
The maximum height of a log weir above the upstream channel invert should be no more
than 1/3 bankfull depth and no greater than 300 mm. Upstream of the weir the bed
should be protected with rock to prevent scour caused by accelerated flows.
Both ends of the weir should be anchored 1-2 metres into the bank or at least 1/3 the
channel width and protected with rock. Undercutting can be avoided by keying the log
weir into the bed.
Bed protection should exist downstream of the log weir a distance of ten times the critical
flow depth achieved in the centre of the weir at the point where the tailwater begins to
drown out the weir flow (i.e. when the tailwater level above the weir crest equals the
upstream critical depth).
Logs The placement of logs within the low flow channel increases the submerged surface area,
traps floating organic matter and provides a resting location for birds, frogs and lizards
(Photo A.8). The surface area and hollows of logs provides living space for invertebrates
such as crustaceans, insects and snails. Hollow logs are also used by vertebrates such as
fish and frogs, functioning as spawning sites and deposition areas for eggs.
If a log is deliberately placed within a watercourse channel to provide a feature, then the log
must be securely anchored to the channel bed to prevent it floating or washing away during
a flood event. In flood problem areas, logs should be placed in the channel so that they
point downstream at an angle no greater than 40% to the direction of flow (Figure A.8).
109
Channel Features
Logs may also be used to create a backwater area to help contain energy dissipation
resulting from a waterfall, drop structure or weir. However, if a log is inappropriately
located, it can generate large scale turbulence resulting in severe bank erosion.
Photo A.7 Log weirs (North Carolina, USA)
Photo A.8 Log placement
110
APPENDIX A
Low flow channels In some creeks, base flow (dry weather flow) may be zero, but a near constant low flow
may occur for some days or weeks following rainfall.
The low flow channel typically meanders within the bed of the channel, occasionally
bouncing off the banks, especially at channel bends. The sinuosity of the low flow channel
within the bed can be totally different from the sinuosity of the main channel within the
valley. It is within this low flow channel that the main pool-riffle system is established.
The size and meander geometry of the low flow channel is usually not a function of the
bankfull flow rate.
Also refer to discussion on Base flow.
Overhanging vegetation Leaf fall from overhanging vegetation can supply organic matter during periods of drought
when organic matter is not supplied by urban stormwater runoff. However, in urban areas
the most important function of overhanging vegetation is to provide shade and thus
moderate extreme temperatures in the low flow channel.
Shading also controls light levels in streams (predation from other fish) and suppresses
weed growth.
Also refer to discussions on Leaf litter.
Pipe crossings Unlike water supply and gas pipes, sewer pipes often cross a channel above the bed level.
As flood waters flow pass a pipe the induced turbulence can cause bank erosion
immediately downstream of the pipe (similar to water flowing around an isolated tree in a
floodplain).
To avoid such erosion problems, rock should be embedded into the bank immediately
downstream of the pipe. Ideally these rocks should be lightly covered with soil and
pocket planted to blend the rocks into the surrounding bank vegetation.
Plan form geometry Plan form geometry is defined by sinuosity (channel length divided by valley length) and
meander wavelength or arc length (see also Channel meanders and Sinuosity).
The design sinuosity may be referenced against similar undisturbed reaches or from
historic aerial photography of the same reach. However, this technique only applies to
channels where the catchment hydrology has remained relatively undisturbed.
Pools In most creek systems, pools provide essential habitat for aquatic life, especially during
dry weather and periods of drought. A pools are also located immediately downstream of
a riffle, waterfall, chute, or drop structure to assist in energy dissipation.
In sandy streams, pools are often found adjacent to structures such as bridges and
culverts. These pools tend to fill with sediment during periods of low flows, and scour
under high flows, such as bankfull flow.
111
Channel Features
Pools located within the low flow channel need to be sized for the given low flow
conditions. Ideally the pools should always contain a moving water surface to prevent
mosquito breeding and maintain healthy water quality (Photo A.9).
Photo A.9 Natural pool in a creek west of Brisbane
Pool-riffle systems Pool-riffle systems are commonly found in gravel bed alluvial streams.
In closed canopy creeks, some degree of bed vegetation (though commonly sparse)
usually exists, and as a result, the pool-riffle system is limited to the width of the low flow
channel rather than the full width of the channel bed (Photos 9 and 14, and Figure 2.12).
While in most open canopy rivers, the pools and riffles often extend the full width of the
channel bed (Photo A.10).
Pools and riffles are usually found in streams with a sediment particle size larger than
coarse sand (0.5-1 mm) and on bed slope of 1 in 200 or less (Keller, 1978; Swales and
OHara, 1980; Lewis and Williams, 1984).
During low flows, the profile of the water surface is stepped with the water surface flat
across the pools and steep over the riffles (Figure A.11). During high flows the pool-riffle
system is usually drowned out and the water surface profile approximates the overall
gradient of the channel.
Rivers and large creek systems tend to have a pool-riffle spacing (the distance between
consecutive pools or riffles) of between 5 to 7 channel widths. The pools typically occur at
channel bends, but may also occur in relatively straight section of a channel.
112
APPENDIX A
Flood flows
Minor flows
Steep bank
Riffle
Pool
Pool
Riffle Riffle Pool

Riffle
Meander wave length
Figure A.11 Traditional full bed width pool-riffle system
Photo A.10 Pool-riffle system in a river system (NSW)

113
Channel Features
Photo A.11 Typical pool-riffle system in a Brisbane creek
Pool Pool
Riffle
Pool Riffle
Pool Pool Pool
Pool
Pool
Pool
Large pools normally

occur at bends
Figure A.12 Pool-riffle system in a typical Brisbane creek
Recessed banks Channel banks can be recessed into the floodplain to form shallow reed pools, or a dry,
low level, floodplain that floods on a regular basis. These recessed, low level banks are
generally sheltered from the high velocity channel flows and are used as bird, fish and frog
breeding areas (Figure A.13).
114
APPENDIX A
Recessed banks should be well shaded to avoid excessive vegetation growth and to
control temperature fluctuations during the day. These areas should also be protected
from domestic animals with the use of vegetation screens and adequate buffer zones.
A A
Main channel Wet or dry

area
Cross section A-A
Plan view
Figure A.13 Recessed channel bank
Riffles Riffles are shallows found in upland streams. Normally composed of gravel, they are part
of deep watershallow water (pool-riffle) sequences inherent to many watercourses. Riffles
provide, aeration, habitat diversity and grade control (control of bed erosion).
The rock found in natural riffles is usually unstable and can be disturbed during bankfull
flows. In constructed riffles the rock size is usually larger to provide stability. Constructed
riffles (Figure A.14) typically have an upstream slope is around 4(H):1(V) and a downstream
riffle slope of 30(H):1(V). Natural riffles can be much flatter.
The natural sorting of gravel and cobbles in riffles during bankfull discharges is an
important process for invertebrate production and fish spawning in natural streams. This
sorting reduces the concentration of fines in the substrate, which can accumulate between
bankfull flows.
A weir, dam or other flood control system if constructed across a stream can adversely
affect riffles in two ways:
(i) it can reduce or remove the supply of suitable rock; and/or

(ii) it can reduce or remove essential, high velocity flood flows that help sort the riffle rock
and maintain an open void, riffle system.
As a general rule, riffle rocks should feel loose under foot. The natural D50 and D85
values for the substrate of a stream are important indicators of preferred substrate
composition of a channel.
When artificial riffles are introduced to clay bed or gravel bed streams that do not naturally
contain a similarly sized rock, then the rocks used to form the riffles need to be of
115
Channel Features
sufficient size to prevent movement during a bankfull flow. This is because any rocks lost
from the riffle will not be replaced by the natural migration of rocks down the stream.
The through-flow of water through the rocks at the crest of a riffle can be desirable, but
can also lower upstream pool level.
Riffle in a narrow low flow channel
0.2 - 0.4 m
(typical)
Part of riffle covered in

vegetation
Part of riffle exposed
to trickle flows
Riffle in a wide channel
Rough, open void surface

with some emergent rocks
in wide riffles
Flow 4
1 30
1
Pool
Pool
Riffle profile
Figure A.14 Typical riffle profiles
116
APPENDIX A
Riparian vegetation It is often difficult to draw the line between riparian vegetation and floodplain vegetation.
The riparian zone usually lies between 1 in 2 year and 1 in 20 year flood level.
Beschta (1991) defines the riparian zone as, "those areas that are saturated by ground
water or intermittently inundated by surface water at a frequency and duration sufficient to
support the prevalence of vegetation typically adapted for life in saturated soil."
The riparian zone is critical to many ecological functions within a watercourse and is
usually characterised by a high diversity of plant species and associated wildlife.
Riparian vegetation provides a long-term contribution to channel stability. However, in

most cases the riparian vegetation will not stop the migration of a stream, it just delays
the process.
The function of riparian vegetation also includes the following:
Reduces downstream flooding by slowing the progress of flood waters and increasing
in- stream storage. However, this may cause local increases in flood levels.
Providing a filter for sediment and nutrients contained within sheet flow as it enters the
watercourse. However, in urban areas most stormwater enters waterways in the form of
concentrated flow, and thus is unaffected by the riparian zone.
Supply of aquatic and terrestrial food, shelter and habitat requirements, including the
supply of logs and essential leaf litter.
Aesthetic and recreational benefits.
Shade of the channel and overbank areas for water temperature control and weed
management.
Reduction in channel velocities during overbank flood flows, especially in channels that
meander across the valley.
Provision of a visual buffer between highly maintained grassed parklands and the low
maintenance, watercourse channel. An irregular, partially eroded channel can appear
unsightly and out of place in the middle of an open park unless surrounded by riparian
vegetation.
Provision of terrestrial movement corridors.
In some circumstances, bank vegetation can decrease flow velocities near the bank and
thus reduce the potential for bank erosion. However, the spacing of woody vegetation
such as trees is critical. Small groups of trees or isolated trees can generate large scale
turbulence and can cause severe bank erosion. As a general rule; riparian trees should be
surrounded by understorey and groundcover plants to prevent soil erosion around the
base of the tree.
117
Channel Features
Minimum riparian widths are usually associated with (but not the same as) terrestrial
corridors widths and have in the past been defined as either a minimum 15 m, 30 m, or 60
m (each side of the channel) measured from the top of bank. Discontinuities within the
riparian zone should be avoided.
Rocks In static creeks (creeks not subject to inbank channel expansion resulting from changing
catchment hydrology) rock protected areas usually exhibit good long-term stability,
especially if suitable deep-rooted vegetation is established over the armour rock (Figure
A.15). In dynamic creeks (creeks subject to inbank expansion or natural meandering) rock
protected reaches regularly fail in the long-term.
Extra protection of the toe is usually required to allow for long-term bed scour, especially
on the outside of bends.
Plants
Fill voids between rocks with
Armour rock
topsoil and seed or pocket plant
Bedding rock
Good soil
Existing poor subsoil

Place small quantity
of topsoil over sublayer
of rock before placing
armour rock
Place a 200mm layer of good,

imported, subsoil (loam) over
existing erodible subsoils
Figure A.15 Rock protection with pocket planting
Cavities around the rocks can provide aquatic habitat and shelter if large rocks are placed
along the water line and if most rocks smaller than 200 mm are removed from this area.
At regular intervals along the low flow waterline, cavities around the rocks can be filled
with soil and planted with ground cover species to provide further habitat diversity.
Significant cavities in the rock can encourage vermin. The use of a wide range of rock
sizes, especially in the upper bank region, may minimise this problem. Rock-protected
banks can also cause significant problems to bank burrowing fauna such as platypus and
some bird speciesprofessional advice should be sought on this matter.
118
APPENDIX A
Fully rock protecting a bank generally results in minimal coverage of trees and shrubs
resulting in limited shelter or protection provided to riparian fauna. Thus rock protection
should only extend over those areas of the bank where it is judged necessary, usually the
bottom half to one third of the bank height.
Weed species typically invade rock protected banks unless soil is placed over the rocks
and desirable vegetation is planted upon establishment of the rock protection. If the rock
is not covered with vegetation, then the colour of the rock can become important for
aesthetic reasons.
For further information on the design and placement of rock in watercourse channels, refer
to Brisbane City Councils publication "Erosion Treatment for Urban Creeks" (1997).
Rock ribs Rock ribs are used to limit the growth of future bed erosion. They are installed in
rehabilitated channels to control bed erosion particularly during the early years of
revegetation (Figure A.16).
Voids filled Suitably sized

with soil rock
Filter cloth
Figure A.16 Cross section of a rock rib
If bed erosion migrates up a channel and approaches a rock rib, the rocks will spill out and
form a short rock chute that should slow the migration of the bed erosion. If bed erosion
does not occur, then the rock ribs will simply blend with the surrounding landscape, and
lie in wait of any future bed erosion.
A rock rib is formed by excavating approximately a one metre by one metre square trench
across a channel, lining the channel with filter cloth, then backfilling the trench with rocks
and soil (level with the existing ground). Finally the ribs are revegetated (pocket planting).
Sediment Sediment flowing down creeks represents possibly the greatest threat to Natural Channel
Design projects.
Sediment smothers pools, riffles and other essential habitat areas, transports phosphorus
which encourages weed growth, and can initiate bank erosion by diverting the low flow
channel and smothering essential bed vegetation.
119
Channel Features
Sediment can also adversely affect fish passage structures, such as artificial pool-riffle
systems constructed on the bed of culverts. Most of the traditional fish ladder designs
used on dams and weirs will fail if installed on the bed of a stream that carries sediment.
Sediment does not affect fish ladders placed on dams and weirs because the sediment is
trapped behind the dam or weir.
To protect a pool-riffle system, or a recently rehabilitated channel, a sediment collection

pond should be installed upstream of the works (Photo A.12).
Photo A.12 Instream sediment ponds with access ramps
Shading and cover Shading is essential for the control of water temperature, particularly in habitat pools.
Cover is essential for the protection of aquatic and terrestrial wildlife from predators.
Shading can be provided by riparian vegetation, both canopy cover trees and
overhanging understorey vegetation. Shading by riparian vegetation reduces the quantity
of light and heat reaching the low flow channel (Figure A.17). It can also help in the
control of weeds and unwanted aquatic vegetation.
Cover can also be provided by overhanging (undercut) banks. Viewing platforms

constructed over stormwater outlets can not only hide the outlet, but can also provide
shading of the dissipation pools that are often found at the end of stormwater outlets.
Snags, boulders, and in-stream vegetation can also provide cover and shade.
As a general rule, channel widths up to 10 metres wide can be adequately shaded by

riparian vegetation.
120
APPENDIX A
Partial to full shade
Figure A.17 Shading
Shelter Shelter is different from cover because cover only hides wildlife from predators, while shelter
can protect wildlife from predators and disturbances such as high velocity flood flows.
In fast flowing waterways the populations of animals can be limited by the amount of
shelter available. A rough, irregular bed can provide many cavities that can act as shelter
for animals during times of flood.
Shelter may consist of cavities between loose bed rock, areas of quiet water behind and
under weed beds, rocks or snags, and backwater areas. Where possible, bed rock should
be uniformly graded (i.e. of similar size) to avoid smaller rocks infilling the voids between
the primary armour rock.
Shelter is different from habitat. The adequate provision of habitat does not necessarily
mean that adequate cover and shelter has been provided.
Sinuosity Sinuosity is the term used to describe the repetitive, though variable, curvature of a
watercourse channel or low flow channel. In natural, dry bed streams, meanders or
sinuosity occurs in both the main channel and the low flow channel.
Newbury and Gaboury (1994) reports that: "In general, stream meanders tend to
complete a full sinusoidal wave form every 12 times the bankfull width measured along
the valley bottom. The average radius of curvature of the meander bends is 2.4 times the
bankfull width."
Within these guidelines the recommended minimum radius of constructed bends is 3

times the bankfull width. However, tighter bends can be produced with the assistance of
rock protection.
121
Channel Features
When reconstructing a watercourse channel, efforts should be made to avoid the

formation of a perfectly smooth and repetitive sine curve meander as shown in Figure
A.18. The recommended sinuosity of the channel should only be used as a design tool to
draft the approximate layout of the channel. The real task is to design and built an
irregular, meandering channel (Photo A.1 and Figure A.6) around the layout of the draft
plan.
Lm
Rc
Wa
ter w el
ay c h a n n
Figure A.18 Channel meander
Snags Snags consist of large woody debris, including logs and fallen trees and branches. They
provide diversity and habitat for fish and other aquatic organisms. Snags are also important
in the aquatic food chain because they trap organic matter which then decomposes to
provide food for invertebrates.
Snags provide shelter and spawning sites for fish. Logs are used by large fish, while
branches, twigs and leaf litter are used by smaller fish. Some species of fish use hollow logs
as spawning sites. Snags also provide roosting and preening sites for birds and reptiles.
On the negative side, snags can aggravate upstream flood levels and can cause scour
holes to be formed around the edge of the snag resulting in bed and bank erosion.
However, when a watercourse channel contains several snags and fallen trees, the
backwater effects can significantly slow the bankfull velocity, thus reducing potential
erosion problems.
If some of the snags are selectively removed, then the resulting increased flow velocity
may begin to cause erosion in locations where problems previously did not exist.
Unfortunately, snag management can often be a case where you are damned if you do,
and damned if you dont.
In fast flowing streams, avoid having snags blocking more than say 10% of the channel
flow area. Protruding branches may be lopped if necessary to prevent hazards or debris
accumulation.
122
APPENDIX A
Generally snags should not be removed for aesthetic reasons. Snags should be removed
or repositioned to avoid hazardous situations and bank erosion. However, it may be
better to treat the bank erosion rather than remove the snag.
Snag retention is possibly more ecologically important in low gradient, sandy bed streams
where the induced bed erosion can form habitat pools. Unfortunately, these are also the
streams most likely to suffer induced bank erosion.
Steep, eroded, Steep, eroded, bare banks often appear in natural stream channels. Such banks usually
bare banks only attract community concern when there is little of no riparian vegetation and the
eroded bank is clearly visible.
These banks can provide nesting areas for certain bird species (Photo 16), as well as
general habitat diversity.
If concerns are raised regarding an existing eroded bank, then first inspect the bank and
assess the age of the erosion. On old erosion sites the soil usually develops an aged
appearance (e.g. a green colour caused by moss, or a dull, grey colour caused by sun
bleaching). If the erosion is old, then it may not require stabilisation.
In the design of creek rehabilitation works it is usually unwise to construct steep, bare,
eroded banks on the outside of sharp bends unless continued creek erosion and channel
migration is the intended aim.
Photo A.13 Birds nest in this steep, unstable bank
Stepping stones The introduction of stepping stones across a watercourse channel should be addressed
with extreme caution (Photo A.14).
123
Channel Features
Photo A.14 Stepping stones.
Avoid placing stepping stones in high velocity channels on near deep pools, energy
dissipaters, drop structures or culverts.
Stormwater outlets In temperate and tropical zones; floodplains, backwaters and side channels are necessary
spawning and rearing habitats for many fish species, birds, frogs and other amphibians.
Therefore, consideration should be given to the suitability of using stormwater inflow
points as side channels (Photo A.15).
Photo A.15 Stormwater outlet recessed into the banks of the creek
124
APPENDIX A
Substrate Substrate is the material forming the channel bed. In general a substrate should be
coarser grained than the bank material and should have low amounts of fines. An excess
quantity of fines will result in the filling of the voids between the larger armour rock thus
reducing essential aquatic habitat and shelter.
Coarse gravels can be important for spawning fish, and loose bed rock can protect the
root system of aquatic plants.
The natural sorting of gravel and cobbles in riffles during bankfull discharges is an
important process for invertebrate production and fish spawning in natural streams. This
sorting reduces the concentration of fines in the substrate, which can accumulate between
bankfull flows.
In a natural system, riffles can be formed and reformed, moved or removed during flood
events. Because of the natural supply of rock down the stream, every rock lost from a
riffle can be readily replaced. This results in the formation of riffles with a uniform rock
size and open voids.
If a weir, dam or other flood control system is constructed across a stream is can adversely
affect riffles in two ways:
(i) it can reduce or remove the supply of suitable rock; and/or

(ii) it can reduce or remove essential, high velocity flood flows that help sort the riffle rock
and maintain an open void, riffle system.
As a general rule, riffle rocks should feel loose under foot. The natural D50 and D85
values for the substrate of a stream are important indicators of preferred substrate
composition of a channel.
TECHNICAL NOTE A1
The term D85 refers to the nominal rock size of which 85% of the rocks are smaller.
When artificial riffles are introduced to clay bed or gravel bed streams that do not naturally
contain a similarly sized rock, then the rocks used to form the riffles need to be of
sufficient size to prevent movement during a bankfull flow.
The flow of water through the rocks at the crest of a riffle can be desirable, but can also
lower upstream pool level.
Terrestrial movement Movement corridors are required for migration, breeding, territory expulsion, predation
corridors and genetic transfer.
125
Channel Features
Waterway corridors provide important corridor links with bushland corridors for various
bird and terrestrial species. It is important for the needs of the terrestrial movement
corridor to be clearly identified in relation to bushland links and open space corridors.
Urban terrestrial movement corridors should be sized in accordance with Councils

Ecological Assessment Guidelines.

The vegetative needs of terrestrial movement corridors varies significantly from species to
species, but generally canopy trees at regular spacing are required for birds and animals
such as sugar gliders, and understorey plants are required to protect terrestrial animals
from predators.
Undercut banks Undercut, eroded banks can provide habitat, shading, and shelter. These features are
difficult to construct and maintain in a rehabilitated stream.
North American guidelines often display examples of constructed overhanging or

undercut banks, but these structures normally rely on the supply of logs made from timber
that does not rot in water.
Water quality As a general rule; to maintain a healthy ecosystem in a typical creek system, the lower the
flow, the cleaner the required water quality.
To maintain the benefits of NCD, stormwater management within the catchment should
promote indirectly connected impervious surface areas. Such drainage systems promote
groundwater infiltration during light rain, but also produce good drainage characteristics
during heavy storms.
Thus, the promotion of indirectly connected, impervious surface areas should increase the
quality and quantity of base flows passing through the pools and riffles during dry
weather. This will help to avoid the formation of stagnant, unsightly pools in urban creek
systems.
Water temperature Water temperature can be a critical component in the establishment of preferred aquatic
species within a watercourse. Shading should be provided, especially around habitat pools.
Trees planted on the northern bank can provide the best shade (also see Shading).
Waterway crossings The order of preference of waterway crossings in fish migration areas is:
(i) bridge;
(ii) precast arch structure (Photo A16);
(iii) buried (earth bed) box culvert;
(iv) box culvert;
(v) pipe culvert.
Also see Aquatic corridors and Terrestrial corridors.
126
APPENDIX A
Photo A.16 Arch bridge
Weirs Large weirs create backwater reaches where flow velocities are reduced. They provide
grade control in a similar manner to riffles. A plunge pool is normally created downstream
of a weir and this pool can provide significant habitat benefits. However, this habitat
accumulation often exists because aquatic life cannot migrate past the weir.
These structures are more suitable to steeper gradient rivers and must be properly
engineered to avoid outflanking or deterioration. The major drawback with weirs and
dams is upstream sedimentation and prevention of fish passage.
If a weir structure is required, try to use several small weirs rather than one large structure.
A typical arrangement of small rock weirs is shown in Figures A.19 and A.20.
Wildlife corridors Refer to Aquatic movement corridors, or Terrestrial movement corridors.
Woody debris Refer to discussion on Snags.
127
Channel Features
Figure A.19 Example of a small rock weir
128
Figure A.20 Example of W-Rock Weirs
129
Appendicies
APPENDIX A
Appendix A - Channel Features
Introduction The geometry of a watercourse is generally determined by the water flow, the substrate
material and the vegetation. However, the ecology of a watercourse is determined by
more than just the water, soil and vegetation. Life within a channel can depend on the
ecology of the entire valley and beyond.
This chapter reviews the basic principles of Natural Channel Design and describes many of
the features that can appear in natural channels. It is the appropriate application of the
channel features for a given site that can have the greatest influence on the successful
integration of the channel design or restoration into the surrounding environment.
The principles of Natural Channel Design can be applied at various scales from the full
catchment down to specific aspects of a given site. The levels of investigation and design
are depicted in Figure A.1.
Catchment All investigations and designs should start from a catchment level. In catchment
management issues investigations the waterway system is studied from a broad catchment or regional point of
view. Useful tools can include Catchment Management Plans, Local Area Plans,
Stormwater Management Plans, Open Space Plans, Park Master Plans and aerial
photography - both current and historical.
It is particularly important to ensure that the design proposal is appropriate within the
overall management context of the catchment, parks and watercourse.
Greatest fish abundance, greatest recreational satisfaction, and greatest ecosystem health
will be achieved by promoting the physical and biological features and processes of
waterways and riparian areas that occur most frequently within a given region. In urban
streams the key elements are water quality, shading, shelter, continuous movement
corridors and habitat diversity.
Without an appropriate catchment level approach to Natural Channel Design it is highly

likely that the constructed or rehabilitated channel will require excessive maintenance and
may in fact fail in the long term.
85
Channel Features
Catchment
Catchment and corridor management planning of overall waterway and valley
w
system.
rvie
ove
Catchment
ted
gra
Plan form of reach
nte
Investigation into channel location and potential upstream
and downstream impacts.
ad i
Channel Bro
Cross section
Design of cross sectional shape and
vegetation.
Channel
cus
Features
Bed form and ecological
features.
w fo
deposition
erosion
arro
dn
aile
Det
Figure A.1 Geographic scale of investigation and design
86
APPENDIX A
Three of the most significant impacts on Natural Channel Design result from excessive
sediment flows, excessive urban encroachment into the buffer zones, and the formation of
breaks in wildlife movement corridors. All three of these issues can be addressed through
appropriate town planning and catchment management.
If excessive sediment flowing down the stream is likely to become a problem, then a
major sediment trap (pond) may need to be constructed upstream of the rehabilitation
works.
Plan form issues The second phase of any investigation or design should consist of plan form studies, i.e.
the horizontal movement of the channel through the valley. In plan form studies the
potential impacts of the channel design on upstream and downstream reaches of the
watercourse are considered, as well as any potential impacts these reaches may have on
the rehabilitated section of the watercourse.
The location of the meandering channel can have significant impacts on flood control,
channel stability, wildlife and human movement corridors and recreational activities.
Photo A.1 Aerial view of a meandering channel
Cross sectional issues In the third phase, consideration is given to the cross sectional shape of the channel.
Channel cross sections are mainly influenced by four flow regimes: base flow, bankfull
flow, riparian flow, and flood flow.
Design considerations need to identify how a watercourse and its floodplains can convey
the various flow regimes in a manner that supports the various functional requirements of
87
Channel Features
the drainage system, such as: flood and erosion control, safety, ecology, low maintenance,
aesthetics, conservation, recreation and cultural issues.
Figure A.2 shows a typical channel cross section.
Floodplain
High bank
Slope
Berm or terrace
Toe
Setback Channel bed

area
Riparian zone
Figure A.2 Channel cross section
Channel features Finally, the design needs to address the desired features of the channel, in particular, the
channel bed features. The guidelines presented in this document focus on the use of
pool-riffle systems, but other bed forms do exist and may be more appropriate for the
given site conditions.
Some species prefer well defined pools and riffles, low levels of fines in the riffle
substrates for spawning, deep pools (ideally with log-jams and/or undercut banks for
juveniles and adults), and a complex, shallow edge composed of mildly irregular shoreline
(ideally with small amounts of wood debris, for fry). While other species require well
developed pools with a good depth and complexity, small side channels or scalloped
margins, and riffles and runs composed of coarse rock and boulders. Some species of fish
are primarily pool dwelling, while others are primarily riffle dwelling.
One of the constraints often experienced in NCD is severe urban encroachment of the
waterway and floodplain. In many situations there may be no mechanism or funding for
the purchase of adequate land to undertake the ideal channel design. However, this does
not mean that a NCD approach should be abandoned. All reasonable and practicable
efforts should always be taken to incorporate as many natural and functional channel
features into the channel as desirable for the local conditions.
88
APPENDIX A
The following channel features are discussed within this section. The features highlighted
in bold are the more notable discussion topics associated with Natural Channel Design.
Aquatic wildlife (ecosystems) Low flow channels

Aquatic movement corridors Overhanging vegetation
Aquatic plants Pipe crossings
Bankfull flow rate Plan form geometry
Banks Pools
Base flow Pool-riffle systems
Berms Recessed banks
Boulders Riffles
Bridges Riparian vegetation
Buffer zones Rocks
Causeways Rock ribs
Channel banks Sediment
Channel beds Shading and cover
Channel meanders Shelter
Chutes Sinuosity
Compound channels Snags
Corridors Steep, eroded bare banks
Culverts Stepping stones
Deep pools Stormwater outlets
Deflectors Substrate
Drop structures Terrestrial movement corridors
Ephemeral creeks Undercut banks
Flows Water quality
Grouted rock Water temperature
Habitat Waterway crossings
Leaf litte Weirsr
Log weirs Wildlife corridors
Logs Woody debris
Aquatic wildlife When considering the environmental needs of aquatic wildlife, it is important not to
(ecosystems) restrict attention to just fish. Aquatic life includes fish, platypus, frogs, insects and
macrobenthic organisms.
Studies into macrobenthic organisms (especially in sediment) can be much more indicative
of stream health.
Most aquatic life share the following basic needs:
Habitat a place to live, usually requiring high quality water during periods of low flow
and shading to control water temperature.
Shelter a place to hide from humans and predators, as well as shelter from high
velocity, flood flows.
89
Channel Features
Sustainable ecosystem a balanced ecosystem of food,predators, habitat diversity and

suitable water quality.
Food supply often closely linked with insect population, aquatic plants, leaf fall from
riparian vegetation and the existence of snags. In urban areas a very high, unnatural
supply of organic matter from street drainage can alter the food supply balance.
Increasing the surface area of submerged objects such as rocks, plants and snags can
also influence food supply.
Aquatic movement Movement corridors are required for: migration, breeding, territory expulsion, predation
corridors and genetic transfer.
Fish are required in waterways for two main purposes:
(i) to assist in achieving an ecological balance within the waterway environment; and
(ii) to control mosquito numbers, thereby reducing human health risks.
To achieve the first requirement, native fish are usually required; however, the second
requirement of mosquito control can be achieved by both native and non native fish.
Some fish migrate along streams, while others may have very limited migration needs.
The fact that a significant break in the aquatic movement corridor may exist downstream
of the proposed rehabilitation works does not mean that the need for fish migration, or
the migration of other aquatic species, should be ignored.
Expert advice should be obtained within Council and from DPIs Fisheries Group on the
need for fish movement and the type of fish species likely to occur within the waterway.
When the likely fish species are known, then Table 1 in Cotterell (1998) should be
reviewed to assess the times of year in which fish migration is expected to occur.
Construction works within the watercourse should be avoided during these periods.
The general requirements for native fish migration are presented below:
(i) Order of preference of waterway crossings:
(a) bridge;
(b) precast arch structure;
(c) buried (earth bed) box culvert;
(d) box culvert;
(e) pipe culvert.
(ii) Low flow or base flow conditions (at the time of fish migration) represented by:
(a) maximum 1 m/s (riffles);

(b) maximum 0.3 m/s (channel or pool flow);
(c) minimum 0.20.5 metres depth;
90
APPENDIX A
(d) low levels of large scale turbulence (relative to flow depth), i.e. avoid creating
whirlpools.
(iii) A rough, irregular bed, similar to natural bed conditions.
Aquatic plants Aquatic plants can be a major source of aquatic food both in the plant material and
through substances that grow on the submerged surface area of the plant. These plants
are needed around the edges of pools.
Some aquatic plants, such as reeds, can adversely affect the hydraulics of the channel. In
straight sections of channel, a thick growth of reeds can initiate the formation of a channel
meander by deflecting flow around the reeds and into the opposite bank. This usually
occurs after an extended period of drought.
Bankfull flow rate The bankfull flow rate can be highly variable along a given reach. In natural streams,
bankfull flow is typically Q1 to Q2. In degraded urban streams, bankfull flow may be as
high as Q20, and is typically Q5 to Q10. However, in a rocky gorge the bankfull flow rate
may be in excess of the Q100 discharge.
The geometry of a channel is usually directly related to the bankfull flow rate. The depth
of bankfull flow (i.e. the height of the lowest channel bank) has a strong influence on the
maximum channel velocity and shear stress, which in turn influences soil erosion.
The stability of most dry bed channels is controlled by the health and density of in-bank
vegetation, except in sandy creeks. The stability of vegetation is governed by the
duration and velocity of bankfull flows.
The duration of bankfull flow is very important. If the channel is located downstream of a
stormwater retardation basin or dam, then the duration of bankfull flow may be much
longer than in natural conditions and significantly more channel instability may result. To
prevent erosion in such cases, the channel may either need lower channel velocities, more
channel protection, or possibly an open canopy to promote a strong vegetation cover.
Banks Introducing a wide variation in bank slopes along a channel and floodplain can be an
effective way of making a constructed or modified watercourse appear natural.
Steep banks on the outside of bends may exist in natural conditions, but they are usually
unstable or formed in rock. Avoid placing steep, unprotected banks in areas where
watercourse stability is desired.
Also refer to Channel banks below.
Base flow Base flow is often referred to as the low flow condition. The provision of habitat and high
quality base flows are probably the two most important factors in the establishment of a
healthy ecosystem.
91
Channel Features
As a general rule; the lower the flow, the greater the need for clean water. Many creek
systems contain turbid water during storm events and flood flows, but most creeks need
clear, high quality water retained in their pools during periods of drought, or flowing
through their riffle systems during periods of low flow.
The base flow rate should be determined from existing conditions, with an allowance
made for expected future changes to the catchment. It has often been reported that
urbanisation decreases base flow due to an increase in impervious surface area. This
increase in impervious area is thought to result in a reduction in rain water infiltration.
However, garden and lawn watering in urban areas can often result in increased base flow
conditions, especially during periods of drought when stream flow would normally cease.
The highest quality base flow usually originates from groundwater inflows and springs.
Wherever practicable, the urban stormwater drainage system should be designed to
maximise the infiltration of stormwater into the groundwater, especially during periods of
light rainfall.
Photo A.2 Base flow in Kedron Brook, Brisbane
92
APPENDIX A
Berms Refer to Channel banks.
Boulders Boulders placed on the bed of the watercourse can increase habitat diversity, while
boulders recessed into the low flow channel or the pools can increase the total
submerged surface area. Increasing the submerged surface area can increase the
available food supply for aquatic life.
Boulders can also help to increase dissolved oxygen levels in the water. However, their
value for oxygenation would be greater in river systems. In small urban creeks, riffle
systems would provide better oxygenation during the critical low flow periods.
Boulders should not be placed in locations that are likely to generate large scale
turbulence during periods in which fish migration.
The effect of boulders on flood levels decreases with increasing flood depth. Generally,
the placement of boulders is unlikely to affect overbank flood levels. The likely affects on
flooding can be assessed hydraulically by reviewing the expected change in Mannings
roughness (refer to Appendix C).
The stability of a boulder cannot be determined by using a standard rock sizing formula
such as (D50 = 40V2) because such design formulas are only suitable for rock that is placed
flush with the channel surface. A boulder is subjected to much higher hydraulic forces
because the majority of the boulder protrudes into the flow.
The boulder size should be less than 1/5 the low flow channel width if located within the
low flow channel, or less than 1/5 the bed width if located on the main channel bed
outside the low flow channel. Boulders are typically 0.6 to 1.5 metres diameter.
Where possible, inspect the watercourse for evidence of existing boulders that have
remained undisturbed within the channel for some time. The size of these boulders will
provide a good indication of the size needed to withstand flood flows in the rehabilitated
channel.
Recommended placement density of isolated boulders is 1 per 27 m2 or 2 per 70 m2.

Alternative, triangular clusters of 3-5 boulders should be placed at least 1/3 the channel
width apart.
WARNING: Boulders can easily deflect high velocity flows onto a bank causing erosion.
Caution should be used when using boulders to dissipate energy as the water jet may
simply deflect over the boulders.
93
Channel Features
Photo A.3 Boulder placement in a natural channel
Bridges Bridges are the preferred means of crossing waterways because they can have the least
impact on terrestrial and aquatic movement corridors.
Unlike most culverts, bed erosion can only be controlled at bridges if significant rock
protection occurs across the channel in association with scour protection around the piers.
Buffer zones Buffer zones are different from riparian zones and wildlife movement corridors. Their main
purpose is to provide an area of land that can buffer the effects of noise, urbanisation
and channel erosion. For example, a buffer zone may be required to separate a football
oval or an industrial complex from a critical bird roosting area.
Thus, a buffer zone aims to control lateral disturbances, while riparian zones are concerned
with aesthetics, habitat and stability; and movement corridors are concerned with wildlife
migration.
Buffer zones are used to protect creeks from adjacent land uses, and to protect adjacent
land uses from the adverse effects of creek migration.
94
APPENDIX A
Causeways Causeways can either be established at normal bed level, or can be raised above base
flow level. Raised causeways usually incorporate a low flow pipe to carry dry weather
flows.
Raised causeways are normally preferred if the causeway has regular vehicular traffic
because these structures can minimise sediment flow into the watercourse. However, bed
level causeways are preferred for fish migration. Raised causeways also tend to trap debris
and redirect high velocity flows causing bank erosion.
When inappropriately designed, causeways can provide a significant barrier to aquatic

movement. If fish movement is required, then the low flow pipe under the causeway
should be designed to meet the aquatic movement corridor requirements specified above
(Aquatic movement corridors).
Channel banks Suggested maximum watercourse bank slopes are presented below.
1(V):2(H) Good, erosion resistant clay or clay-loam soils with a healthy, deep- rooted bank
vegetation formed by suitable riparian groundcover species, shrubs and trees.
1(V):3(H) Sandy-loam soil with groundcover vegetation or a closed canopy channel with
shaded banks and sparse bank vegetation.
1(V):4(H) Sandy soils.
1(V):6(H) Mowable, grassed banks undesirable in Natural Channel Design.
The above bank slopes have been presented as a general guide only. Final bank slope
should be based on bank stability and revegetation requirements.
In some locations, particularly on the outside of bends, a bare, near-vertical bank can
provide a suitable nesting area for some bird species. However, these banks can be highly
unstable.
Channel banks should be able to withstand minor bank erosion without becoming
structurally unstable. There is usually two types of vegetation that play a role in stabilising
watercourse banks. Firstly ground cover plants, such as Lomandra, help to control surface
scour. Secondly, deep rooted plants, such as trees and shrubs, help to anchor the bank
and prevent land slips. Land slips can occur when the bank is undermined by erosion
(undercutting), or the weight of the bank temporarily increases as a result of water
saturation (slumping).
Berms may need to be introduced to steep banks that are higher than three metres, or
where erosion is expected along the toe of the bank. A typical berm width is one metre
(Figures A.3, A.4 and A.5).
The berm presented in Figure A5 can be used to delay the effects of toe erosion on the
upper region of a bank. This delay can allow time for new, deep-rooted vegetation to
95
Channel Features
become established that will hopefully provide long-term stability to the bank. This
approach to bank erosion is unlikely to be of benefit on the outside of channel bends
where the toe erosion is likely to continue to attack the toe of the bank.
Establishing a berm on a watercourse bank can provide the following benefits:
If bank erosion occurs at the toe of the bank and this erosion results in soil slumping,
then the berm can reduce the area of the bank over which this slumping occurs.
Berms can be used to increase the stability of steep banks by either adding weight to
the toe of the bank, or removing weight from the crest of the bank.
Berms can be used to delay the effects of soil erosion around the root system of
establishing trees.
Wide berms can be used to provide pedestrian access to the watercourse banks for
recreational and educational purposes.
As a general rule; to assist in the overall stabilisation of the watercourse channel, the
banks should be hydraulically rougher (i.e higher vegetation cover) than the bed.
Berm or terrace
Figure A.3 Channel bank berm
96
APPENDIX A
Hb
a)
Toe
erosion
b)
Bank slip and loss of vegetation
Expanding
toe erosion
c)
Figure A.4 High bank erosion without berm
97
Channel Features
Terrrace
or berm
a)
Toe
erosion
b)
Extent of
bank erosion
limited by berm
c)
Figure A.5 High bank erosion with berm
98
APPENDIX A
Channel beds The bed form within a watercourse basically depends on the composition of the bed
substrate, the size of the catchment, and the existence or non existence of bed
vegetation.
Even in closed canopy creeks, significant (though commonly sparse) bed vegetation
usually exists. As a result, the pool-riffle system is often limited to the width of the low
flow channel rather than the full width of the channel bed as normally occurs in rivers.
It is generally not considered desirable to establish a particular bed form for the specific
task of encouraging non-native species to an area (i.e. establishing a pool-riffle system
within an area that does not normally contain pools and riffles). However, it is noted that
when urbanisation or other forms of land development significantly change the catchment
hydrology, it is possible for the bed form of a watercourse to also change. These changes
can result in pools and riffles being formed in areas where they previously didnt exist.
As a general rule; the bed should have coarser grained material (soil) than the banks.
Channel meanders Straight streams are rare in nature (Photo 3). Natural streams seldom remain straight
through a distance more than approximately 10 channel widths (Swales and OHara, 1980).
Most natural streams follow an irregular course, termed meandering.
Constructed channels often develop meanders as a result of sediment deposits or the

irregular growth of bed vegetation. Meanders alter the effective slope of the channel and
can increase its resistance to flow through increased form roughness.
The meander should be similar to those exhibited by adjacent reaches of the watercourse
provided the catchment hydrology and physiography are similar. Meanders within a given
reach often have similar dimensions, but with some degree of variability (Figure A.6).
L2
Figure A.6 Channel meanders
99
Channel Features
Chutes Rock chutes are used as an alternative to vertical or steep drop structures. They are in
effect, long riffles, and are often referred to as grade control structures.
To allow fish migration, rock chutes should have a maximum grade of 1 (V):10(H) to
1(V):15(H) and a maximum length of around 10 metres between pool resting areas. Ideally
the grade should be 1 in 30 or flatter; however, this can result in a very long chute.
Because fish can swim at burst speed for only short distances, the greater the flow velocity
down the chute, the shorter must be the distance between rest areas (i.e. pools).
Rock chutes should be used instead of a pool-riffle system when the channel slope is
steeper than say, 1 in 20. Pool-riffle systems typically exist in channels with slopes
between 1 in 20 and 1 in 1000.
Photo A.4 Rock chute and deep pool in a reconstructed channel
Compound channels A compound channel (Figure A.7) is a waterway channel, including its floodplains, than
contains a variety of vegetated states from bushland to open grass. They can perform a
variety of functions from ecological conservation to active parkland.
Through appropriate design, compound channels can have effective flood routing
capacity. These channels can incorporate natural floodplains and can enhance species
richness and diversity.
Compound channels are commonly used in urban areas and in the rehabilitation of urban
drainage channels.
If the high bank is located on the north side of the channel, then it can provide shade.
However, if the soils are moderately to highly erosive, then this may result in the formation
of an unstable, poorly vegetated bank.
100
APPENDIX A
If flood control is a major issue, and the excavated floodplain is required to have few trees
or shrubs, then the floodplain is possibly better located on the southern side of an east-
west aligned channel. This means that shade trees can be planted on the high, northern
bank where they will have the least impact on flood levels.
The above issues are not so critical for creeks that are aligned in a north-south direction.
Lower
berm Upper berm
Main stream
channel
Modified level
Figure A.7 Compound channel
Culverts In fish migration areas, consideration should always be given to the construction of a
bridge in preference to a culvert.
There are various functions that culverts can perform besides the movement of water.
Each of these functions should be given appropriate consideration so that the culvert can
be suitably integrated into its environment. These functions are briefly discussed below.
Aquatic movement: (i) Order of preference of waterway crossings:

(a) bridge;
(b) precast arch structure;
(c) buried (earth bed) box culvert;
(d) box culvert;
(e) pipe culvert.
(ii) Culverts usually require wet and dry movement areas (Photo A5), therefore a low flow
channel is usually required.
101
Channel Features
Photo A.5 New culvert showing dry and wet cells and a cell designed for a future bikeway
(iii) Low flow or base flow conditions (at the time of fish migration) represented by:
(a) maximum 1 m/s at entry-exit points during fish migration periods;
(b) maximum 0.3 m/s through culvert during periods of fish migration;
(c) minimum 0.20.5 metres depth;
(d) low levels of large scale turbulence (relative to flow depth), i.e. avoid creating
whirlpools.
(iv) A rough, irregular bed, similar to natural bed conditions.
(v) Where practicable, the bed of the culvert should reproduce the natural bed conditions
of the watercourse. This may include artificial pool-riffles. However, consideration
needs to be given to potential mosquito problems and debris blockage within the
culvert.
(vi) Light should be encouraged in the culvert, or in the case of a multi-cell culvert, into the
migration cell. In multi-lane roads, stormwater drop inlets can be installed into the
median strip to allow the entry of light as well as stormwater. However, during
nightfall, indirect street lighting of the culvert entrance should be avoided.
Terrestrial movement: In single cell culverts, a dry migration path can be established with the construction of a
0.51 metre wide, stone pitched path. This path needs to be elevated above the base
flow water level (for details, refer to Appendix B).
102
APPENDIX A
Deep pools Deep pools are places of refuge for fish and insects during times of low flow. Ideally the
bed of the pool should contain some rock to allow aquatic life to shelter from flood flows.
Deep pools can also be the only source of permanent water during periods of drought.
These pools can be essential for the retention of mosquito-controlling fish, especially if
hydraulic structures exist downstream that disrupt aquatic movement along the
watercourse.
Deflectors Deflectors can create and sustain pools in wide channels and enhance sinuosity in straight
channel reaches. They can be triangular, straight or angled and normally extend up to half
way across the watercourse (Figure A.8).
They should be used with extreme caution because they can result in severe bank erosion
caused by the formation of large scale eddies. These eddies may also interfere with the
migration of fish. However, they can be used as fish shelter areas (rest spots) in fast
flowing streams.
Deflectors, including retards (Figures A.9 and A.10), groynes, dykes and vanes, can be
used to control erosion on the outer banks of creek bends, but in many cases the
preferred solution is the use of rock protection on the outer bank.
Their benefits are minimal in low velocity streams, say less than 0.9 m/s. Typically they are
used in wide, shallow streams with gradients greater than 3%.
Avoid placing deflectors in streams containing fine, non-cohesive bed material.
The spacing of small habitat deflectors should be 5-7 times the low flow channel width.
The spacing of groynes in a flow deflection system is influenced by a number of factors
and reference should be made to an appropriate design guideline. Typically, groynes will
protect 3 to 5 times their length (perpendicular to the bank to which it is attached).
They should intrude into the low flow channel no more than one third the low flow
channel width. Their height should be not more than 0.15-0.3 metres above the low flow
elevation unless the deflectors are designed to form part of major bank protection and
rehabilitation works (i.e. groynes and retards).
Timber structure needs to be secured into the bank at least 1-2 metres.
Deflectors placed in isolation are typically angles at least 45O to the direction of flow.
However, groynes and retards may be placed perpendicular the direction of flow (noting
that groynes and retards are normally placed in groups).
103
Channel Features
Boulders used to
simulate a channel
meander in
straight channels
Toe protection
on outside
of bend
Well-anchored
log placed
on outside of
Sand bar bend to deflect
water away
from the bank
Pool formed
by eddies
Not to scale
Figure A.8 Rocks, boulders and a log used as deflectors
104
APPENDIX A
Retard barriers
Remove sediment to
open the channel's
flow path
Scour holes will appear

at ends of retards
Promote shrub
growth to
stabilise
sediment that
accumulates
between
the
barriers
Bank protection
Figure A.9 Retard layout (developed from The Binnie Group, et.al., 1987)
105
APPENDIX B
Appendix B - Hydraulic Structures

in Natural Channels
Introduction Hydraulic structures such as culverts and weirs can act as barriers to the general
movement or migration of aquatic and terrestrial wildlife. These structures often also
need to incorporate pedestrian pathways and bikeways.
In this appendix the following issues will be addressed with respect to the impact of
hydraulic structures on the design of watercourse and drainage channels:
(i) migration of aquatic wildlife;

(ii) migration of terrestrial wildlife;
(iii) impacts on riparian and floodplain vegetation;
(iv) public access, movement and safety requirements.
Migration of aquatic The general requirements for native fish migration are presented below:
wildlife
1. Order of preference of waterway crossings:
(i) bridge;
(ii) precast arch structure;
(iii) buried (earth bed) box culvert;
(iv) box culvert;
(v) pipe culvert.
2. Low flow or base flow conditions (at the time of fish migration) represented by:
(i) maximum 1 m/s (riffles);

(ii) maximum 0.3 m/s (channel or pool flow);
(iii) minimum 0.2 - 0.5 m depth;
(iv) low levels of large-scale turbulence (relative to flow depth), ie whirlpools.
3. Rough, irregular bed, similar to natural bed conditions.
4. Low level of large-scale turbulence.
Arch structures Prefabricated arch structures such as CSR Humes Classic Arch and Bebo Arch can be
used on waterways to provide natural bed conditions under a bridge, while also providing
the cost benefits of prefabricated structures (Photo B1).
If appropriately designed and constructed, arch structures may be equivalent to bridges

with regard to their minimal impact on the migration of aquatic wildlife.
130
APPENDIX B
Photo B1 Arch bridge
Bridges In locations where fish migration is required, bridges are the preferred means of crossing
waterways because they cause minimal changes to the waterway channel and thus can
have the least impact on aquatic movement.
To minimise the impact on fish migration, the piers of a bridge should be designed to
avoid the generation of large-scale turbulence, ie whirlpools.
Chutes If located in a fish migration area, the following features need to be considered in the
design of a chute:
(i) ideally, the maximum fall should be 1 m;

(ii) suitable voids should exist in the primary chute armour to allow the shelter of small
aquatic animals during floods;
(iii) recommended chute gradient is 1 in 15 or flatter, but chute grades as steep as 1 in 10
have been used;
(iv) the chute cross-section should promote the formation of a low flow channel;
(v) if rock is used as the primary armour, vegetation should be promoted in the voids
between the rocks, except in the area of the low flow channel.
To allow fish migration, rock chutes should have a maximum grade of 1(V):10(H) to
1(V):15(H) and a maximum length of around 10 metres between pool resting areas.
down the chute, the shorter the distance must be between rest areas (ie pools). However,
tailwater conditions must be checked to see when the chute becomes drowned.
131
APPENDIX B
Figure B2 Multi-cell culvert
Figure B3 Culvert inlet, (Plan View)
Figure B4 Section A - A
133
Hydraulic Structures
To assist in the development of the above conditions in multi-cell culverts, and also to
assist in the control of sediment flow, an inlet weir can be formed. Sediment training walls
can be designed to direct all trickle or base flows to one wet cell as shown in Figures B3
and B4.
The typical height of the training wall that is placed in front of the dry cells is 0.3 - 0.5
metres. For hydraulic reasons, the training walls need to be located at least four times the
height of the wall away from the culvert entrance as shown in Figure B4.
Drop structures Traditional vertical and near-vertical drop structures usually cause serious impediments to
fish passage.
To avoid disruptions to aquatic corridors, several small drop structures should be built
rather than one large structure. In areas where fish movement is required, rock chutes or
grouted boulder chutes should be used.
Weirs Weirs create upstream backwater regions where velocities are reduced and coarse
sediment transported by flood events is allowed to settle. An energy dissipation pool is
normally created downstream of the weir and this pool can provide significant aquatic
habitat benefits. Unfortunately, the accumulation of aquatic species in these areas is
usually the direct result of the fish not being able to migrate past the weir.
To improve fish migration past weirs, fish ladders need to be installed. Fish ladders are
different from the open channel fishways employed in riffle, chute and drop structure
design because they do not need to operate in high sediment load conditions.
The design of fish ladders is not discussed in this document. Reference should be made
to the local Fisheries Authority (ie QDPI, Fisheries Group) for advice on the design and
installation of fish ladders.
Migration of terrestrial Waterway corridors provide important corridor links with bushland corridors for various
wildlife bird and terrestrial species. It is important for the needs of the terrestrial movement
corridor to be clearly identified in relation to bushland links and open space corridors.
Terrestrial wildlife movement corridors are required along waterways and across catchment
boundaries for migration, breeding, territory expulsion, predation and genetic transfer.
The need for terrestrial migration paths along the waterway should have been established
during the planning stage of the proposed works.
Within Brisbane City Council, terrestrial wildlife movement corridors should be sized in
accordance with Councils Ecological Assessment Guidelines.
Discontinuities within the corridor should be avoided. These discontinuities can be

formed by roads, culverts, certain types of drop structures and by the removal of essential
riparian vegetation.
134
APPENDIX B
Where transport infrastructure (road/rail,etc) crosses environmentally sensitive areas with

riparian vegetation, then the Bridge/Culvert should provide adequate spans to maintain a
continuous riparian zone.
The vegetative needs of terrestrial wildlife movement corridors vary significantly from
species to species, but generally canopy trees at regular spacing are required for birds
and animals such as sugar gliders, and understorey plants are required to protect ground-
dwelling animals from predators.
Arch structures It is important to make sure that a dry path is formed under the arch and to make sure
that any future sedimentation that may occur under the arch does not result in the loss of
this dry animal migration path.
Bridges When properly designed, bridge structures generally provide the least interruption to
terrestrial movement.
To allow for terrestrial wildlife movement, the bridge should span the main channel plus a
portion of the floodplain on each side of the channel. In some cases, terrestrial movement
considerations may apply only to one side of the main channel. However in most cases, it
is usually extremely important for each side of a permanently wet stream bed to be
considered as a separate terrestrial movement corridor with its own specifications and
requirements.
In some locations it may be essential for appropriate vegetation to be established under

the bridge to provide cover for terrestrial wildlife migration. This may require special
stormwater drainage consideration to allow the regular supply of rainwater to the plants.
Chutes In most cases, chutes can be surrounded by riparian vegetation corridors that can act as
terrestrial movement corridors.
Design consideration would be the same as for the rest of the watercourse
Culverts Terrestrial wildlife movement through culverts can be affected by a number of factors, but
the most important requirement is for the formation of a dry migration path. In single-cell
culverts, the production of a dry migration path normally requires the construction of a low
flow channel or elevated fauna path.
Fauna path To allow for fauna movement in single-cell culverts, a dry migration path may need to be
established. A fauna path may consist of a 0.5 1.0m wide, stone pitched path elevated
above the base flow water level (Photo B2).
Fencing Fencing, such as Koala-proof fencing, can be used to direct animals to the culvert to avoid
animals moving over the road.
135
Photo B2 Dry fauna path in a single-cell box culvert
Lighting conditions When aquatic passage is provided, light should be encouraged to enter the culvert or, in
the case of a multi-cell culvert, to enter the wet cell(s). In multi-lane roads, stormwater
drop inlets can be installed into the median strip to allow the entry of light as well as
stormwater.
As a general rule, if the road is wide enough to need a raised or painted median, then it is
wide enough to need the introduction of a skylight into the dedicated migration cells.
Lizard run Lizard runs are approximately 100mm wide strips of timber bolted to the waterway
embankment side of a culvert cell. They are introduced to a culvert to enable smaller
reptiles to move through the culvert at an elevated height.
Lizard runs should be located approximately 300 mm below the cells obvert. The lizard
run must extend from ground level at the upstream wing wall, through the culvert to
ground level at the downstream wing wall (Figures B1 and B2).
Median traffic barriers Where practical, any traffic barriers installed along the median should contain breaks to
allow animals to pass over the road. The design of these breaks needs to be discussed
with both fauna and traffic safety experts.
Side wall roughness Ideally the culvert cell wall adjacent to the watercourse bank should be roughened with
texture paint, grout or other suitable material to allow for the movement of fauna such as
lizards.
Any increase in culvert wall roughness must be taken into account in the hydraulic analysis
when designing the culvert. However, it is noted that in a typical road culvert, side wall
friction represents only around 12% of the total head loss.
136
APPENDIX B
Street lighting Many animals migrate only at night. To assist in the migration of these animals, street
lighting adjacent to culverts should be fitted with metal shields to prevent the lighting of
the culvert entry/exit.
In some circumstances it may be desirable to paint the concrete wing walls and apron in a
dark colour (dark green) to minimise the reflection of light.
Vegetation Bank vegetation should be extended up to the edge of the culvert. This is especially
important if a lizard run or fauna path has been installed.
In a critical flood control region, this bank vegetation may need to consist only of flexible
(non-woody) species that provide minimal hydraulic resistance (ie. no shrubs).
Wet and dry cells Most culvert designs will have requirements for some degree of aquatic or terrestrial
movement. Thus, most multi-cell culverts will require both wet and dry cells (Photo B3
and Figure B2).
The dry cells may be elevated above normal bed level, or may be aligned with the
elevation of the adjacent floodplain. Dry cells can also have a dual use as pedestrian
tunnels and bikeways.
Photo B3 Multi-cell, multi-use, wet and dry cell culvert
Wildlife habitat As a general rule, the cells of a culvert should not be modified to provide or promote the
habitat of birds and terrestrial animals. The reason for this is that these habitats and the
residing animals will likely be destroyed during flood events.
137
However, culverts and the bed erosion often associated with culverts can provide
desirable habitat areas for reptiles.
Drop structures Drop structures can cause breaks in terrestrial movement corridors if the structure extends
across the full width of the channel and floodplain.
Breaks in terrestrial movement corridors can usually be prevented by providing rough,

sloping faces to drop structures, especially near the banks of the main channel and within
any floodplains.
Impacts on riparian Vegetation restrictions often exist immediately upstream and downstream of waterway
and floodplain bridging structures. In critical flood control areas, vegetation species that may interfere
vegetations with floodwaters should not be located within the following areas:
(i) upstream of a bridge or culvert within a radius equal to the total bridge or culvert
opening width;
(ii) downstream of a bridge or culvert within the zone defined by a 1 in 4 expansion of

the outlet jet and for a distance equal to three times the floodwater depth;
(iii) between the bridge or culvert opening and the floodplain;
(iv) any areas judged necessary by hydraulic modelling.
Bridges Traditionally, vegetation has not been placed under bridges because it is felt that there
would be insufficient light. However, with the appropriate selection of plant species, this
should not be the case in most circumstances.
If plants are introduced under bridges, whether on the abutments of or the banks of the
watercourse, then it is important for stormwater runoff from or around the bridge to be
directed to these plants.
Drop structures In the Brisbane region, a number of gabion drop structures have either suffered damage
to the wire baskets or have been invaded by undesirable weed species such as vines.
Vine invasion can be a problem in natural channels because the vines can damage native
trees. Gabion structures are generally recommended only if suitable vegetation can be
integrated into the design.
Public access, movement Public safety during storm and flood events is normally assessed based on the depth and
safety requirements velocity product, ie the product of the depth in metres times the flow velocity in metres
per second.
In most cases, the depth velocity product requirements will not apply to Natural Channel
Design works because it is normally unreasonable to expect the public to pass across a
waterway channel during a flood event. However, at recognised public access points and
crossings, such as pathways, bikeways, causeways, culverts and bridges, consideration
must be given to the public safety and thus the depth velocity product at these locations.
138
APPENDIX B
In all cases, public safety should be treated on a case-by-case situation giving due
consideration to Council policies, community expectations, engineering and economic
practicality, the risk of harm and an appropriate storm event or frequency.
Drop structures Drop structures can form a public safety hazard. In most cases, the maximum drop height
should be 1m for public safety reasons.
down the chute, the shorter the distance must be between rest areas (i.e. pools).
However, tailwater conditions must be checked to see when the chute becomes drowned.
139
Mannings Roughness
Table C.2 Modified Cowan method for determining channel roughness

Mannings n = (nb + n1 + n2 + n3 + n4) m
Channel condition n and m Description

values
Channel material Earth 0.020 Clay-based channels.

(nb) Bed rock 0.025 Channels cut into bed rock.
Sandfine gravel 0.024* Sandy creeks.
Coarse gravel 0.026 Gravel-based creeks (otherwise use Eqn C.1).
Degree of Smooth 0.0 Smooth channel.

irregularity (n1) Minor 0.0010.005 Excavated channels in good condition.
Moderate 0.0060.010 Channels with considerable bed roughness and some bank
erosion.
Severe 0.0110.020* Natural channels: pools and riffles, exposed tree roots,
boulders, and/or irregular banks.
Variation in Uniform 0.0 Near-uniform channel section.

channel Gradual 0.0010.005* Large and small cross sections alternate occasionally
cross section (eg. typical NCD n2 = 0.003).
(n2) Severe 0.0100.015 Large and small cross sections alternate frequently (eg. a
significant pool-riffle system).
Effect of Negligible 0.00.004 A few scattered obstructions (boulders, trees, logs)

obstructions that occupy less than 5% of the channel.
(n3) Minor 0.0050.015* Obstructions occupy 515% of the channel and the
obstructions are generally isolated.
excluding Appreciable 0.0200.030 Obstructions occupy 1550% of the channel.
vegetation Severe 0.0400.050 Obstructions occupy more than 50% of the channel (eg.
severe debris collection).
Amount of Small 0.0020.010 Grasses and/or weeds with the flow at least three
vegetation times the height of the vegetation.
(n4) Medium 0.0100.025* Grass and/or weeds with the flow one to two times the
height of the vegetation; or reeds or tree seedlings
Consideration growing with the flow two to three time the vegetation
should be given height; or minor bed vegetation with medium bank
to the obstruction vegetation.
caused by Large 0.0250.050 Grasses and/or weeds with flow depth equal to vegetation
vegetation height; or weedy beds with thick bank vegetation; or
relative to moderate shrub growth across the bed and banks.
channel width Very Large 0.0500.100 Grass and/or weeds more than twice the height of flow
and depth depth; or dense, strong reed growth; or significant shrub
growth within the channel; or significant inflexible vegetation
within channel.
Degree of Minor 1.00 Channel sinuosity is 1.0 to 1.2

channel Appreciable 1.15* Channel sinuosity is 1.2 to 1.5
meandering Severe 1.30 Channel sinuosity is greater than 1.5
(m) or; m = 0.57 + 0.43 (Sinuosity), but 1.30
(*) Typical NCD channel roughness n = (0.024 + 0.003 + 0.012 + 0.005 + 0.015) 1.15 = 0.068
142
APPENDIX C
Table C.3 Mannings n for a watercourse floodplain
Minimum Normal Maximum Description
A. Pasture, no brush
0.025 0.030 0.030 Short grass use design charts for grass
0.035 0.035 0.050 High grass use design charts for grass
B. Cultivated areas
0.020 0.030 0.030 No crop
0.040 0.040 0.050 Mature crop
C. Brush
0.035 0.040 0.070 Scattered brush, heavy weeds
0.050 0.060 0.100 Light brush and trees
0.070 0.080 0.160 Medium to dense brush
D. Trees (also refer to Table C.5)
0.080 0.100 0.110 Heavy stand of timber, a few fallen trees, little undergrowth,
tree branches above flood level.
0.100 0.120 0.150 As above, but with tree branches below flood level.
0.120 0.160 0.200 Dense tree cover
143
Mannings Roughness
Table C.4 Modified Cowan method for floodplain roughness

Mannings n = (nb + n1 + n2 + n3 + n4) m
Floodplain condition n and m Description

values
Floodplain Earth 0.020* Clay-based soil.

material Bed rock 0.025 Smooth, flat rock floodplains.
(nb) Sand 0.024 Sandy soils.
Gravel 0.026 Gravel-based soils (otherwise use Eqn C.1)
Degree of Smooth 0.0 Smooth, flat, floodplains.

irregularity Minor 0.0010.005 Slightly irregular shape. A few rises and dips.
(n1) Moderate 0.0060.010* Regular rises and dips.
Severe 0.0110.020 Very irregular floodplains. Pasture furrows perpendicular to the
flow.
Variation in
floodplain 0 Not applicable.
cross section
(n2)
Effect of Negligible 0.00.004 A few scattered obstructions (debris, stumps, logs, boulders)
obstructions occupying less than 5% of the floodplain flow area.
(n3) Minor 0.0050.015* Obstructions occupy 515% of the flow area.
excluding Appreciable 0.0200.030 Obstructions occupy 1550% of the flow area.

vegetation
Amount of Small 0.0020.010 Grasses and/or weeds with the flow at least twice
vegetation the height of the vegetation.
(n4)
Medium 0.0100.025 Grass and/or weeds with the flow one to two times the
Consideration height of the vegetation; or tree seedlings growing
should be with the flow two to three time the vegetation height.
given to the Large 0.0250.050 Grasses and/or weeds with flow depth equal to vegetation
obstruction height, or irregular shrub growth across the floodplain.
caused by Very large 0.0500.100* Grass and/or weeds more than twice the height of
vegetation flow depth; or significant shrub growth, woody weeds,
relative to or other inflexible vegetation growing across the floodplain.
the depth
of flow. Extreme 0.1000.200 Dense bushy shrub growth, or heavy stands of trees with
understorey vegetation and a few fallen trees, or a heavy stand
of trees with branches below flood level.
Floodplain
meander 1 Not applicable.
(m)
(*) Example calculation: n = (0.020 + 0.008 + 0.0 + 0.012 + 0.090) 1.0 = 0.130
144
APPENDIX C
Table C.5 Floodplain revegetation density guidelines for various

Mannings n roughness values
Mannings n Description
0.03 Short grass with the water depth >> grass height.
0.04 Short grass with water depth >> grass height on a slightly irregular earth surface. Trees at 10 metre
spacing, area is easy to mow.
0.05 Long grass on an irregular (bumpy) surface with few trees. Irregular ground could make grass
cutting difficult.
Alternatively, trees at 8 metre spacing on an even, well-grassed surface, no shrubs, no low
branches.
0.06 Long grass, trees at 6 metre spacing, few shrubs. The vegetation is easy to walk through. Area not
mowed, but regular maintenance is required to remove weeds and debris.
0.07 Trees at 5 metre spacing, no low branches, few shrubs, walking may be difficult in some areas.
0.08 Trees at 4 metre spacing, some low branches, few shrubs, few restrictions to walking.
0.09 Trees at 3 metre spacing, weeds and long grasses may exist in some locations. Walking becomes
difficult due to fallen branches and woody debris.
0.10 Trees at 2 metre spacing, low branches, regular shrubs, no vines. Canopy cover possibly shades
weeds and it is difficult to walk through.
0.12 Trees at 1.5 metre spacing with some low branches, a few shrubs. Slow to walk through.
0.15 Trees and shrubs at 1 metre spacing, some vines, low branches, fallen trees, difficult and slow to
walk through.
Alternatively, a continuous coverage of woody weeds with sparse leaves and no vines.
0.20 Trees and shrubs at 1 metre spacing plus thick vine cover at flood level and fallen trees. Very
difficult to walk through.
Alternatively, a continuous coverage of healthy shrubs and woody weeds from ground level to
above flood level.
145
Mannings Roughness
Photo C1
Straight, excavated, tidal channel.
Bed: n = 0.02
Banks: n = 0.06
Bankfull: n = 0.024
Photo C2
Slight meandering, regular cross section,
well maintained grass channel.
Bankfull: n = 0.028
Photo C3
Mown grass channel, regular cross

section, slight meander.
Bankfull: n = 0.028 (clean)

n = 0.030 (some shrubs)
146
APPENDIX C
Photo C4
Regular cross section, slight meandering,

mown overbanks.
Bankfull: n = 0.04
Overbank grass: n = 0.03 (shallow flow

depth assumed)
Photo C5
Mown grass banks, unmaintained

wetland plants on bed, regular cross
section, very slight meander.
Bed: Mannings n is variable depending

on flow depth.
Bankfull components:
bed n = 0.035
bank n = 0.030
resulting in a bankfull n =0.035
Photo C6
Canopy trees in early stages of growth,

straight, regular channel.
Bankfull: n = 0.04
Overbank: n = 0.15
147
Mannings Roughness
Photo C7
Rock size approx. 300 mm, this results in

a Mannings n = 0.034 assuming deep
water flow.
Bed: n =0.04
Photo C8
Deep channel, irregular cross section,

meandering channel.
Bankfull: n = 0.045
Photo C9
Near straight channel, full canopy cover

with few weeds, pool-riffle system,
shallow pools with boulders.
Bed: n = 0.045
Bank: n = 0.09
148
APPENDIX C
Photo C10
Pool-riffle bed system, meandering

channel, thick shrub growth on banks,
deep pools.
Bed: n = 0.04
Left bank: n = 0.06
Right bank: n = 0.20
Bankfull: n = 0.06
Photo C11
Channel vegetation in early growth

stage, gradual bends, regular cross
section, deep water, pool-riffle system.
Bankfull (existing): n = 0.04

Long-term (full vegetation)
bed: n = 0.05
banks: n = 0.15
Photo C12
Irregular, meandering, constructed

channel with boulders.
Bankfull: n = 0.05
Bank vegetation: n = 0.15
149
Mannings Roughness
Photo C13
Irregular channel with meanders and

woody debris (logs).
Bankfull: n = 0.05
Overbank: n = 0.10
Photo C14
Bed is a combination of thick, flexible

vegetation and open rock pools and
riffles. banks have sparse trees and
woody shrubs. Irregular channel shape
with slight meandering.
Bed: n = 0.06
Bank: n = 0.12
Bankfull: n = 0.07
Photo C15
Weedy channel passing through a long

grass floodplain. Irregular channel cross
section with some meanders.
Bankfull: n = 0.08 (assuming low velocity

and shallow depth that will not flatten
reeds)
Overbank: n = 0.03 to 0.10

(depends on flow depth and velocity)
150
APPENDIX C
Photo C16
Irregular mountain creek with flexible

understorey plants, few vines or woody
shrubs.
Bankfull: n = 0.10 to 0.12
Photo C17
Overbank vegetation at approximately 8

metre spacing with no shrubs.
Overbank: n = 0.05
Photo C18
Overbank vegetation consists of tall

truck trees, no low branches or shrubs.
Tree spacing of approx. 8 metres.
Overbank: n = 0.05
151
Mannings Roughness
Photo C19
Irregular channel with meanders.
Channel: n = 0.04 to 0.05 depending on

channel irregularity and debris content.
Overbank area consists of single truck

trees with no low branches or shrubs.
LHS (5 m spacing): n = 0.055
RHS (6-7 m spacing): n = 0.05
Photo C20
Trees at approx. 5 metre spacing, no low

branches.
Overbank: n = 0.055
Photo C21
Irregular natural channel and wetland

system with many weeds.
Overbank: n = 0.06
152
153
154
Printed on recycled paper

Brisbane City
For more information
Council Information
please telephone
GPO Box 1434
Brisbane Qld 4001
on (07) 3403 8888
www.brisbane.qld.gov.au K2000-452

NCD - Natura Channel Design

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NCD - Natura Channel Design

Uploaded by

Copyright:

Available Formats

6945 Nat Channel/Guide 4/5/01 2:43 PM Page 2

Natural Channel Design

Dedicated to a better Brisbane

Issue A December 2000 Original

Natural Channel Design

Brisbane City Council

ISBN 187609 141X

Natural Channel Design Guidelines

Natural Channel Design Guidelines

Executive Summary ..........................................................1 Step 9...............................................................................

Flood relief works ................................................................22 Step 10-8(b) .....................................................................

Step 3............................................................................... Planting patterns and hydraulic constraints.........................60

Natural Channel Design Guidelines

The principle of NCD is based on providing the required hydraulic conveyance of a

Unfortunately, we cannot have a totally natural creek in an unnatural catchment, but by

Chapter 3 presents a channel design procedure. This chapter concentrates on the

Chapter 5 provides step by step proceedures to evaluate NCD Projects. Monitoring

Natural Channel Design Guidelines

Choosing the appropriate site and deciding the extent of works

To minimise erosion, flooding and maintenance while maximising

Natural Channel Design Guidelines

How do the guidelines assist me

Flood mitigation Self-sustaining systems

DEVELOPMENT DEVELOPMENTS THAT: Guidelines are a referenced document

HIGH LEVEL OUTCOMES

There are no universally agreed-upon design guidelines/techniques that can be applied to

However, many of the relationships presented relate strongly to universally accepted

An ecologist is required to effectively incorporate the channel features described in

(i) channel geometry; and

Natural Channel Design Guidelines

The determination of minimum corridor widths during catchment and development

The rehabilitation or reconstruction of a channelised or otherwise degraded

Photo 1. Rehabilitated flood channel in Denver, Colorado

Preservation of an ecosystems integrity by conserving or promoting diverse

Development of a self-sustaining system to minimise long-term maintenance costs and

Photo 2. Habitat features added to a concrete drainage channel, Brisbane

(i) the Field of Dreams Method; and

Natural Channel Design Guidelines

Whichever method is used, it is important not to concentrate on pure aesthetics, but to

Unfortunately, we cannot have a natural creek in an unnatural catchment, but by

Following release of the 1996 guidelines Council embarked on a process of surveying a

D [m] Channel depth from bankfull water level to channel invert

Dp [m] Maximum pool depth

d/s [-] Downstream

N [-] Number of riffles within a reach

n [-] Mannings roughness value

q [m3/s/m] Discharge per unit width for a nominated design event

Qf [m3/s] Bankfull discharge

Qfds [m3/s] Bankfull discharge immediately downstream of the works

Qfus [m3/s] Bankfull discharge immediately upstream of the works

Natural Channel Design Guidelines

So [m/m] Average channel slope over a selected reach

Sr [m/m] Slope of downstream face of riffle

Si [-] Sinuosity = L2/L1 or L3/L2

u/s [-] Upstream

V [m/s] Average flow velocity

Wp [m] Maximum pool width measured at low flow water level

Wr [m] Riffle bed width

[deg] The angle measured between two consecutive inflection points

(iv) Select the section and extent of works.

Natural Channel Design Guidelines

Ecological requirements, including: riparian vegetation, shading, wildlife corridor

Aesthetic requirements, including: natural appearance, diversity of plant species,

When rehabilitating or reconstructing a watercourse there are various levels of work,