Anda di halaman 1dari 31

7

Marine Outfa 32. Marine Outfalls

Part C | 32
Peter M. Tate, Salvatore Scaturro, Bruce Cathers

32.2.3 Data Collection


Marine outfalls are used to discharge treated liquid
for Outfall Design ........................ 11
waste to the environment. Not all contaminants in
liquid waste can be removed by treatment. A prop- 32.3 Predicting Near-Field Dilutions .............. 12
erly designed, constructed, and operated marine 32.3.1 Physical Models........................... 13
outfall effectively dilutes the discharged waste 32.3.2 Positively Buoyant Jets
which then substantially reduces the concentra- and Plumes ................................ 13
tion of contaminants in the wastewater. In turn, 32.3.3 Negatively Buoyant Jets ............... 14
32.3.4 Model Validation ......................... 14
this reduces the risk to biota and human users of
32.3.5 Far-Field Numerical Modeling ...... 17
the marine environment. An introduction to some
32.3.6 Data for Running the Models........ 17
of the main aspects of marine outfalls is provided.
32.3.7 Conceptual Design
Five areas are covered, commencing with the for Positively Buoyant
main influences associated with the decision to Discharges .................................. 18
build a marine outfall. Included is an overview
of the wastewater treatment process. Near-field 32.4 Hydraulic Analysis and Design ................ 19
numerical modeling is described and it is demon- 32.4.1 Governing Hydraulics ................... 19
strated how this tool can be used to assist with 32.4.2 Diffusers Hydraulic Design ......... 21
32.4.3 Flow Variability ........................... 25
the design of a marine outfall. Outfall hydraulics
32.4.4 Hydraulic Integration ................... 26
is discussed, detailing a range of features includ-
32.4.5 Air Entrainment........................... 28
ing head losses, manifolds (or diffusers), seawater
32.4.6 Sedimentation ............................ 29
intrusion, and air entrainment. A very brief sum-
mary of the construction of a marine outfall is 32.5 Outfall Construction ............................... 30
provided. The final area covered describes envi- 32.5.1 Construction Materials ................ 30
ronmental monitoring that should be undertaken 32.5.2 Construction Methods ................. 30
to confirm the putative impacts associated with 32.5.3 Some Considerations ................... 31
a marine outfall. 32.6 Environmental Monitoring ..................... 32
32.6.1 Change Versus Impact .................. 32
32.6.2 Pre- and Postconstruction
32.1 Terminology .......................................... 8 Monitoring ................................. 32
32.2 Governance ........................................... 9 32.6.3 Long-Term Monitoring ................. 34
32.2.1 Drivers for a Marine Outfall .......... 9 32.6.4 Summary .................................... 35
32.2.2 Wastewater Treatment ................. 10 References..................................................... 35

Work presented in this chapter concentrates on the outline some initial considerations to help those new
discharge of wastewater to the environment through to the subject area. Understandably, the present chap-
marine outfalls. Marine structures are required for in- ter does not cover all areas in detail; the focus is on the
takes for drinking water (e.g., desalination plants) and design and monitoring aspects of marine outfalls. Infor-
water for industrial or commercial use (e.g., flushing of mation on some of the problems drawn from experience
toilets). The focus here is on marine outfalls; marine in- with marine outfalls is provided and reference material
takes are not considered further. with additional detail is identified.
The objective here is to provide practitioners with an
overview of the fundamentals of marine outfalls and to
8 Part C Coastal Design

32.1 Terminology
Part C | 32.1

The following terminology is used throughout this of each riser and number of outlet nozzles on each riser
chapter. Sewage is the raw input to a municipal wastew- will depend on the specific needs for each outfall.
ater treatment plant, the product of which is effluent. Some points to note are provided later:
Seawater is the raw input to a desalination plant, the
product of which is brine. Wastewater refers to either ef-  It is advantageous to locate the diffuser in fast flow-
fluent or brine. An outfall refers to the disposal system ing ambient waters. This will enhance dilution of
from the treatment plant to the discharge outlets. The the wastewater and rapidly transport the wastewater
manifold is that part of the outfall from which there is away from the diffuser.
a series of offtakes, termed risers. A diffuser comprises  For effluent discharge, the outlet nozzles are usually
that section of the outfall which includes the manifold horizontal. Effluent is less dense than the ambient
and the risers. marine waters and will rise to the sea surface or,
Two generalized marine outfalls are shown in if the stratification is sufficiently strong, become
Fig. 32.1. Figure 32.1a shows the discharge of pos- trapped below the surface. In contrast, outlet noz-
itively buoyant effluent from a wastewater treatment zles associated with a brine discharge are angled
plant and Fig. 32.1b shows the discharge of negatively toward the surface (often, an angle of 60 to the hor-
buoyant brine from a desalination plant. Both schemat- izontal is used). The density of brine is greater than
ics show an inclined tunnel from the wastewater treat- that of seawater and it will fall toward the sea bed.
ment plant, the outfall tunnel, a diffuser comprising Angling the outlet nozzles toward the surface and
several risers, and the wastewater plume emanating discharging the brine with high velocity will maxi-
from outlet nozzles on the top of each riser. The out- mize its dilution.
fall tunnel is inclined upward to ensure any air trapped  A pipeline may replace the tunnel as shown in
in the declined tunnel exits though the outlet nozzles Fig. 32.1. The pipeline is anchored to the sea bed
(or ports) and does not remain in the tunnel. The num- and discharge is through outlet nozzles fixed to
ber of risers, separation distance between risers, length the pipeline. Risers are not used in these con-

a) Wastewater
treatment plant

Near-field Far-field
Mean sea level
Plume Ambient Plume
current

Tunnel
Risers
Upwards sloping tunnel Diffuser

b)
Desalination plant

Mean sea level

Plume Plume
Pipe Pipe Nozzle
Diffuser

Fig. 32.1a,b Schematics of (a) a wastewater treatment plant outfall, and (b) a desalination plant outfall, showing side
views and end views
Marine Outfalls 32.2 Governance 9

figurations. (Risers are vertical structures used to ficiently high to prevent sediments from settling in

Part C | 32.2
transfer the wastewater from an outfall tunnel to the pipeline.
the outlet nozzles. They may be tens of metres in  Outlet nozzles may be fitted with nonreturn check
length). valves (also called duckbill valves). These valves
 The number of outlet nozzles attached to each riser are closed when the wastewater flow is zero and pre-
is usually restricted to eight or less. If more than vent the ingress of seawater into the pipeline. One
eight outlet nozzles are used, the plumes from ad- advantage of check valves is that they enhance di-
jacent nozzles interfere with each other and reduce lution, compared with a round nozzle of the same
the effective dilution of the wastewater. cross-sectional area [32.1]. However, they may be
 The outfall pipeline and diffuser may be tapered to fouled by biota or fishing nets, rendering them per-
ensure the velocity of the wastewater remains suf- manently open or closed.

32.2 Governance
There are many factors affecting the decision to build zone, which needs to be clearly defined prior to con-
a marine outfall. struction of the marine outfall.
Municipal wastewater collects at the bottom of the
catchment. For a coastal city, this is at the edge of the Social
marine environment. There are large costs associated What does the community expect from a marine out-
with the movement of wastewater to the top of a catch- fall? What are the values that are important to the
ment for potable reuse, including construction of a pipe community? This will vary among and within different
network, pumps, and energy required to operate the geographical regions and cultural groups. Some com-
pumps. Furthermore, there may be high costs associated munities will comprise a large number of beach users.
with the conversion of wastewater to potable water. The To them, the concept of a marine outfall may not be
disposal of wastewater through a marine outfall may palatable unless it can be clearly demonstrated that the
be the best overall use of resources. Despite this, the marine outfall poses minimal risk to their use of the ma-
decision to proceed with a marine outfall should first rine environment.
examine other options and maximize the beneficial uses
of recycled wastewater. Public Health
Is it safe to swim in the marine waters? What are the
32.2.1 Drivers for a Marine Outfall types and concentrations of substances that will be dis-
charged to the marine environment? Will they be of
SPHERE is an acronym we use to describe the main harm to us? Much information is available to inform
factors overlying the need for government sponsored us about of the potential harm of substances that may
development (social, public health, environmental, reg- be discharged through a marine outfall. Most countries
ulation, economic). The first three elements of SPHERE synthesize this information into a set of guidelines ap-
represent the main aspects in which a marine out- plicable to their marine environment. There is a tacit
fall has an impact (i. e., the community values). The assumption that, provided the concentrations of the sub-
last two elements of SPHERE represent the con- stances are kept below harmful levels, the health of the
straints on the marine outfall regulation tending users of the marine environment will be maintained.
toward high treatment and consequential high cost and This does require knowledge of the types, concentra-
economic tending toward low cost and consequential tions, and variability of the substances in the wastew-
low treatment. Below, some of the considerations of ater. It should be noted here that all the substances are
SPHERE are described in the context of a marine potentially toxic given sufficiently high concentrations
outfall. and the environment into which they are discharged.
Most countries have environmental guidelines that
need to be met during the design of a marine outfall. Environmental
These guidelines are unique to each country and all can- Will the discharge of substances through the marine
not be detailed here. Suffice to say that they include outfall cause harm to marine organisms? Will the ma-
meeting concentrations of contaminants which may in- rine environment be degraded into the future? Will the
clude pathogens, nutrients, metals, and organics. These beaches and marine waters be free from visible pol-
guidelines usually apply at the boundary of a mixing lution oil, grease, rags, etc.? As noted under Public
10 Part C Coastal Design

Health, most countries have environmental guidelines. are not isolated from the environment and infiltration
Part C | 32.2

Provided these guidelines are met, it is assumed that the of water during storms may also occur. The composi-
marine environment will be protected. The guidelines tion of wastewater depends on the relative contribution
are usually in the form of concentrations of substances of these three main sources and on the type and size
(e.g., metals, nutrients, and bacteria) that should be met of industry and/or commercial activity. Each wastewa-
at a specific distance from the outfall (this distance de- ter system is unique and treatment plants are designed
fines a mixing zone). This implies that there will be to deal with the quantity and quality of wastewater pro-
a region inside the mixing zone in which the guidelines duced by a specific system.
may not be met. The consequences are that the biologi- Wastewater comprises particulate matter, patho-
cal diversity inside the mixing zone may not be the same gens, nutrients, organic, and inorganic material. Severe
as that in reference areas. environmental damage can result if wastewater is dis-
charged undiluted or without treatment. Therefore the
Regulation main objective of sewage treatment is the elimination
What are the regulations that govern the discharge of or reduction in concentration of these materials.
any substance to the marine environment? Regulations Different concentrations of substances will invoke
are often in the form of licence conditions restricting different responses in different species. Metals may be
the types, concentrations, and/or loads of substances adsorbed onto particulates that may be ingested by fish
that can be discharged to the marine environment. As and shellfish. Organics are often adsorbed by the fatty
noted above, there is a tacit assumption that keeping tissues in aquatic animals. Reducing the concentrations
within these restrictions will ensure the safety of hu- of suspended solids, oil, and grease during the wastew-
mans, and the protection of flora and fauna in the marine ater treatment process, reduces the quantity of metals
environment. and organics that may affect marine organisms.
Wastewater treatment can be broadly divided into
Economic three levels: primary, secondary and tertiary (or ad-
Governments will invest a large amount of money for vanced). The levels are modular, subsequent treatments
the construction of a marine outfall. Ultimately, this being bolted onto lower levels of treatment. Within
money is raised through taxes and governments are ac- each level of treatment there are multiple options that
countable for the wise use of the taxes they collect. produce wastewater of similar quality. The distinc-
Outfall dollars will be competing with funding areas as tion among the treatment levels themselves is blurred
diverse as education, security, and care for the aged. and will depend on how individual levels are operated
What does the community value? What is the com- and maintained. Usually, concentrations of suspended
munity willing to pay to protect both humans and the solids, biochemical oxygen demand (BOD), and indi-
environment? The marine outfall is just one of many cator bacteria in the effluent are used to distinguish the
options that should be considered. Ultimately there is levels of treatment. The type of wastewater treatment
a balance between the level of protection offered and plant adopted is often based on the collective experi-
the cost incurred by each option. It is the responsibil- ence of the engineers and process workers within an
ity of the engineer and scientist to evaluate each option organization.
and provide the government with the most effective
solution. Primary Treatment
Primary treatment removes debris that could damage
32.2.2 Wastewater Treatment the wastewater treatment system. This is done by pass-
ing the sewage through trash racks and screens. Sewage
Our main focus in this section is on municipal outfalls. then flows through sedimentation tanks at low veloci-
Critical to a marine outfall is knowledge of what is be- ties ensuring residence times of 23 h or more [32.2].
ing discharged, particularly the types, concentrations, This allows sufficient time for negatively buoyant solids
and variability of contaminants in the wastewater. Dis- to settle at the bottom of the tank and positively buoy-
charge of contaminants from other sources including ant oils and greases to rise to the surface of the tank.
private outfalls, rivers and estuaries, atmospheric in- Chemicals can be added to the sewage to accelerate the
puts, discharges from vessels, and illegal dumping are settling process. Both the solids and oil and grease can
not considered. The reader is referred to Tchobanoglous then be easily removed. Primary treatment also helps
et al. [32.2], which provides considerable detail on regulate the flow of sewage to subsequent levels of
wastewater treatment. treatment. Primary treatment may be used in isolation,
Wastewater discharges from domestic, commercial, but this usually depends on the environment into which
and industrial sources. Often, the wastewater systems the wastewater is discharged.
Marine Outfalls 32.2 Governance 11

Table 32.1 Median concentrations of substances in sewage and after various levels of treatment. The numbers are indica-

Part C | 32.2
tive only and may vary in time and between sewage treatment plants
Substance Units Raw sewage Primary Secondary Tertiary
Faecal coliforms cfu=100 ml 107 106 104 10
Suspended solids mg=l 250 100 10 <5
BOD mg=l 200 100 10 <5
Oil and grease mg=l 50 20 <5 <5
Total nitrogen mg=l 50 40 20 10
Total phosphorus mg=l 10 7 5 3

Secondary Treatment brine are about 60 psu, although it may vary between 40
Secondary treatment covers a wide range of biological and 80 psu. The median salt content in seawater is about
processes including: activated sludge, trickling filters, 35 psu. Marine organisms can tolerate salt concentra-
rotating biological contactors, aerated lagoons, oxidiz- tions to about 39 psu [32.3], although this value varies
ing beds, and membrane bioreactors. The basic objec- with different organisms. Therefore, the configuration
tive of all of these processes is the removal of organic of an outfall discharging brine to the marine environ-
material and suspended solids. Secondary treatment ment should ensure a rapid reduction in salinity to less
may also include disinfection to reduce the concentra- than 39 psu.
tions of bacteria in the wastewater. A common form
of secondary treatment is activated sludge in which 32.2.3 Data Collection for Outfall Design
microorganisms are mixed with the wastewater under
aerobic conditions for about 48 h. The microorgan- While the preliminary design of a marine outfall can
isms metabolize the organic matter in the wastewater, be undertaken using minimum data, the detailed design
ultimately producing inorganic materials. usually requires considerable data. The main aim for
these data collection programs is identification of the
Tertiary Treatment site (or sites) for the marine outfall. That is, for the low-
Tertiary treatment often involves the further removal of est cost, identifying the level of treatment and site that
suspended materials using sand filters. High levels of best meets the environmental (and other) guidelines.
nitrogen and phosphorus may remain in the wastewa- The type and volume of data required depends
ter after secondary treatment, which can contribute to on the marine outfall being considered. Broadly, data
excessive primary production and eutrophication. The include: volume and flow rates of effluent to be dis-
basic premise for nitrogen removal is to convert nitrate charged, water quality (both in the treated effluent and
to nitrogen gas, which is then discharged to the atmo- in the marine waters), ocean currents, and stratification.
sphere. Biological processes and chemical precipitation A critical aspect of monitoring, often overlooked, is the
are two methods used to remove phosphorus from the variability of these data. Our favored approach is to use
wastewater. Once removed, phosphate can be used as the data variability in a Monte Carlo approach, running
a fertilizer. the models for many different combinations of input
Microfiltration and reverse osmosis are specific values. This results in a statistical distribution of the
forms of advanced wastewater treatment. The wastew- concentrations of contaminants in the marine waters,
ater is forced through a fine membrane. The size of the which can be synthesized in, for example, a probabil-
membrane mesh is sufficient to allow the passage of wa- ity of exceedance plot.
ter, but larger materials are captured and removed from Historical data or data collected from different
the wastewater. Increasingly, micro- or milli-filtration projects can be used. The difference between the data
are added to primary or secondary treatment processes. needs and the historical data defines a gap that a data
When combined with an effective outfall diffuser, the collection program needs to fill. Much of the data col-
diluted wastewater may achieve licence requirements. lected as part of these studies can also be used in
An indication of the median effluent concentrations Sect. 32.6. Some of the main data collection programs
of selected substances after treatment is given in Ta- are outlined below.
ble 32.1. It is stressed that these are general values that The volume of effluent flow can be estimated from
will differ for specific wastewater treatment plants and human population projections. This information allows
are highly variable. assessment of when environmental guidelines are likely
A desalination plant discharges brine in which the to be exceeded; hence when upgrades of the treatment
primary contaminant is salt. Median concentrations of plant are likely to be needed.
12 Part C Coastal Design

Water quality in the treated effluent will be a func- and remotely sensed data (e.g., via satellites or airborne
Part C | 32.3

tion of the level of treatment. Estimates can be obtained scanners). The number and duration of moored cur-
from other, similar treatment plants or from the indica- rent meters will depend on the size of the outfall under
tive values provided in Table 32.1. Ongoing monitoring consideration. For a moderately sized outfall, a sin-
of the effluent quality after construction will help ensure gle profiling current meter moored for 12 months and
maintenance of environmental standards. serviced monthly, provides the minimum data require-
Measuring water quality in the marine waters (into ments. A roving current meter (deployed at different
which the wastewater is discharged) provides back- locations for one month at a time) may provide a com-
ground concentrations of contaminants. The back- promise between the number of instruments and spatial
ground concentrations must be added to the modeled coverage.
concentrations to estimate the total concentration of Density stratification of the water column largely
contaminants in the marine waters. Background con- governs the height of rise of the wastewater. (The ef-
centrations may already exceed environmental guide- fect is much reduced for brine discharges). In coastal
lines, in which case, they made need to be relaxed marine waters, density is a function of both tempera-
or another outfall location sought. To obtain a rep- ture and salinity, both of which should be measured. For
resentative picture of marine water quality, sampling shallow outfalls (outfalls in water depths less than about
should take place over large spatial and temporal scales 10 m), stratification has little effect. However, for out-
and should include replication. Instruments that can be falls in deep waters, relatively small stratification may
moored in the field for long periods of time are increas- produce a submerged plume resulting in lower dilutions
ingly being used to obtain water quality measurements, and a nonvisible plume (at least to a surface observer).
although the accuracy of such results is less than can be Moored temperature/salinity strings provide a profile
achieved in the laboratory. of density throughout the water column, although ma-
Current speed and direction are important for plume rine fouling will reduce the quality of data from salinity
dilution. Moored current meters can provide detailed sensors. Such data can also be collected during the ser-
temporal information at a point in space (or a pro- vicing of moored instruments, which may be monthly
file throughout the water column). However, they are over a period of 12 months.
expensive to deploy, maintain and retrieve, and care- Other data such as surface waves and tides may also
ful consideration needs to be made in regard to the be important, particularly for shallow outfalls where
number and location of such moorings. Spatial in- the changes in water depth resulting from such pro-
formation can be obtained by profiling currents from cesses may represent a significant proportion of the
a vessel underway, drifters drogued at specific depths water column.

32.3 Predicting Near-Field Dilutions


The design of a marine outfall centers on the dilution Together with cost, the above factors are used to
required to meet the relevant guidelines. Occasionally, optimize the location and configuration of the marine
guidelines may be met after an appropriate level of outfall. This is further discussed in Sect. 32.3.7.
sewage treatment. However, many substances will rely After discharge from the marine outfall, efflu-
on the dilution with marine waters to meet these guide- ent rises (whereas brine descends) due to buoyancy
lines. Dilution depends on: (Fig. 32.1). The wastewater (effluent or brine) then
mixes with the ambient currents and is diluted. Two
 Wastewater flowrate types of models are used to quantify this process: near
 Depth of water into which the wastewater is dis- field and far field. This separation is made because the
charged time and space scales of the processes in each model
 Length of the diffuser are substantially different.
 Outlet diameter (and whether a single or multiple In the near field, the motion of the wastewater is
outlets will be used) dominated by its initial momentum and buoyancy; the
 Configuration of the diffuser (e.g., whether T-sec- velocities and rates of dilution are high. Up to 90% of
tion outlets or gas-burner type rosettes are used, wastewater dilution takes place within the near field at
whether nonreturn check valves are used) the end of which most regulations apply. The engineer
 Ocean conditions (e.g., currents, stratification of the can configure the outfall design to maximize dilution in
water column, tides, and ocean turbulence). the near field.
Marine Outfalls 32.3 Predicting Near-Field Dilutions 13

In the far field, the wastewater is passively trans- Tate and Middleton [32.1, 6]. Central to either approach

Part C | 32.3
ported by the ambient currents and the rates of dilution are the conservation equations for mass, momentum,
are much lower than in the near field. Far-field mixing is and buoyancy. In a Lagrangian framework, they are:
dominated by natural processes, over which the design
engineer has little control.
 Mass conservation
While the use of both near field and far-field models @ .V/
may be necessary for the detailed design of a ma- D a fent Uent A
@t
rine outfall, we argue that near-field modeling alone
may be adequate for the initial design and empha-  Momentum conservation
sis in the following sections is placed on near-field @ .Vui / @ .V/
modeling. D Ui C g0 V
This section provides a broad introduction to near- @t @t
field modeling. Wood et al. [32.4] provide considerable  Buoyancy conservation
detail on near-field modeling and many of the problems
that may be encountered in the design of a marine out- @ .g0 V/
D N 2 ui V ;
fall. @t

32.3.1 Physical Models where  is the density of the jet/plume, a is the density
of the ambient fluid, fent is an entrainment function, Uent
While the focus of this section is on near-field numer- is the entrainment velocity, V is the volume of a buoy-
ical modeling, it is recognized that physical modeling ant fluid element, A is the cross-sectional area through
can also play an important role in the design of a ma- which ambient water is entrained, ui is the velocity of
rine outfall. Scaled physical models of prototype marine the buoyant fluid, Ui is the velocity of the ambient fluid,
outfalls are sometimes constructed in the laboratory g0 is the buoyancy modified gravity (D .  g/ =ref ,
and used to examine the behavior of jets and plumes where ref is a reference density), and N is the Brunt
in the near field. They provide good visualization of Visl frequency
the plumes, particularly interactions between multiple s  
plumes and include effects that are not found in most g @a
ND  :
numerical models. ref @z
The fluids used in the model are typically fresh
and saline water. The scales for such models are ex- These governing equations are also applicable to nega-
pressed as a ratio of a prototype quantity to a model tively buoyant jets and plumes.
quantity. Model design requires the selection of (i) the If the buoyant jet/plume (a) lies well away from
fluids which yield the reduced gravity ratio (g defined its source (i. e., beyond the influence of the initial mo-
below), (ii) the length scale to ensure that the model mentum), (b) is moving with the ambient fluid, and
Reynolds numbers are sufficiently high to guarantee (c) the Boussinesq approximation is applied, then the
turbulent model flows and (iii) a scaling criterion which, above equations can be solved analytically to give what
in this case, is the densimetric Froude number (Fr), i. e., is known as the asymptotic results. These equations
there is a point-to-point correspondence of Fr in pro- (Table 32.2) are equivalent to the advected thermal
totype and model. This scaling criterion, together with equations in Wood et al. [32.4] and the corresponding
the length scale, yields the velocity scale. Other scales flow classifications are detailed in Jirka and Akar [32.7]
for time, pressure and buoyancy force can then be de- and Jirka and Doneker [32.8].
termined. The inclusion of ambient currents in physical Solutions to the asymptotic governing equations for
models is possible, but places greater demands on labo- positively buoyant plumes emerging from round (i. e.,
ratory facilities and data acquisition systems. axisymmetric) outlet ports and from a slot (i. e., line
Results from physical models may include infor- source), in a flowing ambient fluid, with both linearly
mation on dilutions, trajectories, the velocity field, and stratified or nonstratified marine waters, are presented
interactions between neighboring plumes. in Table 32.2. It should be noted that these asymptotic
solutions below should only be used at the conceptual
32.3.2 Positively Buoyant Jets and Plumes stage of outfall design. For preliminary and detailed
design, the full set of conservations equations above
Two basic approaches to near-field numerical modeling should be used and solved numerically.
are available: Eulerian and Lagrangian. A Lagrangian The entrainment function has evolved from the con-
approach is followed in both Lee and Cheung [32.5] and stants used in Morton et al. [32.9] to a complex function
14 Part C Coastal Design

Table 32.2 Solutions to the governing asymptotic equations for positively buoyant plumes when the ambient current
Part C | 32.3

speed is nonzero and the marine water density is linearly stratified (after [32.6]). Solutions to the asymptotic equations
are applicable only at the end of the near field and do not include the outlet port diameter, angle of discharge, or the exit
velocity
Axisymmetric source Line source
 1=3  1=2
BS 1cos. Nx
U / BS 1cos. Nx
U /
z.x/ D 0:98
2=3 UN 2
z .x/ D 1:00
1=2 UN 2
fent fent

2b .z/ D 2:00fent z 2b .z/ D 2:00fent z


CS nports CS
.z/ D C.z/ D 3:14fent
2 Uz2
Q S .z/ D C.z/ D 2:00fent Uz LQD
 1=3  1=2
BS BS
zmax D 1:24
2=3 UN 2
zmax D 1:41
1=2 UN 2
fent fent
 1=3  1=2
1=3 BS 1=2 BS
2b .zmax / D 2:48fent UN 2
2b .zmax / D 2:82fent UN 2
 1=3  1=2
CS 2=3 UB2S nports CS 1=2 UBS LD
S .zmax / D C.zmax / D 4:84fent N4 Q S .zmax / D C.zmax / D 2:83fent N2 Q

CS
z is the elevation of the plume above the outlet [m], zmax is the maximum elevation (i. e., height of rise) of the plume [m], S .z/ D C.z/
is the average dilution at elevation z, CS is the concentration at the source [kg=m ], C .z/ is average concentration at elevation z
3

[kg=m3 ], x is distance downstream from the outfall [m], fent is the dimensionless entrainment function, U is ambient current velocity
g
[m=s], N is the BruntVisl frequency [1=s], where N 2 D a d dz , and a denotes a representative seawater density, 2b.z/ is the
a

diameter (or thickness) of the plume [m], Q is the flow through the outfall [m3 =s], LD is the length of the diffuser [m], nports is the
total number of outlet ports on the diffuser and, BS is the buoyancy flux at the source, BS D g a
S
S Q
nports for an axisymmetric source
and BS D g a
S
S Q
LD for a line source, where S is the density of the wastewater at the source.

of the densimetric Froude number, plume geometry, the of interest along the jet trajectory are the centerline
velocity of the fluid inside the plume, and the veloc- peak (zm / and return point (xr / which are both defined
ity of the ambient current [32.5]. Wood et al. [32.4] use in Fig. 32.2. A range of experimentally derived val-
a spreading function to model the entrainment of ambi- ues for each of the proportionality coefficients compiled
ent fluid into the plume. from various experimental studies are presented in Ta-
ble 32.1, as reported in Lai and Lee [32.11]. Note that
32.3.3 Negatively Buoyant Jets these coefficients are only valid for single jets discharg-
ing into quiescent ambient conditions from a nozzle
Research conducted on negatively buoyant jets over orientated at 45 to the seabed. Coefficients for other
the past several decades has sought to quantify jet discharge angles can be found throughout the research
behavior using a variety of analytical and experimen- literature [32.1218]. Current research on negatively
tal techniques. Results from these studies have led to buoyant jets focuses on multiport diffusers and dis-
the development of proportionality coefficients which charge into receiving waters with ambient currents.
relate the jet densimetric Froude number and nozzle di-
ameter to trajectory and dilution. The particular points 32.3.4 Model Validation

The information presented in Table 32.2 and Fig. 32.3


z
are based on asymptotic models i. e., results only at the
end of the near field and should only be used at the con-
zt
xm ceptual stage of outfall design. Full numerical models
detail the movement of the wastewater from the outlet
nozzle to the end of the near field and include the noz-
zle size, the initial momentum of the wastewater and its
zm
x trajectory. A limited set of results from the laboratory
z0 xr experiments of Fan [32.10] for a single outlet, discharg-

ing positively buoyant water into a flowing, unstratified
ambient fluid are compared with several near-field mod-
Fig. 32.2 Trajectory of a negatively buoyant jet els that have been used by the authors. The models are:
Marine Outfalls 32.3 Predicting Near-Field Dilutions 15

Part C | 32.3
Dilution F = 10, k = 4 Dilution F = 20, k = 12
100 100

10 10

1 1
0 50 100 150 0 50 100
Nondimensional downstream distance Nondimensional downstream distance
Nondimensional height of rise Nondimensional height of rise
30 50
40
20
30
CORJET CORJET
PLOOM 20 PLOOM
10 OSPLM OSPLM
IMPULSE 10 IMPULSE
JETLAG JETLAG
0 0
0 50 100 150 0 50 100
Nondimensional downstream distance Nondimensional downstream distance

Fig. 32.3 Near-field model results for plume dilution and trajectory compared with laboratory data from Fan [32.10]

Table 32.3 Experimentally derived coefficients for a single negatively buoyant jet discharging into quiescent ambient
conditions at an angle of 45 to the seabed (after Lai and Lee [32.11])
Description Equation Experimentally derived coefficients
C1
Jet terminal rise height zt D DFr 1:43  C1  1:61
C2
Horizontal location of return point xr D DFr 2:82  C2  3:34
Dilution at return point Sr D CFr3 1:09  C3  1:55
C4
Vertical location at jet trajectory centerline peak zm D DFr 1:07  C4  1:19
C5
Horizontal location at jet trajectory centerline peak xm D DFr 1:69  C5  2:09

IMPULSE [32.19], JETLAG [32.5], CORMIX [32.7, it is recognized that different models may behave dif-
8], OSPLM [32.4], and PLOOM [32.1, 6]. Fans data ferently for different regimes and selection needs to
set is used here because it is independent of the labo- be appropriate to the problem under investigation. For
ratory data collected by any of the above authors. The example, a particular model may provide good esti-
unique identifiers for Fans experiments are Fr, the den- mates of dilution for outfalls comprising a single point
simetric Froude number discharge but poor estimates of dilution for outfalls
comprising a long diffuser with multiple risers and out-
uport
Fr D p ; let nozzles. It is stressed that there are other near-field
g0 dport models available [32.2022] that would likely provide
similar results.
where uport is the velocity through the outlet port, dport While laboratory studies are often used to calibrate
is the diameter of the outlet port and parameters within a model, field experiments are used
to validate the model predictions for a specific marine
uport outfall. This is undertaken using outfall dilution studies.
kD
U Obviously, such validation studies can only be carried
out after the marine outfall has been constructed. Out-
and fall dilution studies involve the continuous injection of
  a tracer into the wastewater and the measurement of
a  S its concentration downstream from the discharge point.
g0 D g :
S The tracer is injected at a known rate and concentration,
and the flow of wastewater is also known. Therefore, by
Compared with the laboratory data of Fan [32.10], measuring the concentration of the tracer in the marine
these models all produce similar results (Fig. 32.3) pro- waters, the concentration of the wastewater can be de-
viding confidence in the models themselves. However, termined.
16 Part C Coastal Design

Many tracers are available, for example: rhodamine individual risers have merged, the overall width of the
Part C | 32.3

WT, fluorescein and the isotopes gold-198, technetium- plume has increased and the concentration of the tracer
99m, and tritium. Natural tracers such as salinity have (or plume) has markedly reduced.
also been used, but the variability in the data is usually
too large to produce meaningful results. Preference is Problems with Tracer Studies
given to use a tracer that has little or no background Some problems encountered by the authors in conduct-
signal; hence contact with the tracer will result in un- ing tracer studies are outlined below.
ambiguous readings. The tracer sensing device (such as Some marine outfalls may have intermittent flow,
a fluorometer or scintillation counter) may be towed be- particularly early in the life of the outfall when the de-
hind a vessel and/or profiled through the water column sign flow capacity is not yet reached. With intermittent
to build a three-dimensional picture of the location and flow, the time history of the patch of wastewater is un-
size of the plume. A critical element of the work is accu- clear. However, it may be possible to temporarily store
rate position fixing, now usually done with differential the wastewater, to enable a continuous and steady flow
GPS (global positioning system). over the duration of the field experiment.
Simultaneously, the wastewater flow, ambient cur- Locating the plume in the field may be difficult. The
rent speed and direction, and the density of the water tracer may not be visible when the ambient waters are
column are measured. These data are used as input to stratified and the plume is trapped below the water sur-
the model. A direct comparison between the observa- face. A conductivity-temperature-depth probe can be
tions and model results can then be made. However, used to identify stratification in the water column and
models are only approximations to the real world and hence the likely depth at which the effluent will reside.
there will be uncertainty associated with the results. Isotope tracers decay with time. The half-life of
Based on the results from many such experiments, pre- technetium-99m is 6 h which is comparable with the du-
dicted dilutions within a factor of two of actual dilutions ration of many tracer experiments. If technetium-99m
are generally acceptable. is used as a tracer, the initial signal will change signif-
An example of the results obtained from a tracer ex- icantly over time and needs to be accounted for in the
periment is shown in Fig. 32.4. The transect lines run data analysis. Tritium, with a half-life of about 12 yr,
parallel to the diffuser, 100 (lower panel) and 1000 m can be used for long duration tracer experiments or
(upper panel) downstream from the outfall. Multiple when there is considerable transport time between the
transect lines are shown in each panel. In the lower nuclear facility that produces the isotope and the exper-
panel, plumes from each of the nine risers comprising iment site.
this outfall can be clearly identified. At a distance of When a positive contact is made with the labeled
1000 m downstream from the outfall, plumes from the plume, it is not possible to know where this contact oc-

1000 m downstream
0.004

0.003

0.002
Concentration of tracer (mg/l)

0.001

0
0 100 200 300 400 500 600 700
100 m downstream
0.004
Fig. 32.4 Example of tracer con-
0.003 centrations obtained from field
0.002
studies. Concentration data were
collected from 1 m below the sur-
0.001 face, at distances of 100 and 1000 m
downstream from the outfall. Note
0 the uneven distribution of concentra-
0 100 200 300 400 500 600 700
Distance (m) from start of transect tion along the diffuser indicating an
uneven distribution of flow
Marine Outfalls 32.3 Predicting Near-Field Dilutions 17

curs in the wastewater plume. One solution is to take the unknowns consist of two velocity components and

Part C | 32.3
many tracer readings closely separated in space and a depth. In a 3-D model, the mesh includes the 2-D hori-
time to identify the plume boundaries and the region zontal plane and the mesh points or nodes in the vertical
of highest tracer concentration. dimension. The unknowns at each 3-D mesh point or
The fluorescence of rhodamine WT is highly tem- node, typically consist of two horizontal velocity com-
perature dependent. It loses about 3% of its fluores- ponents and pressure.
cence for every one degree Celsius drop in water Usually, 2-D models require substantially less com-
temperature. In very cold environments, it may not be putational time and less data for calibration and running
possible to detect a signal at all as happened to the than 3-D models. In water depths exceeding about 20 m,
authors when first using rhodamine WT in Antarctica. the velocity vector at a single location in plan, may
vary in magnitude and direction throughout the water
32.3.5 Far-Field Numerical Modeling column. If resolution of this variability is considered
significant from the point of view of pollutant move-
The emphasis in this chapter is on near-field modeling ment, a 3-D model may be preferred to a 2-D model.
rather than far-field modeling. The reason for this is be- The horizontal spacing between mesh points or nodes
cause most of the dilution of the discharged wastewater will depend on the bathymetry and the presence of
occurs in the near field, and environmental guidelines islands, headlands, and submarine canyons. The near-
and licence conditions are usually applied at the end field model results need to be incorporated into the
of the near field. However, far-field modeling is impor- far-field model. To achieve this effectively, it may be
tant when assessing discharges into relatively shallow desirable to refine the far-field mesh in the vicinity of
waters when mixing in the near field is incomplete or the near field.
when examining potential impacts at sites remote from Far-field models run under various flow scenarios
the outfall, e.g., beach bathing waters or sensitive ma- can be used in the early stages of an investigation to
rine habitats or communities. guide data collection programs before, during, or after
Far-field modeling usually includes hydrodynamic commissioning of an outfall. Such model studies may
and water quality components. The hydrodynamic be conducted using a coarse mesh for quick turnaround
models are based on the principles of mass and momen- of results.
tum conservation of the marine waters; water quality
models are based on mass considerations of the contam- 32.3.6 Data for Running the Models
inant(s) or tracer(s) being discharged. Hydrodynamic
models are usually based on a fixed mesh in space (i. e., A range of information is required to run the numer-
an Eulerian formulation) and produce the depth and ve- ical models. This includes: the outfall configuration,
locity fields as output. The water quality models require wastewater flow, and oceanographic data (currents and
the velocity field as input; their formulation may be stratification of the water column). Over the long term,
based on the same mesh as the hydrodynamic model. In the model results can be used to examine changes in
another formulation (i. e., the Lagrangian formulation), outfall performance. Below is a summary of the in-
many parcels of contaminant or tracer may be tracked formation required to run the models and how that
as the velocity field transports and disperses them. The information may be obtained.
results from an Eulerian formulation yields the con-
taminant concentrations on the fixed mesh, while the Outfall Configuration
Lagrangian formulation yields the number of contami- The concept outlined in the following section can pro-
nant parcels contained in each volume of fluid bounded vide a starting point for the outfall design. In the
by mesh points or nodes; these numbers can then be design phase, the outfall configuration can be changed
converted into contaminant concentrations. and refined until the relevant environmental guidelines
The most common types of Eulerian models used are met and engineering feasibility assessed. Once the
are finite difference (i. e., point-wise approximations outfall has been constructed, its configuration is essen-
of the variables), finite elements (piecewise approxi- tially fixed. However, some flexibility may be enabled.
mations of the variables), or finite volume (based on For example, twin pipelines may be built and only
fluxes of mass or momentum within each mesh cell). one pipeline used for present wastewater flows (the
The meshes can be regular (i. e., structured) or irregular second pipeline being saved for use when wastewa-
(i. e., nonstructured); they can be 2-D (2-dimensional; ter flow increases with future growth in population).
i. e., depth averaged) or 3-D (three-dimensional). Similarly, a multiport diffuser may have one or more
In a 2-D, hydrodynamic model the mesh is in outlet ports blanked, again in anticipation of future
the horizontal plane. At each mesh point or node, growth.
18 Part C Coastal Design

Information needed for the outfall configuration in- density are dominated by changes in temperature and
Part C | 32.3

cludes: salinity. The height to which a wastewater plume rises


in the water column is largely governed by the strength
 Water depth in which the diffuser section is located. of the stratification.
 Length of the diffuser section. Measurements of temperature and salinity are often
 Configuration of the diffuser (e.g., a single or mul- made using a conductivity, temperature, depth (CTD)
tiport outlet). probe. (Salinity is calculated from conductivity and
 Diameter of each outlet port. temperature). The CTD probe can be lowered from
 Whether the outlet ports are fitted with nonreturn a boat providing a continuous profile through the wa-
check valves. ter column. CTD probes can also be moored, thereby
providing a time series of density data at a fixed point.
Wastewater Flow Historically, conductivity data from moored CTDs drift
Wastewater flow is usually measured in the outlet pipe with time due to the gradual build-up of film on the
at the end of the treatment processes. A range of flow sensors. While there have been substantial improve-
measuring devices are available including flows based ments in the reliability of moored conductivity sensors
on electromagnetic, pressure, ultrasonic, or capacitance in recent years, the quality of the data may still be
sensors. Also important is the density of the wastewa- highly variable. Temperature sensors (unless heavily
ter in relation to the density of the marine waters into fouled with marine growth) do not suffer the same prob-
which the wastewater is discharged. Usually it is safe to lem. Hence changes in stratification using data from
assume that the density of the wastewater is close to that long-term moored systems are usually estimated from
of fresh water [32.2], although large amounts of partic- temperature sensors alone.
ulate material in the wastewater may alter the density of
the wastewater. 32.3.7 Conceptual Design
For numerical modeling purposes, wastewater flow for Positively Buoyant Discharges
is usually assumed to be uniform throughout each outlet
port. This may not be necessarily the case. Energy loss Wilkinson [32.23] described a method by which the
may be significant over long diffuser sections, resulting minimum length of a simple outfall could be deter-
in reduced flows through outlets lying further offshore. mined. In his concluding remarks, Wilkinson [32.23]
Low flows may result in the intrusion of seawater into was careful to point out that this provides a preliminary
the diffuser and a reduction in its performance. To help estimate only, and he provided some suggestions on
establish uniform flow, diffuser sections may be tapered ways in which the outfall configuration could be further
(Fig. 32.1) and to help prevent the intrusion of seawater, refined. This analysis is only intended as a starting point
outlet ports may be fitted with nonreturn check valves. for outfall design. Detailed analyses are site specific and
must be undertaken for final design. Some site specific
Currents factors include: the bathymetry, environmental guide-
Currents determine the movement and dilution of the lines, level of wastewater treatment, and the likelihood
wastewater. Often a moored Doppler profiler is used of plumes reaching the surface or sensitive ecological
to measure the current speed and direction throughout areas. Wilkinsons [32.23] approach is modified here, by
the water column. Doppler profilers can also be ship using the single set of equations (Table 32.2) and intro-
mounted, which allows a spatial picture of the currents ducing construction cost as criteria for outfall design.
to be obtained. Remote sensing and shore-based radar The following analysis is applicable to nonzero am-
systems can provide detailed spatial coverage of the bient currents, which is usually applicable to marine
surface currents. However, it is subsurface current data waters (e.g., currents near the Sydney, Australia deep
which is critical for running the near-field models. water ocean outfalls exceed 0:05 m=s more than 90%
The choice of mooring location should be as close of the time).
as possible to the diffuser. However, a compromise is The total cost (Tc ) of a marine outfall can be ex-
often made, balancing the proximity of the diffuser to pressed as
the mooring, with the health of workers who service the
mooring (in waters that may be contaminated with di-
Tc D lLp C mLD C nnports (32.1)
luted wastewater) and the security of the mooring itself.

Stratification where l is cost per meter of the outfall pipeline or tun-


Stratification is a rapid vertical change in the density nel [$=m], Lp D length of the outfall pipeline or tunnel
of the marine waters. In coastal waters, changes in the [m], m is the cost per meter of the diffuser [$=m], LD
Marine Outfalls 32.4 Hydraulic Analysis and Design 19

is the length of the diffuser [m], n is the cost per outlet where S is the dilution required to comply with licence

Part C | 32.4
port [$], and nports is the number of outlet ports. conditions or environmental guidelines.
The basic premise used in Wilkinson [32.23] is that To minimize the total cost, the above expression
the profile of the water depths as a function of distance is differentiated with respect to the water depth (z),
offshore (i. e., the length of the marine outfall, Lp ) can equated to zero and solved. The result gives the depth
be expressed as the power curve, Lp D rzs , where r and at which the minimum cost for the marine outfall is
s are constants that express the least-squares, best-fit achieved. Substituting this value for depth into the
shape of the across shelf bathymetry that may be ob- equations in Table 32.2, gives the length of the diffuser
tained from navigational charts and z is the water depth. and the number of outlet ports that comprise the marine
Expressions for the length of the diffuser and the outfall.
number of outlet ports can be obtained from Table 32.2 Actual costs do not need to be known. If the rela-
and the total cost can then be rewritten as tive costs among l, m, and n are known, then the total
    cost of the marine outfall can be expressed in terms of
SQ 1 SQ 2 a normalized cost, Tc =l. Again, it is stressed that this
Tc D l.rz /Cm
s
z Cn 2
z ;
2fent U 3:14fent U analysis is preliminary and is only intended as a start-
(32.2) ing point for outfall design.

32.4 Hydraulic Analysis and Design


It is often necessary to define the physical extent of figuration is an example of parallel flow. For parallel
a brine or wastewater outfall system for project plan- flow, the total head loss must be equal through each par-
ning and design purposes. While the ambient sea in allel flow path.
the vicinity of an outfall structure typically defines the The energy equation applied between (0) and (1)
downstream boundary of an outfall system, defining the is written in terms of total head (i. e., energy per unit
upstream boundary may not necessarily be as straight- weight of effluent),
forward. For the purposes of this chapter, the upstream
boundary is assumed to be a free surface which exists V02 p0 V2 p1 X
C C z0 D 1 C C z1 C HL.0!1/ ;
somewhere upstream of the outfall conduit entrance. 2g e g 2g e g
Typical locations for this boundary could be the effluent (32.3)
level in outfall shafts, deaeration chambers, sedimenta-
tion basins, or pumping wet wells. For configurations where V is the velocity [m=s], p is the pressure [Pa], z
without any free surfaces between the treatment pro- is the elevation above an arbitrary datum [m], P zSL is the
cess and outfall conduit, the upstream boundary may elevation of sea level above datum [m], H L.0!1/
be taken at some hydraulically arbitrary point. Regard- is the total head loss between locations (0) and (1) [m],
less of its physical location, this boundary represents e is the density of effluent [kg=m3 ], a is the density of
a key design interface that must be properly integrated ambient seawater [kg=m3 ].
with the treatment plant as a whole. Determining the Working with gauge pressures and assuming negli-
piezometric head at the upstream boundary of an outfall gible effluent velocity in the outfall shaft, the first two
system is therefore a critical hydraulic design objective. terms on the left-hand side of (32.3) are reduced to
zero. Using the assumption that the pressure at (32.1)
32.4.1 Governing Hydraulics is hydrostatic based on the density of seawater, or p1 D
.zSL  z1 / a g, this relationship is substituted into (32.3)
Consider the gravity-driven outfall system shown in to yield an expression for effluent level in the outfall
Fig. 32.5 which includes an upstream shaft to cap- shaft, z0 .
ture plant effluents, as well as rosette style outfall     X
structures installed on the seabed. (A rosette structure V12 a
z0 D C .zSL  z1 / C z1 C HL.0!1/
typically has multiple nozzles that are arranged around 2g e
its perimeter.) The piezometric head at the outfall shaft, (32.4)
defined by point 0 on the fluid surface, can be obtained
by applying the energy equation between this point and Equation (32.4) demonstrates that the effluent level in
point 1 which is located precisely at the tip of the noz- the outfall shaft is a combination of (i) the head required
zle. It does not matter which outfall structure or nozzle to drive effluent out of the nozzle at the specified veloc-
is used to define point 1 because this multiple riser con- ity, V1 , (ii) elevation to the center of the exit port (z1 /
20 Part C Coastal Design

Fig. 32.5 Typical outfall system with


Part C | 32.4

Sea surface
0 (elevation zSL) rosette style outfall structures
Nozzle
e a 1
Rosette style
outfall structure
Riser
Diffuser
(or manifold)
Outfall shaft Conduit (pipe or tunnel)

plus discharge depth below sea level (zSL  z1 / scaled as well as storm surge. Statistical methods can be ap-
by the ratio of fluid densities, and (iii) a summation of plied to sea level time series at the project location in
head losses through the system. Each of these compo- order to determine exceedance probabilities and recur-
nents is briefly described in the following sections. rence intervals. Using sound engineering judgment in
conjunction with project requirements and/or local de-
Nozzle Exit Velocity sign standards for infrastructure design life, the results
The overall diffuser configuration including total num- of the statistical analysis can be used to select design
ber of ports or nozzles, and nozzle diameter (or exit values for minimum and maximum sea level. To cap-
velocity) is typically provided as an input to the hy- ture any seasonal trends which could include wind and
draulic design based on the results of near-field model- barometric effects, sea level data used in the statisti-
ing. The nozzle configuration affects the efficiency with cal analysis should include field measurements taken
which effluent is diluted in the near-field region and is regularly throughout the year. It is imperative that data
generally selected based on the maximum outfall flow specific to the project location is used because sea level
rate. In some outfall systems, such as those at desalina- characteristics can vary greatly from one locale to an-
tion plants, maximum nozzle exit velocities of the order other, regardless of the distance between them.
of 10 m=s may be required to ensure the brine is ade- An allowance for sea level rise due to the effects
quately diluted. The corresponding velocity head would of climate change should also be included because it
likely be the largest component of the outfall shaft wa- could have a significant impact on maximum outfall
ter level for exit velocities of this magnitude, especially shaft fluid level. Some statistical models estimate that
for a relatively short outfall conduit with low conduit the sea level will rise more than 1 m by the year 2100
friction loss. (Seneviratne et al. [32.24]).

Sea Level Head Losses


If seawater is discharged through the outfall system The total system head loss represented by the third term
.e D a / the second term of (32.4) simplifies to zSL of (32.4) consists of the sum of conduit friction losses
and the outfall shaft fluid level is equal to sea level for and the sum of local head losses through all fittings,
the no-flow case. For sewage outfalls in which e < a , system components (e.g., bends and contractions), and
the effect of the density difference is to increase the dividing flows in the manifold/diffuser. These losses
outlet shaft effluent level. Conversely, the level in the can be expressed as
outfall shaft is decreased when brine .e > a / is dis- X X V 2  Lc  X V 2
charged. Density differences tend to be of the order H L D c
fc C L
.KL /; (32.5)
of .je  a j =a /  100  3% for most sewage and de- 2g Dc 2g
salination applications. Although this difference corre- in which the first term on the right-hand side of the
sponds to a relatively small change to outfall shaft level equation is the DarcyWeisbach equation for conduit
when discharging into shallow waters, (32.4) shows that friction loss and fc D conduit Darcy friction factor ,
the density ratio effect is amplified for deeper outfall Dc D conduit diameter [m], Lc D conduit length [m],
discharges. Vc D velocity through the conduit [m=s], KL D local
In addition to changes in plant operating conditions head loss coefficient at fitting or component , and
which could increase or decrease outlet flows, changes VL D velocity through fitting or component [m=s].
in sea level will also cause the outfall shaft effluent level In (32.5), the term conduit refers to the tunnel or
to vary. The outfall system design should therefore con- pipe which delivers flow to the manifold and risers.
sider the entire range of sea levels that could occur over The Darcy friction factor can either be determined from
the project design life, accounting for tidal fluctuations manufacturers charts for particular wall roughnesses,
Marine Outfalls 32.4 Hydraulic Analysis and Design 21

computed iteratively using the implicit Colebrook ity. Using an incorrect reference velocity could result in

Part C | 32.4
White formula given in (32.6), or approximated using significantly higher or lower head losses.
the explicit SwameeJain equation given in (32.7). The For complex hydraulic systems, it is important to
SwameeJain approximation is accurate to within a few note that the total local head loss may not simply be the
percent of the value computed using the Colebrook sum of individual local loss components as (32.5) sug-
White equation over the typical ranges of roughness gests. Rather, head loss coefficients are typically subject
values and fully turbulent Reynolds numbers. to certain limitations. For example, the coefficient for
! a single tee-junction can only be applied to a series of
1 kS 2:51 tee junctions (such as in a dividing manifold) if the sep-
p D 2 log C (32.6)
fc 3:7Dc Rec  fc1=2 aration distance between successive junctions is, say,
0:25 5 to 10 times the manifold diameter. Correctly apply-
fc  h  i2 (32.7) ing local head loss coefficients will help to minimize
log kS
3:7Dc
C 5:74
Re0:9
under- or over-prediction of total local head loss. For
c
cases in which the loss coefficient limitations are not
where: kS is the Nikuradse equivalent sand grain rough- clearly defined or the system configuration cannot eas-
ness of the conduit wall [m], Rec is the conduit ily be broken down into standard components (such
Reynolds number [] D VcDc ,  is the effluent kine- as through a rosette-style outfall structure), physical
matic viscosity [m2 =s]. and numerical modeling can be used to confirm head
The kinematic viscosity of water for various tem- losses.
peratures and salinities can be obtained using the re-
lationships provided in Sharqawy et al. [32.25]. Wall 32.4.2 Diffusers Hydraulic Design
roughness values for common pipe materials can be
found in any hydraulics data handbook, while rough- Outfall systems usually consist of a manifold (also re-
ness values for segmentally lined tunnels are presented ferred to as a diffuser) whereby a common pipe or tun-
in Pitt and Ackers [32.26]. It is customary to make an nel supplies flow to multiple risers, ports, or branches.
allowance for increased wall roughness over the conduit Although a manifold is an example of parallel flow in
design life to account for aging and degradation. which the head loss between the outfall shaft and each
Local head losses arise from flow through bends, exit port is the same, it is critical to note that the flow
tee or wye junctions, flow or pressure control devices rate out of each port will not necessarily be the same.
(e.g., valves), and expansions or contractions in cross- The variation in flow rate can be attributed to (i) de-
sectional flow area. Local head losses will also occur creasing flow and total head along the length of the
at conduit entrances, at submerged discharges, and any manifold, (ii) changing depth along the manifold, and
location in the system where flow separation occurs. As (iii) head loss coefficients for tee junctions which are
shown in the second term of (32.5), local head loss is a function of (a) the ratio of conduit diameter to branch
expressed as a multiple of the velocity head at the par- diameter, and (b) the ratio of local flow through the con-
ticular component of interest. The local loss coefficient, duit to flow through the branch. Head loss curves for
KL , depends on the component geometry and is de- a range of tee-junction configurations can be found in
termined experimentally. Loss coefficients for common Miller [32.27].
system components can be found in any hydraulics data Given that effluent dilution is directly related to port
handbook. Several references such as Miller [32.27] exit velocity, the overall design objective should be to
and Idelchik [32.28] are devoted entirely to local head achieve equal exit velocities (or as close to equal as pos-
loss coefficients and include many components present sible) at each port to ensure consistent dilution levels
in marine outfall systems. over the length of the manifold. Traditional analyti-
As local head loss coefficients are always based on cal methods of solving for the hydraulic performance
some reference velocity, it is important to ensure con- of manifolds involve making an initial guess about the
sistency between a given KL and the velocity, VL , in the flow conditions at the most downstream port, then using
associated velocity head term (VL2 =2g). For components an iterative approach to progressively work upstream,
with constant cross-sectional area such as certain bends, port by port, until a final solution is reached. An alterna-
KL is normally based on the average velocity through tive approach using simultaneous equations is presented
the bend. However, there is no standard reference veloc- in this section. The resulting set of equations can be
ity for components like nozzles or sudden expansions quickly solved using a spreadsheet application with
or contractions which have multiple cross-sectional ar- built-in equation solver. The spreadsheet can be set up
eas; some sources may use the upstream velocity for to optimize port velocities by varying known quanti-
reference, while others may use the downstream veloc- ties such as the port and manifold diameters and/or port
22 Part C Coastal Design
Part C | 32.4

Total head line along tunnel

Head loss between Tn and Pn


0 Total head just inside nozzle

Tn Velocity head at nozzle


Tn1 Total head line for
T2 locations P1 Pn
T1
e
Pn Pn1 P2 P1

a
Pn Pn1
P2 Tn just upstream of riser
P1
just inside outlet nozzle
Tn Tn1
T2 Pn short distance outside nozzle
T1
zSL just downstream of riser

Datum
Diffuser (or manifold)

Fig. 32.6 Schematic for a manifold or diffuser flow calculation

spacing. Once the final manifold configuration is se- must be equal to the total head loss between points
lected, the same spreadsheet model can also be used to TiC1 and
determine the resulting port velocities and outfall shaft X 
fluid levels over a range of flow rates and/or effluent PiC1 HL.TiC1 !PiC1 /
densities.
(Table 32.4, (8k)). Note that exit loss should be in-
Consider the diffuser with n ports arranged along
cluded in the expression for the head loss between
a tunnel manifold as shown in Fig. 32.6. Points P1
Ti and Pi (Table 32.4, (8f)) because the velocity
through Pn correspond to locations downstream of the
head has been fully dissipated by the time the dis-
individual port openings, while points T1 through Tn are
charge has reached any point Pi . The total head at
located along the tunnel centerline, just upstream of the
all points Pi are equal.
port with the same subscript i. Flow conditions through
this system or any similar system can be solved Subscript Pi denotes individual ports, subscript
using the set of simultaneous equations provided in Ta- Ti denotes individual tunnel manifold sections, sub-
ble 32.4. Note that the values in brackets in the 3rd and script Pn denotes the most upstream port, and subscript
5th columns indicate that there is an equal number of Tn denotes the tunnel section between the outfall shaft
equations and unknown variables (8n C 1). The param- and port Pn (Fig. 32.6).
eters in the 6th column labeled inputs are assumed to In cases where loss coefficients KLTi!P in (8f) are
i
be known values. The governing principles reflected in functions of a flow ratio between the manifold and
this set of equations are: individual ports, the set of simultaneous equations in
Table 32.4 may need several iterations in which ad-
 Continuity the sum of individual port flow rates justments are made to the loss coefficients after each
(QPi / must equal the total outfall flow rate (QT /. iteration until the change in solution between succes-
(Table 32.4, (8a)) sive iterations is negligible.
 Energy equation the total head at point Pn must be Note that the head loss coefficients in (8f) of Ta-
equal to the total head at point 0, less
P the total head
ble 32.4 are presented in terms of the port exit velocity
loss between these two points H L.0!Pn / . (i. e., the velocity at which the effluent is discharged
(Table 32.4, (8j)) into the sea). For cases in which the port geometry
 Parallel flow the total head loss between points is more complex with varying diameters and multiple
TiC1 and head losses, all loss coefficients used in (8f) of Ta-
X  ble 32.4 must be converted such that they are based on
Pi H L.TiC1 !Pi / the port exit velocity. To convert a head loss coefficient
Marine Outfalls 32.4 Hydraulic Analysis and Design 23

Table 32.4 Set of simultaneous equations for manifold flow calculations

Part C | 32.4
Description Equations No. of Unknowns No. of Inputs Equation
equations unknowns
P
n
Continuity: total flow QT D QPi 1 QPi n QT (8a)
rate is equal to sum of iD1

individual port flow


rates
QPi
Port exit velocity VPi D  n VPi n DPi (8b)
4 D2Pi

P
i
QPi
Velocity in manifold V Ti D 1
 n V Ti n DT i (8c)
4 D2Ti
section between adja-
cent ports
VTi DTi
Reynolds number ReTi D  n ReTi n  (8d)
in manifold section
between adjacent ports
Darcy friction factor fTi D "
0:25
!#2 n fTi n ksi (8e)
ks
in manifold section log 3:7Di C 5:74
Ti Re0:9
T
between adjacent ports i

P VP2i
P  P P
Head loss between HL.Ti !Pi / D 2g KLTi !Pi n HL.Ti !Pi / n KLTi !Pi (8f)
manifold station i and
just downstream of
port i
P P
Head loss between HL.0!Pn / D 1 HL.0!Pn / 1 lTn (8g)
VT2n
 P 
outfall shaft and just l D
PTn
fTn DTTn C KL.0!Tn / C
downstream of the P2g n KL.0!Tn /
n-th port HL.Tn !Pn /
P P
Head loss between HL.TiC1 !Pi / D n1 HL.TiC1 !Pi / n1 lP
Ti (8h)
manifold station VT2i
 P  KL.TiC1 !T i /
lTi
(i C 1) and just down- 2g fTi DTi C KL.TiC1 !T i / C
P
stream of port i HL.Ti !Pi /

Pressure just down- pPi D zSL  zPi a g n pPi n zSL (8i)
stream of port i, zPi
expressed in terms a
of the ambient seawa-
ter density
V02 p0
Energy equation ap- 2g C e g C z0 1 z0 1 V0 (8j)
plied between point 0 VP2n pPn P p0 ;
and the most upstream D 2g C e g CzPn C HL.0!Pn / e
port (n/ zPn
P
Parallel flow: head H L.TiC1 !Pi / D n1 - - - (8k)
P
loss is the same be- H L.TiC1 !PiC1 /
tween manifold station
i C 1 and just down-
stream of either port i
or port i C 1
Total: 8n C 1 8n C 1

based on some velocity into an effective loss coefficient Manifold Section Diameters
based on another velocity, the following equation can The equations in Table 32.4 are kept general so that
be used, geometric parameters can be varied along the length
of the manifold. In some cases, it may be necessary
V12 V2 V2 to progressively reduce the manifold diameter in order
KL1 D 2 KL01 ) K L01 D 12 KL1 : (32.8)
2g 2g V2 to maintain velocities that prevent settlement of solids
24 Part C Coastal Design

Fig. 32.7a,b Tunnel velocity


Part C | 32.4

a) Tunnel velocity (m/s) b) Tunnel velocity (m/s)


variation along manifold
(a) constant diameter and
(b) stepped diameter . The
dotted line indicates a mini-
mum self-cleansing velocity;
Position (m) Position (m) refer to Sect. 32.3.7 for de-
tails

along the invert. (Due to the potential for excessively cients. In order to satisfy the principle of parallel flow
high head losses, it may not be possible to maintain suf- such that head loss between the outfall shaft and each
ficient self-cleansing velocities over the entire length of port is the same, the flow rate through each port will be
a conduit with a constant diameter.) Velocities through different.
a typical diffuser are shown for two cases in Fig. 32.7 The flow variation among ports is not necessarily
constant diameter and stepped diameter. The dotted line a problem from a hydraulic perspective. However, the
represents a nominal minimum velocity required to pre- near-field effluent dilution may become unbalanced if
vent solids from accumulating. It should be noted that the flow differential among ports becomes too great.
reducing the manifold section diameters is usually only This condition may prevent near-field dilution targets
feasible for applications involving piped outfalls. For from being met which could lead to adverse environ-
tunneled solutions in which sedimentation can occur, mental consequences. In the case of sewage outfalls in
an allowance for sedimentation build-up or a provi- which the effluent density is less than that of seawa-
sion for periodic removal of accumulated sediments ter, the variation in port discharges can create another
should be included in the design of the system. Refer undesirable condition seawater intrusion. One way
to Sect. 32.3.7 for a discussion on sedimentation. to prevent seawater intrusion from occurring is to use
smaller port diameters to ensure adequate port dis-
Port Diameters charge velocities. Modeling a less dense fluid flowing
The number of ports as well as the port exit velocity through an orifice into a more dense fluid shows that
required to ensure adequate dilution are typically de- seawater intrusion can be prevented when the port den-
termined from the results of near-field modeling. These simetric Froude number is greater than approximately
constraints effectively set the port diameter which, in 1.6 [32.29],
turn, establishes hydraulic performance of the system.
After the port diameter has been selected, the system VPi
Fr D r   > 1:6 : (32.9)
should be checked for excessively high head losses,
unbalanced flow distribution, and seawater intrusion g a
e
e
DPi

(for sewage outfalls only) to minimize the potential for


adverse operating conditions. In practice, an iterative In practice however, port densimetric Froude numbers
design process is usually required in order to strike an are typically kept well above this threshold value. For
appropriate balance between dilution performance and example, the port densimetric Froude number at the
favorable hydraulic conditions. Sydney deep water ocean outfalls is of the order of
Achieving a consistent discharge velocity at all 2030.
ports along a line diffuser is not necessarily a simple A sensitivity analysis of port diameters using the
task due to the varying head loss coefficients along the set of simultaneous equations in Table 32.4 shows
length of the manifold. An analysis of head loss coeffi- that port flow rates along the manifold do not vary
cients for dividing flows shows that they are a function significantly after port diameters are reduced below
of both the flow ratio and the area ratio, between some critical value. Moreover, reducing the port di-
individual ports and the corresponding conduit sec- ameter will increase the overall system head loss and
tion [32.27]. In general, loss coefficients are greater lead to higher outfall shaft effluent levels. In some
for larger flow ratios and smaller area ratios. Assum- cases, the maximum possible head that can be accom-
ing constant conduit and port diameters, upstream ports modated at the outfall shaft may dictate the smallest
will therefore have relatively low loss coefficients while allowable port diameters. If these diameters still do
downstream ports will have relatively high loss coeffi- not lead to sufficiently high velocities, an alternative
Marine Outfalls 32.4 Hydraulic Analysis and Design 25

configuration may be required. One option is to use design to provide a way for seawater to be bypassed

Part C | 32.4
variable-orifice nozzles such as the duckbill valves de- directly into the outfall system during periods of low
scribed in Sect. 32.4.3. outlet flow. The bypassed seawater effectively increases
Another option is to use rosette style outfall struc- the outlet flow rate and nozzle exit velocity, thereby
tures. A similar analysis of flow conditions through providing additional momentum for dilution by mixing.
a manifold configured with rosettes shows that port dis- The bypassed seawater also acts as an initial diluting
charges can be more evenly distributed. This better flow agent before effluent is discharged into the sea; as such,
balance occurs because head loss coefficients along the adequate levels of dilution may be obtained with lower
manifold are nearly uniform along its length, provided nozzle exit velocities for a given brine flow. The amount
the friction losses and local losses between the conduit of bypassed seawater required for any particular combi-
and risers are kept low. In this case, flow rates into each nation of outlet flow rate and salinity can be determined
rosette will be nearly equal. Furthermore, the flow rate using results of the near-field modeling and basic con-
through each port on a given rosette structure will be servation principles.
the same for axisymmetric rosette designs because the Although this option provides a near instantaneous
total head loss coefficient will be the same regardless level of control for managing the outfall system (when
of which port the effluent flows through. Rosettes can the appropriate flow control devices are included in the
have any number of nozzles, however, there tends to be design), the addition of bypassed seawater will lead to
an upper limit beyond which the additional nozzles will increased energy use and operating costs due to ad-
impede individual jet mixing processes. This behavior ditional pumping of seawater. These costs should be
occurs because neighboring jets tend to coalesce due to considered when deciding whether this option is an ap-
reduced pressures caused by entrainment of the ambi- propriate strategy for managing low flows. Depending
ent water between them. The result is to reduce overall on the frequency and amount of bypassing required, this
dilution performance. Near-field modeling can be used solution may, in fact, be found to be the most practical
to determine the maximum number of nozzles for a spe- or cost effective when compared with other alternatives.
cific rosette structure configuration. The incorporation of a seawater bypass system also pro-
vides a way to commission both the marine intake and
32.4.3 Flow Variability outfall systems independently from the rest of the plant.

The wide range of flows that may be experienced at Staged Construction of Marine Works
treatment plants is one of the greatest challenges in Many treatment plants are designed to be initially oper-
designing an outfall system. The preceding discussion ated at a low flow rate, then periodically upgraded until
assumes that the nozzle diameters have been sized to the maximum design capacity is attained. To manage
provide the required dilution for the maximum design the low flows associated with early stage plant opera-
flow rate, while at the same time maintaining sufficient tion, the marine construction works can be staged such
velocity to prevent salt water intrusion at the ports and that only some of the nozzles or rosette structures re-
avoiding problems arising from sedimentation. In real- quired for the ultimate flow case are initially installed.
ity, treatment plants operate over a wide range of flows, Additional nozzles, diffuser structures, or sections of
and it is quite possible that the design flow might not a line diffuser can be installed each time the plant is
be achieved until many years after the plant is initially upgraded; near-field modeling will inform the number
constructed. In addition, plant start-up and ramp-down of nozzles to be installed at each stage.
operations as well as any major changes to process The main disadvantage with this solution is that ad-
flow requirements could create situations in which the ditional marine construction works are required when-
plant will operate well below the design flow rate. The ever the plant is upgraded. Moreover, it assumes that
frequency and duration of any such events that re- once the plant is upgraded to ultimate capacity, it will
sult in low dilutions should be considered. If low flow not operate at the lower early stage flows. This assump-
scenarios are likely to result in sustained periods of in- tion may not be valid if demand can fall below the
sufficient dilution, the following mitigation schemes for plants maximum capacity after it is upgraded (for ex-
managing low-flow scenarios can be used. Each of these ample, lower demand from a desalination plant due to
options has advantages and disadvantages, and a hybrid a temporary period of increased rainfall). This solution
solution may turn out to be the most suitable approach. also does not address periods of low flow that will occur
even for ultimate capacity when the plant ramps up or
Supplemental Flows shuts down due to operations or maintenance require-
For desalination plants, a connection between the in- ments. These periods of low flows may not be an issue
take and outfall systems can be incorporated in the plant if they occur infrequently and for short durations, and if
26 Part C Coastal Design

environmental licences allow for temporary periods of periods of no flow. Due to stiffness of the valve mate-
Part C | 32.4

lower dilution. As such, plant processes and the feasi- rial, the valves gradually open in an elliptical shape as
bility of periodic marine construction works need to be the flow rate increases resulting in higher velocities at
considered should this option be pursued. lower flow rates as compared to fixed-diameter nozzles.
This eliminates or reduces the need for blanking nozzles
Nozzle Sizes and Blanks or introducing bypassed seawater into the outlet flows is
If it is not feasible to stage the installation of marine reduced or eliminated. The check valve feature of duck-
outlet structures or sections of a line diffuser, another bill valves also prevents backflow from the sea into the
option for managing low flows is to install blanking outfall system during periods of zero or low flow. Sea-
plates at the nozzle openings. This solution provides water intrusion into wastewater outfalls can therefore be
greater flexibility than staged construction because each prevented with this option.
nozzle can be brought online or taken offline individu- As material stiffness dictates the degree to which the
ally to suit the flow requirements. In addition, it allows valves will open for any given flow rate, these valves
the total flow rate to be evenly distributed since nozzles are usually custom designed for each particular appli-
on multiple structures or at any location along a line dif- cation. Hydraulic testing of the valves is recommended
fuser can be blanked off. Near-field modeling is needed prior to selection and incorporation into the final design,
to confirm the number and location of nozzles to be given the specificity of duckbill valves to each outfall
blanked off for each long-term operating flow rate. In- design. Hysteresis of the material should also be inves-
put from the plant operators should be sought when tigated to ensure that the valve will perform consistently
considering this option, as making adjustments to noz- over time. This consideration is particularly important if
zle blanks more than two or three times per year may flows are expected to be highly variable and/or prevail
render this option infeasible. for extended periods (i. e., months or longer). The local
A similar option is to install smaller diameter noz- conditions in which the valves will be installed is an-
zles for the initial stages, then remove and replace with other important aspect that should be taken into account
larger diameter nozzles each time the plant is upgraded. when deciding whether duckbill valves are an appro-
To minimize installation activities, each set of nozzles priate solution. If high levels of marine growth are ex-
should be designed to have the same entrance diam- pected, the valves may not close properly if organisms
eter such that modification to the manifold or rosette attach themselves to the area near the valve opening.
structures is avoided. This option provides the same
degree of flexibility as nozzle blanking if demand or 32.4.4 Hydraulic Integration
process requirements lead to an extended period of low
flow; the smaller nozzles can be re-installed at any Following a hydraulic assessment, the range of piezo-
time. metric heads at the upstream boundary of an outfall sys-
Although these options will also require marine tem must be considered with respect to the hydraulics
works each time the plant is upgraded, the scope and through the entire treatment plant. Failure to properly
duration of such works is likely to be considerably less integrate the outfall system with other plant systems
than for staged installation of marine structures or the could result in adverse operating conditions such as
addition of more diffuser length. These options also decreased plant capacity, spillage and flooding, and in-
carry some operational risk due to the fact that the efficiencies for pumped systems. The choice between
plant could still be operated at any given flow rate, even gravity-driven or pumped outfalls depends on several
with an incorrect number or size of nozzles. A nozzle factors including site topography and plant elevation,
configuration that is not aligned with the target operat- management of low flow scenarios and long-term oper-
ing flow could lead either to insufficient dilution (for ating or maintenance costs.
lower flow), or excessively high head losses through
the system (for higher flow). The operating procedures Gravity Systems
for these two options must ensure that plant capacity is Gravity-driven outfall systems are often preferred be-
linked to the installed nozzle configuration at any given cause they are relatively simple to operate (few electri-
time. cal and mechanical components) and have virtually no
long-term operating costs. Apart from periodic inspec-
Variable-Orifice Nozzles tion and maintenance, the passive nature of a gravity-
Variable-orifice nozzles, also known as duckbill valves, driven system ensures continuous reliability and mini-
can be used instead of fixed diameter nozzles. These mal operator intervention. The free surface that exists
valves are made from a variety of flexible materials such at an outfall shaft is convenient when outflows from
as rubber or neoprene and are designed to close during different parts of the treatment plant must be collected
Marine Outfalls 32.4 Hydraulic Analysis and Design 27

and combined into a single outfall stream. In addition, levels should not adversely impact the treatment plant

Part C | 32.4
a hydraulic break is usually required for certain process processes or reduce the plant operating capacity.
schemes such as removal of entrained air or settlement
of suspended particles; refer to Sects. 32.4.5 and 32.4.6 Vertical Drops
for a discussion of these issues. For cases in which the treatment plant is located at
The range of piezometric heads in the outfall shaft a significantly higher elevation than the outfall shaft
should first be considered in relation to the vertical fluid levels, a means of transferring the outflows down
alignment of the outfall conduit and ground surface a large vertical distance may be required. Allowing out-
level at the shaft location. Given the complexity asso- let flows to simply drop over a weir into the outfall shaft
ciated with installing an outfall conduit, the key factors causes a great deal of air entrainment and results in
in selecting its vertical alignment typically include suit- rapid dissipation of energy which can cause operational
ability of ground conditions, constructability, cost, and instabilities and structural damage if not properly man-
construction scheduling. For example, a trenched so- aged. A vortex, drop shaft system can be incorporated
lution which follows site topography and bathymetry into the design using methods such as those outlined
is often desired as it can be the cheapest and easiest in Hager [32.30] in order to gradually dissipate energy
to build. However, this alignment may cause adverse and minimize impact on the structure, and reduce the
hydraulic conditions such as hydraulic jumps and accu- potential for air entrainment.
mulation of air at localized high points. The potential As an alternative, a mini-hydroelectric scheme can
for such conditions should be assessed before finalizing be installed to recover the energy that would other-
the outfall conduit alignment. wise be lost by the vertical drop between the plant
In general, the entrance to the outfall conduit should and the outfall shaft. Whether the recovered energy is
be fully submerged for all flow rates and sea levels used to power other parts of the plant or sold back
in order to prevent operational instabilities. Otherwise, to the power grid, a substantial portion of the costs
hydraulic jumps would be likely to occur at the down- associated with pumping of intake flows (e.g., at a de-
stream end of unpressurized steep sections or where salination plant) may be recoverable. The potential cost
free surface flow conditions develop along an align- savings associated with any such mini-hydro scheme
ment that follows a ground profile containing inter- should be compared with the initial cost of all nec-
mediate peaks and valleys, as illustrated in Fig. 32.8. essary mechanical and electrical components, as well
Hydraulic jumps would likely be dynamic, unstable, as their ongoing maintenance and replacement costs.
and unpredictable, and could lead to air entrainment The frequency and duration of flow/head combinations
(Sect. 32.4.5 discusses the effects of air entrainment in should also be considered when determining viability of
greater detail). a mini-hydro scheme. While a large amount of energy
Regardless of whether the outfall shaft can be de- might be available for recovery during peak flow con-
signed and constructed to accommodate the range of ditions, lower operating flow rates over the long term
operating levels, these levels must also be compatible might render this option uneconomical. If a mini-hydro
with the upstream treatment plant systems. Ideally, the scheme is included, a passive system such as the vortex
plant is configured in such a way that shaft fluid levels drop shaft may also be required as an emergency by-
do not influence the upstream hydraulics. At the very pass or back-up in case the turbines need to be taken
least, any backwater effect caused by the outfall shaft offline.

Outfall shaft Outfall shaft

Hydraulic jump e
and aeration
Entrapped air
e
a
a

Hydraulic jump and aeration

Fig. 32.8 Examples of configurations that could entrain or entrap air and cause hydraulic jumps
28 Part C Coastal Design

Pumped Systems contractions and expansions. This condition could lead


Part C | 32.4

If specific site conditions render a gravity-driven sys- to reduced system capacity and the potential for ex-
tem unfeasible, a pumped solution will be needed to cessively high outfall shaft water levels (in a gravity
provide the driving head required for maintaining the system) or inefficient pump operation (in a pumped
required nozzle exit velocities. The hydraulic analysis solution). As it is typically not feasible to include air re-
for a pumped system is the same as for a gravity-driven lease mechanisms along an outfall conduit, trapped air
system because the pumps will need to deliver flow at pockets can remain in the system for extended periods
a head equal to the fluid level in a hypothetical out- of time.
fall shaft placed at the pumping station location. Pump Air pocket accumulation can also be problematic
selection should be based not only on the required even if there are no high points along the alignment.
piezometric head on the discharge side of the pump In the case of a conduit with constant slope (includ-
(which is a function of both flow rate and sea level), ing zero slope), entrained air will tend to come out of
but also on any flow- or time-based fluctuations of the solution and accumulate along the conduit soffit. De-
upstream water level or pressure. Selecting a single type pending on the conduit slope and diameter, and the
and size of pump to efficiently cover the entire flow and effluent flow rate, the air pocket may move upstream
head range may not be possible. However, using a sin- or downstream. Regardless of direction, the air pocket
gle type and size of pump is generally desired from an may not start moving until it reaches some critical vol-
operations and maintenance perspective because it adds ume or pressure, at which point it may move suddenly
flexibility and redundancy to the system. and violently. This sudden release of air at either end
Pumping can also be used as a way to manage low of the conduit results in a condition called blow-back
flow scenarios or ensure constant nozzle exit velocity or blow-out which can cause strong vibrations and even
over the entire flow range. This option provides near structural damage in extreme cases. Operational insta-
instantaneous control for managing nozzle exit veloc- bilities in the outfall system and farther upstream may
ities, but the associated operating costs will also be be experienced, especially in cases where the gradual
relatively high. Nonetheless, a pumped outfall system accumulation of air followed by a sudden release (i. e.,
may be required to manage low flows at a desali- gulping), becomes repetitive.
nation plant outfall when bypassing seawater is not It is important to recognize the processes by which
possible. air becomes entrained so that the system can be de-
signed to minimize the adverse conditions described
32.4.5 Air Entrainment above:

Managing air that becomes entrained in the treatment  Flow conditions at the upstream end of the out-
plant effluent stream is an important consideration for fall conduit can be a major contributor to entrained
outfall system design. If air bubbles are not captured air if the effluent is conveyed vertically through
and released before effluent enters the outfall conduit, a drop shaft, over a weir, or by way of any other
the bubbles could eventually agglomerate and form air hydraulic structure which causes a jet to plunge
pockets. Depending on the particular conduit configu- through air.
ration, air pockets can become trapped at localized high  Hydraulic jumps can occur in open channels at lo-
points along the outfall conduit and alter the hydraulic cations where the cross-sectional flow area or depth
performance of the system. Localized head losses can changes abruptly. It is even possible to have hy-
be induced at these locations due to the associated flow draulic jumps occur within the outfall conduit if the
upstream end is not fully submerged.
 For wastewater outfalls, gases can be generated over
Minimum fluid level the entire conduit length due to biological processes
in the effluent.
Air and effluent Effluent flows
directed upwards downwards, but air The two main approaches to removing entrained air
towards surface tends to stay near are: (i) remove it upstream of the entrance to the out-
surface for low fall conduit, and (ii) configure the outfall conduit such
Flow effluent velocities
that air bubbles either move upstream or downstream
without forming large air pockets. A deaeration cham-
ber or channel can be included immediately upstream
of the outfall conduit entrance to capture and release
Fig. 32.9 Indicative deaeration system using baffles air from the effluent in a controlled manner. The gen-
Marine Outfalls 32.4 Hydraulic Analysis and Design 29

eral idea behind these schemes is to provide a flow path very sensitive to fouling and require the removal of

Part C | 32.4
of sufficient length such that air bubbles have enough solids as well as organic compounds. Nonetheless, the
time to rise to the surface before entering the conduit. solid matter removed in the desalination pretreatment
Indicative air bubble terminal rise velocities are pro- processes may end up being added to the brine efflu-
vided in Falvey [32.31] and Lauchlan et al. [32.32]. To ent downstream of the reverse osmosis filters if not
minimize the deaeration chamber or channel flow path disposed of by other means. The design of an outfall
length, effluent velocities can be reduced by increasing conduit for a desalination plant must therefore consider
the cross-sectional flow area. Properly configured baf- all plant processes that discharge an effluent stream into
fle walls can also be used to direct flow upward and the outfall system.
facilitate air removal through the fluidair interface, or The build-up of granular sediments, organic matter,
to permit only the effluent at the bottom of a chamber and other solids along the invert of an outfall con-
or channel to continue downstream toward the outfall duit will eventually lead to increased head losses and
conduit; effluent near the invert is likely to have less decreased hydraulic capacity. In extreme cases, the con-
entrained air than that near the surface for low effluent duit section can become clogged and the function of the
velocities. Refer to Fig. 32.9 that shows an indicative discharge nozzles impeded. There are two general ap-
deaeration system using a baffled configuration. proaches to reduce the risk of sediment build-up, each
In other cases where site constraints limit the of which is described below. For some applications,
amount of space available for upstream deaeration a combination of these options may prove to be the most
schemes, or where a majority of entrained air is the effective solution.
result of biological activity occurring in the outfall con-
duit, an alternative approach can be taken. The conduit Minimize the Amount of Solids
can be designed to allow air bubbles to travel to ei- that Enter the Conduit
ther the upstream or downstream ends of the system. By providing a way for solids to settle out of the effluent
Falvey [32.31] presents a series of curves which sum- stream upstream of the conduit entrance, deposition and
marize a range of conditions for air bubble and air removal becomes a matter that can be managed locally
pocket movement in a closed and fully flowing conduit; at the plant site, rather than along the entire length of the
refer to Fig. 32.10 in which Q is the conduit flow rate, conduit. This approach using sedimentation basins or
D is the conduit diameter, and  is the angle of the con- settling tanks may be the preferred solution for cases in
duit to the horizontal. which the outfall conduit is long and/or inaccessible for
periodic inspections and cleaning. It may also be used
32.4.6 Sedimentation if it is not possible to design the conduit to achieve self-
cleaning velocities. The general design principle behind
Sediment management tends to be a more relevant is- this approach is that the sedimentation basins/tanks re-
sue at wastewater treatment plants than at desalination duce the effluent stream velocity to a point where the
plants; desalination inflow (i. e., seawater) generally has time it takes for suspended particles to settle is less
fewer solids than sewage, and its pretreatment processes than the residence time of the fluid in the basin or
generally remove a wider spectrum of material. Indeed, tank. Because particle settling time increases as the tar-
the filtration membranes in reverse osmosis plants are get particle size decreases, a potential drawback of this

sin
0.7
Air bubbles move
0.6 downstream large air
Air bubbles move pockets move upstream
upstream
0.5

0.4
Air bubbles and
0.3 pockets move
downstream D (m)
0.2
Q
(m 3
0.1 /s)
Reduced potential for blow-back

0 Fig. 32.10 Air movement in pressur-


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Dimensionless flowrate, Q2/(gD5) ized conduits (after Falvey [32.31,
Fig. 29])
30 Part C Coastal Design

option is that a large basin or tank footprint may be re- scenario. If, however, the chosen diameter leads to ex-
Part C | 32.5

quired. As such, this option may only be feasible for cessively high velocities and head losses during peak
removing larger solids. To decrease the basin/tank foot- flow, it may not be possible to achieve self-cleaning
print and increase its efficiency, internal baffle walls can velocities for all flow conditions. In this case, the de-
be installed to lengthen the flow path. However, the re- sign objective should aim for self-cleaning velocities on
sulting effluent velocity should remain sufficiently low a frequent basis (i. e., at least once per day) and for sus-
to allow sediments to settle out. tained periods of time (i. e., hours) in order to prevent
build-up of sediments and organic matter over time. The
Design for Self-Cleansing Velocities research literature contains a great deal of information
If upstream sedimentation control is not feasible, the on the velocities and shear stresses required to move
outfall conduit diameter can be selected to yield ve- solids through the conduit and to re-suspend solids that
locities that are high enough to prevent deposition of may have already settled. Velocities to keep solids in
solid materials. Given the wide flow variability asso- suspension range from approximately 0:52:0 m=s de-
ciated with many types of treatment plants, the outfall pending on conduit diameter, particle size and specific
conduit diameter should be based on the lowest-flow gravity, and particle concentration [32.33].

32.5 Outfall Construction


An effective design is not possible without input from the concrete. The density of concrete is about one-third
those engaged to construct the outfall. The design must that of steel (although still about double that of seawa-
be detailed in conjunction with how the marine outfall ter) and the concrete pipeline may need to be anchored
will be built. Grace [32.34, 35] and Wood et al. [32.4] to the sea floor.
provide good information on outfall construction ap- High-density polyethylene is relatively light, with
proaches and techniques. a density slightly less than that of seawater. Its big ad-
There are two fundamental types of wastewater de- vantage is that it is flexible and can be relatively easy
livery: via tunnels or pipes (either on trestles or in to deploy. Long sections are welded together on shore,
trenches). They can be used in combination, a tunnel the pipeline then towed into position and anchored to
for the bulk of the outfall length and a pipeline on the the sea floor. Purging of air from the pipe is critical to
sea bed comprising the outlet ports. The decision is prevent the pipe from floating back to the surface.
often governed by the geography of the region tun-
nels may be preferred in areas with rocky coastlines 32.5.2 Construction Methods
or where beach access is difficult, while pipelines may
be preferred in areas with easy access such as a sandy Pipelines are often manufactured in sections and assem-
shoreline. bled at an access point close to the proposed line of the
outfall. When in the water, each section is buoyed to fa-
32.5.1 Construction Materials cilitate easy movement. Offshore, a vessel is anchored
and pulls a section of the pipe into a trench along the
The materials used for outfall construction include steel, line of the outfall. The next section is welded on and
reinforced concrete, and high density polyethylene. The the pull continues. This is a popular technique for steel
marine environment is corrosive and steel structures pipelines, which are very strong in tension. Concrete
must be protected by coating the steel to prevent direct blocks or ballast rock may be used to anchor the pipe to
contact with sea water (often concrete is used) and/or us- the sea floor.
ing sacrificial anodes. Ongoing inspections are required The trench is then backfilled either mechanically
to monitor the integrity of the protective coating or the or by natural means. Techniques have been devised
anodes and replace them as necessary. whereby trenching, pipe-laying and backfilling are all
Steel reinforcement in concrete is susceptible to cor- done in a single operation. The end of the pipeline ter-
rosion by chloride salts in marine waters. To help in minates in a diffuser and the wastewater is discharged
preventing such corrosion, low permeability concrete through outlet nozzles.
is used, often in conjunction with additives that inhibit The active wave climate of the surf zone makes it
corrosion. Hydrogen sulphide in wastewater is corro- the most critical region. Often, a temporary trestle and
sive to concrete and concrete pipes are lined (usually sheet piling is used to protect the pipeline through this
with plastic) for protection. Regular inspections are un- region. Sometimes, wave action can damage the trestle
dertaken to ensure the lining has not peeled away from and sheet piling (Fig. 32.11).
Marine Outfalls 32.5 Outfall Construction 31

infauna displaced and particulate material placed into

Part C | 32.5
a)
suspension. Apart from the obvious physical destruc-
tion of habitats, suspended matter may reduce light in
the water column (affecting photosynthesis), clog the
gills of fish and, on resettling, may smother marine
plants and infauna.

Route selection
The shortest distance may not be the optimal route
for the outfall. Other influencing factors include: the
presence of rocky outcrops, wave and current climate,
maritime activity, fishing zones, and ecological consid-
b)
erations. The route of the Wollongong outfall (Aus-
tralia), constructed in 2005/2006, was changed in its
early design phase to avoid the habitats of the weedy
seadragon (Phyllopteryx taeniolatus) a local protected
species.

Wave and Current Climate


Waves can induce drag and lift forces on pipelines.
Breaking waves are most prevalent in the surf zone.
They can produce very large forces on a pipeline, al-
though their duration is short lived. Abnormally large
Fig. 32.11a,b Photographs from the same site, during (a) waves can be generated from storm surges, rogue waves
and after (b) a storm that caused extensive damage to sheet and tsunamis. Currents over a pipeline exert a drag
piling. The sheet piling was placed to protect the outfall force. The weight of the pipeline needs to be suffi-
pipeline (located in a trench) as it passes through the surf cient to prevent such forces from moving the pipeline.
zone Grace [32.34] provides some simple calculations that
estimate the force exerted on a pipeline by waves and
Tunnels are expensive to construct but their big ad- currents. In part, this problem can be largely overcome
vantage over pipelines is that they avoid the surf zone by burying the pipeline in a trench, which is then back-
where damage to a pipeline is more likely to occur. This filled.
is particularly important along high wave energy coast-
lines. Sediment Movement
One of the two tunnel construction techniques is The current and wave climate may resuspend sedi-
usually adopted: full face tunnel boring or drill and ments. Once in suspension, their movement and distri-
blast. The latter technique is more commonly used on bution can be widespread. Of particular importance is
shorter outfalls where the cost of a tunnel boring ma- the potential for sediments to erode from beneath the
chine cannot be justified. The drill and blast technique pipeline, potentially placing considerable stress on the
can be dangerous, releasing gases into the tunnel or pipeline itself. Conversely, an accretion of sediments
weakening the tunnel structure. may smother the pipeline inhibiting or preventing the
Long risers (tens of metres or more in length) are discharge of wastewater from the outlet ports. This may
drilled from the sea floor down to the tunnel. The risers also cause sediments to enter the outfall pipeline, reduc-
are capped with high velocity nozzles through which ing the efficiency of the outfall.
the wastewater is discharged. On the three Sydney deep
water outfalls, between four and eight outlet nozzles are Head Loss Monitoring
fitted in a rosette configuration to each riser. The hydraulic grade line for an outfall can be deter-
mined as a function of the discharge. At each discharge,
32.5.3 Some Considerations upper and lower limits for the hydraulic grade line
can be established and by maintaining the discharge
Environmental Impacts of Construction between these limits, we can optimize the hydraulic per-
During the construction of a marine outfall there will formance of the outfall. Below the lower limit, seawater
inevitably be environmental damage caused by trench- intrusion into the outfall is likely to occur and operators
ing and drilling activities. Habitats may be removed, can increase the wastewater flow accordingly (although
32 Part C Coastal Design

increasing the flow may not always be possible). The at high velocity. This also has the advantage of clearing
Part C | 32.6

addition of salinity sensors on individual risers can con- the outfall of sediments.
firm whether seawater intrusion has occurred.
Outfall Maintenance
Seawater Intrusion and Purging Regular inspections of the marine outfall should be car-
A rule-of-thumb for preventing seawater from enter- ried out as part of a regular maintenance program to
ing the outfall is to keep the port densimetric Froude ensure there is (a) no physical damage to any compo-
number (32.9) well above unity (e.g., on Sydneys nents and (b) no blockage of any outlet nozzles. This
deepwater ocean outfalls, the port densimetric Froude can be carried out by divers although this is accom-
number is of the order of 2030). Occasionally, the port panied by substantial health and safety concerns (e.g.,
densimetric Froude number may become small and sea- diving in contaminated waters, wave action, and decom-
water may enter the outfall. One method to help prevent pression requirements for deep diving). Our preference
such an intrusion of seawater is to use nonreturn check is to carry out inspections using remotely operated ve-
valves. However, not all outfalls are so equipped and hicles. The frequency of maintenance inspections will
it may be necessary to periodically purge the seawater depend on a range of factors such as environmental
from the outfall. This is done by backing up the wastew- conditions and construction materials. Often, annual in-
ater in the treatment plant and releasing the wastewater spections are adopted.

32.6 Environmental Monitoring


The fundamental objective of environmental monitor- 32.6.1 Change Versus Impact
ing is to quantify impacts that may arise as a result of
the discharge of wastewater to the marine environment. The marine environment is in a state of continual flux.
Some of the questions that an environmental monitor- The challenge of a monitoring program is to separate
ing program should address include: change that occurs via natural processes from change
that is a direct result of the discharge from the ma-
 Is it safe to swim? rine outfall (i. e., impact). It is critical to define what
 Is it safe to eat the seafood? constitutes an impact prior to the execution of the post-
 Will the marine communities be protected into the commissioning monitoring program. This places clear
future? bounds on the interpretation of change that may be ob-
 Will the beaches be free from contamination? served in the marine communities.

Two distinct environmental monitoring programs 32.6.2 Pre- and Postconstruction


are often implemented. The first is a pre- and postcon- Monitoring
struction monitoring program. This is usually an intense
program of short duration (perhaps five years) aimed Broadly, the pre- and postconstruction monitoring pro-
to quantify the initial impacts of the discharges from gram compares conditions before and after the marine
the marine outfall. The second is a long-term (ongo- outfall is commissioned. It is critical that sufficient time
ing) monitoring program, which usually takes a subset is allowed to carry out the before monitoring. An after
of the information obtained from the pre- and post- monitoring program can be conducted at any time af-
construction monitoring program, and uses it to define ter commissioning of the marine outfall. However, we
acceptable limits of change in the environmental indi- only get one chance at the before phase of the monitor-
cators. Long-term monitoring is designed to determine ing program.
whether a subsequent change lies outside these limits. A monitoring program may be of several years
There is an assumption, often tacit, that the en- duration. Consequently, the cost of an environmental
gineering aspects of a sewage treatment system and monitoring program may be large perhaps 5% of the
marine outfall are operating as designed. If there is construction costs may be needed to properly address
a treatment bypass, breakdown or blockage of some the main environmental issues.
of the outlet ports, the quality of the effluent (or its
delivery to the marine environment) will be less than Monitoring Philosophy
expected. Monitoring the operational performance is an Some of the characteristics that constitute a good mon-
important element of environmental monitoring. itoring program are outlined below. These characteris-
Marine Outfalls 32.6 Environmental Monitoring 33

tics combine to define our philosophy for environmental Sources of Contaminants. There are many potential

Part C | 32.6
monitoring. sources of contaminants discharged to marine waters,
Environmental monitoring programs should be de- including rivers, sediments (as both a source and sink),
signed to establish cause-and-effect between the dis- private and industrial outfalls, illegal dumping, and
charge and the environmental response. The approach coastal wastewater treatment plants. An important con-
we favor uses weight-of-evidence. This simply means sideration to environmental monitoring is separating the
applying different techniques and approaches to the relative contribution of an outfall from contributions
same problem. If the answers that they provide are from the other sources. It may be possible to charac-
consistent with what might be expected from such dis- terize (and isolate) the different sources via the types,
charges, then we have greater confidence in the overall concentrations, and variability of contaminants being
result. Our weight-of-evidence is a three-pronged at- discharged from each source.
tack, which is detailed in the following section. Marine organisms may be impacted if exposed to
Most countries have guidelines (or licence condi- wastewater. Toxicity testing is a technique used to de-
tions) that reflect the values that need to be protected. termine the concentration of substance which is likely
These values include social, public health and environ- to harm to marine organisms. Whole effluent toxicity
mental aspects. While some values may be subjective, (WET) testing examines whether a toxic response is
many can be quantified in terms of safe levels of identified when an organism is subjected to the com-
contaminants that can be discharged. There is a tacit plex wastewater matrix.
assumption, that by adhering to these concentration lim- If toxicity testing is carried out using a single test or-
its, there will be no irreversible damage done to the ganism, it is then inferred that all marine organisms will
marine environment. exhibit the same toxic response. This is not necessar-
The most critical task in undertaking a monitoring ily the case and multiple organisms are recommended
program to quantify environmental impacts of a marine for toxicity testing: different organisms display differ-
outfall is to ask the right questions. This requires an un- ent responses to different substances at different stages
derstanding of what is being discharged, what effects in their life cycle. Ideally, a range of fish, inverte-
these discharges are likely to have on the marine en- brates, and algae at different stages in their life cycle
vironment and the level of change that we are willing should be considered for toxicity testing. Toxicity tests
to accept. An answer to the wrong question, no matter can be either acute or chronic and both types of tests
how accurate is that answer, will not allow impacts to should be included. The former determine the concen-
be quantified. tration of wastewater that is lethal to the test organism.
Environmental monitoring programs must be sci- Chronic testing examines the reduced capability (e.g.,
entifically robust and defensible. We must test a hy- impaired development or reproductive ability) of the
pothesis and gather empirical evidence to support (or test organism.
disprove) the hypothesis. The experiment must be re-
peatable, cover a range of conditions and, in theory, Movement of Wastewater. Once we know what is be-
repeated experiments should arrive at the same con- ing discharged, we determine the path that the wastew-
clusion. Marine outfalls may be contentious and the ater will take after discharge and how dilute it becomes.
monitoring program may need to be defended. One Emery and Thompson [32.36] describe many techniques
way in which this can be done is via peer review of to measure and/or monitor the physical properties of the
the work and publication of the results in a reputable marine environment.
journal. Instrumented moorings provide excellent temporal
Conditions of approval for a marine outfall are often coverage although the horizontal spatial coverage is
reliant on predictions. The objectives of an environmen- limited by the number of such systems that the mon-
tal monitoring program should include the verification itoring program can afford to deploy. Remote sensing
(or otherwise) of those predictions and whether the provides good spatial coverage of the region, although
environmental values have been maintained. Further, in general, the temporal scales are much longer than
the program should include a mechanism by which those associated with the movement of wastewater
problems identified in the monitoring program can be plumes.
rectified or mitigated. Numerical modeling enables us to predict how the
wastewater will react to different marine conditions and
The Three-Pronged Attack to changes in the volume or method of discharge. It is
An approach to collecting weight-of-evidence for im- critical to calibrate and validate any numerical model
pact assessment includes three components. These are to ensure confidence in the results under existing con-
detailed below. ditions and for future scenarios. Coupled with this is
34 Part C Coastal Design

a need for confidence limits to be placed on model than this and the light attenuation starts to inhibit
Part C | 32.6

results. Too often model results are provided without the growth of organisms on the settlement panels).
a real appreciation for the confidence that can be placed  Fish, shellfish, and planktonic communities move
on those results. in, and with, marine waters. The variability among
control sites may be as much as the variability be-
Quantifying Change. There are a number of ap- tween putatively impacted sites and control sites.
proaches to quantifying change in response to a marine Therefore, it may be difficult to isolate the marine
outfall. One favored by the authors is the beyond BACI outfall as the cause of change in such communities.
or mBACI approach [32.37]. BACI is an acronym for  Sediments have been used to assess the accumu-
before, after, control, impact. Sites close to, and remote lation of contaminants and to examine infauna
from, the outfall are monitored on several occasions community variations. Our experience is that such
both before and after commissioning of the marine out- studies have limited success. Unless the contami-
fall. Perhaps the main reason that this is a favored nant signal is very strong, it is unlikely to register in
approach is its emphasis on statistical power and sta- a sediment sample.
tistical error.  Bioaccumulation studies are sometimes recom-
mended as impact assessment indicators. However,
Statistical Errors. A Type I statistical error occurs caution is needed when interpreting the results of
when the results of our analysis incorrectly predicts that such studies because:
a change has occurred. We can protect against making Fish move, and it is not always possible to know
such a mistake by specifying the level of significance. the region from where contaminants were ac-
Usually this is set at 5% (equivalent to 95% confidence cumulated. This necessitates fish home range
limits), i. e., there is a 5% chance that our analysis studies, which can be expensive and, perhaps,
makes a Type I statistical error. inconclusive.
A Type II statistical error occurs when the results Species that are caught at one particular time
of the analysis incorrectly predicts that a change has and location may not be caught at other times
not occurred. From an environmental point of view, this or locations and pooling into higher biologi-
is more insidious than a Type I error because it incor- cal levels may be required. Different species
rectly leads us to believe that there is no environmental may accumulate different substances at different
problem. It is difficult to protect against a Type II er- rates and the pooling process may mask poten-
ror as this requires a priori knowledge of the variability tial impacts.
in the system being measured. Such knowledge is ac- Moored systems comprising oysters or mussels
quired only after the monitoring is complete. Therefore, are often used in bioaccumulation studies. How-
we need to use experience to estimate the system vari- ever, this may involve removing the animals
ability and design experiments accordingly. To protect from their natural habitat thereby, adding stress
against making a Type II statistical error, we usually de- to the organisms and confounding the results
sign our experiments to target statistical power above that are obtained.
80%. However, if statistical power is too high, very
small changes become statistically significant and we 32.6.3 Long-Term Monitoring
question whether such small changes are meaningful. It
is critical to check the statistical power after the experi- The pre- and postconstruction monitoring program will
ments have been completed. identify whether there has been a step change in the
Type III statistical errors occur when we arrive at baseline conditions as a result of the discharge from
the correct answer but have asked the wrong question. the marine outfall. Long-term monitoring is used to
Some of the marine components measured to de- identify whether further change occurs well after the
tect change as a result of a new marine outfall include marine outfall has been commissioned. It can be used
water quality, sediment quality and community studies to extrapolate trends and, where necessary, design and
(such as intertidal, plankton, pelagic, benthic, and ses- implement an appropriate mitigation strategy to pre-
sile communities). Marine community studies should vent, or reverse, the trend. The three-pronged attack
be tailored for specific outfalls and might include: described in Sect. 32.6.2 is also applicable to long-term
monitoring.
 Intertidal community studies for shoreline or short However, there is a temptation to simply implement
outfalls. the bulk of the pre- and postconstruction monitoring
 Subtidal settlement panel studies for marine outfall program as part of the long-term monitoring program.
in waters less than about 20 m depth (much deeper Apart from the expense associated in maintaining a de-
Marine Outfalls References 35

tailed monitoring program over a long period of time, 32.6.4 Summary

Part C | 32.6
environmental changes will be masked by the small
time steps between consecutive sets of readings. It may This chapter provides a synopsis of the discharge of
be more effective to implement a low level, long-term wastewater to the marine environment. Flows of ef-
monitoring program and revisit the detailed monitoring fluent from municipal sewage treatment plants and
program, for example, over two consecutive years ev- brine from desalination plants are primarily considered
ery 10 yr. If the long-term program indicates a potential within the context of near-field modeling and outfall hy-
problem, it provides the motivation for a more detailed draulics. The designers of outfalls are under increasing
investigation. pressure from social, public health and environmental
Where such an approach is adopted, source charac- constraints, within a regulatory and economic frame-
terization combined with numerical modeling are used work, all of which need to be considered. This compact
to estimate environmental impact. If the results indicate overview exposes the reader to fundamentals of outfall
a possible impact, confirmation studies can be imple- design and identifies some of the traps and problems
mented. that may arise.

References

32.1 P.M. Tate, J.H. Middleton: Buoyant jets of elliptic 32.14 S. Nemlioglu, P. Roberts: Experiments on dense
shape: Approximation for duckbill valves, J. Hy- jets using three-dimensional laser-induced fluo-
draul. Eng. 130(5), 432440 (2004) rescence (3DLIF), Proc. 4th Int. Conf. MWWD, Antalya
32.2 Metcalf & Eddy Inc., G. Tchobanoglous, F.L. Burton, (2006)
H.D. Stensel: Wastewater Engineering: Treatment 32.15 G. Kikkert, M. Davidson, R. Nokes: Inclined nega-
and Reuse, 4th edn. (McGraw-Hill, New York 2003) tively buoyant discharges, J. Hydraul. Eng. 133(4),
32.3 P. Palomar, I.J. Losada: Impacts of brine discharge 545554 (2007)
on the marine environment. Modelling as a pre- 32.16 D. Shao, A.W.-K. Law: Mixing and boundary inter-
dictive tool. In: Desalination, Trends and Tech- actions of 30 and 45 inclined dense jets, Environ.
nologies, ed. by M. Schorr (InTech, Rijeka 2010) Fluid Mech. 10(5), 521553 (2010)
pp. 279310 32.17 I.G. Papakonstantis, G.C. Christodoulou, P.N. Pa-
32.4 I.R. Wood, R.G. Bell, D.L. Wilkinson: Ocean Disposal panicolaou: Inclined negatively buoyant jets 1: Ge-
of Waste (World Scientific, Singapore 1993) ometrical characteristics, J. Hydraul. Res. 49(1), 312
32.5 J.H. Lee, V. Cheung: Generalized Lagrangian model (2011)
for buoyant jets in current, J. Environ. Eng. 116(6), 32.18 I.G. Papakonstantis, G.C. Christodoulou, P.N. Pa-
10851106 (1990) panicolaou: Inclined negatively buoyant jets 2:
32.6 P.M. Tate, J.H. Middleton: Unification of non- Concentration measurements, J. Hydraul. Res.
dimensional solutions to asymptotic equations for 49(1), 1322 (2011)
plumes of different shape, Boundary-Layer Mete- 32.19 V.H. Chu: L.N. Fans data on buoyant jets in cross-
orol. 94(2), 225251 (2000) flow, J. Hydraul. Div. 105(4), 612617 (1979)
32.7 G.H. Jirka, P.J. Akar: Hydrodynamic classification 32.20 P.J. Roberts, H.J. Salas, F.M. Reiff, M. Libhaber: Ma-
of submerged multiport-diffuser discharges, J. Hy- rine Wastewater Outfalls and Treatment Systems
draul. Eng. 117(9), 11131128 (1991) (IWA Publishing, London 2010)
32.8 G.H. Jirka, R.L. Domeker: Hydrodynamic classifica- 32.21 B. Henderson-Sellers: Modeling of Plume Rise and
tion of submerged single-port discharges, J. Hy- Dispersion-The University of Salford Model: USPR,
draul. Eng. 117(9), 10951112 (1991) Lecture Notes in Engineering, Vol. 25 (Springer,
32.9 B. Morton, G. Taylor, J. Turner: Turbulent gravi- Berlin Heidelberg 1987)
tational convection from maintained and instan- 32.22 H.B. Fischer, E.J. List, R.C.Y. Koh, J. Imberger,
taneous sources, Proc. R. Soc. A 234(1196), 123 N.H. Brooks: Mixing in Inland and Coastal Waters
(1956) (Academic, New York 1979)
32.10 L.-N. Fan: Turbulent Buoyant Jets into Stratified or 32.23 D.L. Wilkinson: Optimal design of ocean outfalls. In:
Flowing Ambient Fluids, Technical Report (CALTECH, Environmental Hydraulics, ed. by C.C. Lai, J.H. Lee
Pasadena 1967) (Balkema, Rotterdam 1991) pp. 275279
32.11 C.C. Lai, J.H. Lee: Mixing of inclined dense jets in 32.24 S.I. Seneviratne, N. Nicholls, D. Easterling,
stationary ambient, J. Hydroenviron. Res. 6(1), 9 C.M. Goodess, S. Kanae, J. Kossin, Y. Luo,
28 (2012) J. Marengo, K. McInnes, M. Rahimi, M. Reich-
32.12 P.J. Roberts, A. Ferrier, G. Daviero: Mixing in in- stein, A. Sorteberg, C. Vera, X. Zhang: Changes in
clined dense jets, J. Hydraul. Eng. 123(7), 693699 climate extremes and their impacts on the natural
(1997) physical environment. In: Managing the Risks of
32.13 A. Cipollina, A. Brucato, F. Grisafi, S. Nicosia: Extreme Events and Disasters to Advance Climate
Bench-scale investigation of inclined dense jets, Change Adaptation, ed. by C.B. Field, V. Barros,
J. Hydraul. Eng. 131(11), 10171022 (2005) T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mas-
36 Part C Coastal Design

trandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, 32.33 D. Butler, R.W.P. May, J.C. Ackers: Sediment trans-
Part C | 32.6

M. Tignor, P.M. Midgley (Cambridge Univ. Press, port in sewers part 1: Background, Proc. ICE-Water
Cambridge 2012) pp. 109230 Marit. Energy, Vol. 118 (1996) pp. 103112
32.25 M.H. Sharqawy, J.H. Lienhard, S.M. Zubair: Ther- 32.34 R.A. Grace: Marine Outfall Systems. Planning, De-
mophysical properties of seawater: A review of sign and Construction (Prentice-Hall, Upper Saddle
existing correlations and data, Desalin. Water Treat. River 1978)
16(1-3), 354380 (2010) 32.35 R.A. Grace: Marine Outfall Construction: Back-
32.26 J. Pitt, P. Ackers: Hydraulic Roughness of Segmen- ground, Techniques, and Case Studies (ASCE, Reston
tally Lined Tunnels (CIRIA, London 1982) 2009)
32.27 D.S. Miller: Internal Flow Systems (Miller Innova- 32.36 W.J. Emery, R.E. Thomson: Data Analysis Methods
tions, Bedford 2009) in Physical Oceanography (Pergamon, Oxford 1998)
32.28 I. Idelchik: Handbook of Hydraulic Resistance 32.37 A. Underwood: Beyond BACI: Experimental designs
(Jaico, Mumbai 2003) for detecting human environmental impacts on
32.29 D.L. Wilkinson: Avoidance of seawater intrusion temporal variations in natural populations, Mar.
into ports of ocean outfalls, J. Hydraul. Eng. 114(2), Freshw. Res. 42(4), 569587 (1991)
218228 (1988) 32.38 32nocite]Please cite this reference in the text. Oth-
32.30 W.H. Hager: Wastewater Hydraulics: Theory and erwise it must be deleted.M.S. Field, R.G. Wilhelm,
Practice (Springer, Berlin Heidelberg 2010) J.F. Quinlan, T.J. Aley: An assessment of the po-
32.31 H.T. Falvey: Air-Water Flow in Hydraulic Structures, tential adverse properties of fluorescent tracer dyes
Technical Report (Water and Power Resources Ser- used for groundwater tracing, Environ. Monit. As-
vice, Denver 1980) sess. 38(1), 7596 (1995)
32.32 C.S. Lauchlan, R.W.P. May, R. Burrows, C. Gahan: Air 32.39 C. Leibundgut, S. Hadi: A contribution to toxicity of
in Pipelines: A Literature Review (HR Ltd., Walling- fluorescent tracers. In: Tracer Hydrology, Vol. 97, ed.
ford 2005) by A. Kranjc (Balkema, Rotterdam 1997)
http://www.springer.com/978-3-319-16648-3