Hoofstuk 6: wave induced structural response of coastal structures

Conceptual design vs. detailed design

Conceptual design of structural parts is first phase where we make use of simple
formulas, tables, guidelines,…. This design is based on the chosen design level:

- For failure mode damage, for example the determination of armour layer
block weight.
- For hydraulic responses, for example the crest height based on run-up or
overtopping criteria.

For this conceptual design we don’t make use of a probabilistic approach,

because it is to advanced. We will make use of a stochastic approach.

The different structural responses are described by a semi-empirical formula

in the coastal engineering. That means that the applied formulas are a
combination of two parts:

- Parameters based on physical consideration and insight

- Empirical constants, which are found by fitting on experimental data
obtained by doing testing.

By using a semi-empirical formula, we have a rather simple and not too complex
formula. This would not be the case when we try to include all the physics, which
is then not longer a semi-empirical formula.

After performing a conceptual design, then we can take that design to a lab and
start testing to have an idea of the real response (ex. looking to the occurring
wave overtopping). Here, a final optimalisation can be done by making use of
physical scale modelling testing.

Structural response types

Here, 4 different structural response types are considered:

- Wave run-up, defines the design crest level in case where no overtopping
is accepted (ex. breakwater with pedestrians).
- Wave overtopping, defines the design crest level and design of the upper
part of the structure as function of the allowable overtopping discharges.
- Wave transmission, see chapter 2
- Wave reflection, defines wave disturbance in harbour basins which
affects the manoeuvrability at the harbour entrances (see chapter 2).
Wave run-up and wave overtopping are described in this chapter, and both are
based on the EurOtop manual.

The EurOtop manual

The EurOtop manual is the Europese Overtopping manual which is written as a
common manual for all projects in Europa. This manual is based on manuals which
already existed before. These already existing manuals was one of Holland,
Germany and United Kingdom.

Wave run-up
Wave run-up and run-down on structures
Wave run-up Ru = maximum elevation of water-surface measured vertically from
SWL of the breaking wave run-up on the slope.

Wave run-down Rd = minimum elevation of water-surface measured vertically

from SWL of the breaking wave run-up on the slope.

This wave run-up and down can happen on both impermeable and permeable slope:

- Due to a wave that is moving up and down on the slope results in a minimum
and maximum vertical distance.
- The wave run-up on an impermeable slope will be lower than on an
permeable slope, because part of the energy is already dissipated by
water flowing through the porous of the skeleton.
- The combination of the run-up Ru and run-down Rd will indicate the active
zone where we can expect wave energy. So, in the not active zone, it is
possible to make use of smaller stones.
The values of Ru and Rd depend on the 2 families of parameters:

- Wave characterises:
o Incident wave height Hi
o Wave steepness si (= H/L), which will define a type of wave
breaking by the so-called breaker parameter ξ 0 = f(si)
- Geometrical characteristics:
o Slope angle α
o Surface roughness (natural or artificial)
o Permeability and porosity of the slope

In general, a smoother impermeable slope results in a larger R u .

Wave run-up on rock slope

In practice, we will allow some wave overtopping, because when we would avoid
wave overtopping, then the structure becomes not cost-effective. So, by
accepting some wave overtopping we obtain a economical acceptable solution.

In lab’s, the wave overtopping, wave run-up and run-down can be measured for
different situations and applications.

Vb. dia 8

Wave run-up on dikes and seawalls

Definition
The goal is to predict the wave run-up heights R u on impermeable structures. To
start, we have stochastic characterises of the waves, but also stochastic
information of the wave run-up. Reason is that each individual wave height gives
another wave run-up height, but which one needs to be used for the design.

The choice for the design parameter is the parameter R u2% which is the wave
run-up height exceeded by 2% of the incident waves at the toe of the structure.
With this parameter Ru2% known, it is possible to define the level of the
structure crest. Often this parameter is given in dimensionless form by dividing
it by the wave height Hm0 :

 Dimensionless form = Ru2% / Hm0

 Nowadays, the design is based on the allowable overtopping, but this R u2% is still
valuable for different reasons:

- Basic input for number of overtopping waves

- Upper or lower degree of armour protection, or structure elements (toe,
superstructure,…)
- Indicator for overtopping or transmission

Semi-empirical formula
The formula of wave run-up Ru2% / Hm0 with the mean value approach is given as
followed, and presented as the solid line in the figure:

This is a semi-empirical formula which means that we combine empirical defined

factors with parameters that describe the physics:

- Factors 1.00 ; 4.0 ; 1.5 are the semi-empirical factors defined based on
experimental data of testing and best-fitting techniques. These factors
will results in a better working formula.
- Reduction parameters to take different effects into account:
o γb = influence factor of the berm [-]
o γf = influence factor of the roughness of the slope [-]
o γβ = influence factor for oblique wave attack
- parameters which describes the physics
o ξm-1.0 = breaker parameter [-]
o Ru2%/Hm0 = relative wave run-up height [-]

The solid line shows this formula graphically, where we can see 2 different
regions. The first region is linear with a significant increase of ξ m-1.0 when
Ru2% / Hm0 is increasing. The second region is more gradual development with a
parabolic fit. So, in general, we can say that when the parameter ξ m-1.0 is
increasing (so going from spill to surge waves), then the wave-runup is higher.

Is the 5% exceedance line

The solid line represents the

average value of the 2%
measured wave run-up heights

= mean value approach

Multiple data sets are used for

different boundary conditions
to find empirical factors that
results in a general formula
The wave run-up heights are defined as the vertical distance between the
highest point of wave run-up and SWL. Due to the stochastic nature of the
incoming waves, each wave will result in a different run-up level. In the
Netherlands and Germany, many dike heights have been designed to a wave run-
up height Ru2% . This is the wave run-up height which is exceeded by 2% of the
number of incoming waves at the toe of the structure.

The idea behind this was that if only 2% of the waves reach the crest of a dike
or embankment during design conditions, the crest and inner slope do not need
specific protection actions other than clay with grass. For that reason, much
research in the past has been focused on the 2%-wave run-up height.

The choice for 2% as a design basis for crest level of dikes was long ago made.
The origin stems from the closing of the Sea in the Netherlands by the
construction of a 32 km long dike (Afsluitdijk in 1932). The M101 report gives
only the 2% wave run-up value and this must have been the time that this value
would be the right one to design the crest height of dikes.

Further tests (1939-1941) on wave run-up used only the 1% wave run-up value.
Other and later tests report the 2%-value, but for completeness also give 1%,
10%, 20% and 50% values.

It can be concluded that the choice of 2% was made in 1936, but the reason why
is not clear (design report is missing). Important to remark, this 2% goes about
the incoming waves, and not the number that runs up the slope.

In the past decade, the design and safety perspective has been changed to
allowable overtopping instead of wave run-up. Still a good prediction of wave run-
up is required as it is the basic input for calculation of number of overtopping
waves over coastal structure, which is needed to calculate overtopping volumes,
velocities and flow thicknesses.
The general formula that can be applied for the 2% wave run-up height was given
with the equations above. This describes the relationship between the relative
wave run-up height Ru2% / Hm0 and the breaker parameter ξm-1.0 .

The relative wave run-up height increases linearly with increasing ξ m-1.0 in the
range of breaking waves and small breaker parameters (ξ m-1.0 < 1.8). For non-
breaking waves and higher breaker parameters, the increase of R u2% / Hm0 is less
steep (parabolic development).

The relative wave run-up height Ru,2%/Hm0 is also influenced by the geometry of
the coastal dike or embankment seawall, the properties of the incoming waves,
and possibly by the effect of wind:

- Geometry of the coastal dike or embankment seawall are taken into

account by the parameters γb and γf .
- It is not recommended to consider the influence of wind on wave run-up
for coastal dikes or embankment seawalls.
- The properties of the incoming waves are considered in the breaker
parameter ξm-1,0 and the influence factor for oblique wave attack γβ .

Breaker parameter ξm-1.0

The relative wave run-up Ru2%/Hm0 is related to the breaker parameter ξm-1.0 :

 ;

As described in another chapter, the spectral average wave period Tm-1.0 is the
most suitable for the calculation of the wave run-up height for complex spectral
shapes as well as for theoretical wave spectra (ex. Jonswap). This spectral
period Tm-1.0 gives more weight to the longer wave periods in the spectrum and
therefore well suited for all kinds of wave spectra.
Peak period Tp (used in older studies), is difficult to use in some spectra (ex.
bimodal seas, flat spectra in shallow water,…), because they can result in large
inaccuracies. The relationship between peak period T p and spectral average wave
period Tm-1.0 is given as:

 Tp = 1.1 · Tm-1.0

This is a rule of thumb to find Tm-1.0 because mostly Tp is obtained.

Transition point ξtr

Like we said, the relative wave run-up height increases linearly with ξ m-1.0 in the
range of breaking waves. In other words, this happens for small breaker
parameter values ξ , which remain smaller than a certain transition value ξ tr .

The increase of the relative wave run-up is less steep (nearly horizontal) for
non-breaking waves. Or in other words, this happens for larger breaker
parameter values (ξ > ξtr).

On the graph, it can be seen that this transition point is situated around a
breaker parameter value ξ = 1.8

Influence of the slope

For a given slopes (ex. 1:6 slope), the relative wave run-up height decreases with
an increasing wave steepness.

For a given wave steepness (ex. s = 0.03), the relative wave run-up height
increases with an increasing slope angle (ex. 1:6 towards 1:3)
Reduction factors

Like we said, reduction factors γb , γf , γβ and γv are used to take influence

factors into account. Reason is that some physical process can result in a wave
run-up reduction, which can be easily taken into account by these reduction
factors (γ < 1). The values of each parameter is given in tables or with formulas.

The different reduction factors γ are:

- γb which takes the influence of the berm into account

- γf which takes the influence of the roughness of the slope into account
- γβ which takes the angle of direction and directional spreading in short
crested waves into account. In case of head-on waves, this factor γ β is
equal to 1 (β = 0°).
- γv which takes the influence of a wave wall into account. This parameter
will results in large reduction on the wave run-up (but also on the
overtopping volumes).

This formula was a semi-empirical formula obtained by using different data sets
which is covering different situations. So, when we are also using reduction
factors, we need to be sure that these situations are covered with the used
dataset’s. Reason is that we are not sure if we can extrapolate outside that
region. For that reason, also an applicability range is given, to make sure that the
used formula can be used and results in acceptable results. Here, the formula is
valid in the range:

 0.5 < 𝛾𝑏 𝜉𝑚−1,0 < 8-10

The influence factors are found by experimental investigations. But a

combination of the influence factors needs to be higher than 0.4 . A combination
of influence factors is often required in practice which reduces wave run-up and
wave overtopping significantly.
Uncertainty of the formula

Due to the large international dataset for all kind of structures a significant
scatter was observed present, which cannot be neglected for application. There
are several ways to include this uncertainty for application. This is done based
on the formula describing the mean and including now an uncertainty around this
mean approach.

In short, the mean value approach gives the average expectation and should also
give the reliability around this average. This is often done by giving a standard
deviation. The design approach includes some safety (one standard deviation σ)
and can straightforwardly be used for of coastal structures.

The previous formula can be seen as a probabilistic design (= mean value

approach) which made use of a normal distribution and:

- Parameter 1.65 with mean value of 1.65 and a standard deviation σ of 0.10
- Parameter 1.00 with mean value of 1.00 and a standard deviation σ of 0.07

For a deterministic design we add one standard deviation σ on the mean value
that was previously described (that value comes from international standards).
This results in value of 1.75 (= 1.65 + σ = 1.65 + 0.10) instead of 1.65, and a value
of 1.07 (= 1.00 + σ = 1.00 + 0.07) instead of 0.07 which should be used. This
results in the following equation:

So, with laboratory data we obtain the values for a probabilistic design. By
adding the +/- 5% exceedance lines we make use of µ + 1.64 · σ , where σ :

 σ = σ’µ = 0.07µ

(see summary)
Wave run-up on armoured permeable slopes
Permeable slopes results in a storage capacity of water, so wave run-up in the
case of permeable slopes will be smaller than or impermeable slopes. Hereby,
there is not a new formula developed, but the basic formulae is modified to take
the permeability into account.

This influence effect will be less important for rock slopes and rubble mound
breakwaters.
Wave overtopping
Wave overtopping over structures
For the wave overtopping over structures, different kind of structures are
treated separately:

- Smooth sloping structures, for example dikes and seawalls

- Rubble mound breakwater, for example breakwaters and rock slopes
- Vertical structures, for example caisson and sheet pile walls

In the new EurOtop manual, the case of a vertical breakwater is integrated in

the case of a sloping structure, which is more simpler now.

For wave overtopping, the crest freeboard Rc will be the most crucial parameter
which will also occur in the formulas. For wave run-up, that most important
parameter will be the breaker index ξm-1.0 .

Crest freeboard Rc = vertical distance between SWL and the crest of the
structure, and will be a crucial parameter for wave overtopping.

Wave overtopping happens if the crest level of the dike or embankment seawall
is lower than the highest wave run-up level Rmax . Wave overtopping depends on
the freeboard Rc and increases for decreasing freeboard height R c .

Overtopping characteristics
Usually the wave overtopping for dikes or embankments is described by an
average wave overtopping discharge q, which is given in m³/s per m width, or in
L/s per m width.

The overtopping discharge from wind-generated waves is very unevenly

distributed in time and space, because the amount varies considerably from wave
to wave. The major part of the overtopping discharge during a storm is due to a
small fraction of the waves. In fact the local overtopping discharge (in m³ /s
per meter structure) from a single wave can be more than 100 times the average
overtopping discharge during the storm peak.
Nevertheless, most information on overtopping is given as the time averaged
overtopping discharge q [m³/s per meter structure]. However, some limited
information exists on the probability distribution of the volume of overtopping
water per wave.

The average overtopping discharge q provides no information about the

discharge intensity of the individual overtopping waves (in other words, how
many waves will overtop and how much water will be overtopped in each wave)
However, such information is important because most damaging impacts on
persons, vehicles, and structures are caused by overtopping of large single
waves. For that reason, the overtopping volume per wave V will be used as design
parameter. Generally, most of the overtopping waves are fairly small, but a small
number gives significantly larger overtopping volumes.

The maximum volume overtopped Vmax in a sea depends on the mean discharge q,
on the storm duration and the percentage of overtopping waves. A longer storm
duration gives more overtopping waves, but statistically, also a larger maximum
volume. Many small overtopping waves (like for river dikes or embankments) may
create the same mean overtopping discharge as a few large waves for rough sea
conditions. The maximum overtopping wave volume will, however, be much larger
for rough sea conditions with large waves.

In the left figure it can be seen that only a few waves results in a high
overtopping volume. Therefore, we are not making use of an average value, but
the value Vmax .

On the right figure, we see the effect of “flow thickness” , this is the thickness
of the flow over the crest. This go very fast to a peak value, and then a
gradually decrease. Further it can be seen that not every single waves gives
wave overtopping.
 Average overtopping discharge q
Overtopping occurs when the run-up Ru is higher than the crest freeboard Rc .
The individual overtopping volumes Vi are unevenly disturbed in time and space,
resulting in variability between individual waves. By time-averaging this, we can
find the average overtopping discharge q:
mean volume of water during a certain time of period 3
 q= [m / s per m∨L/s per m]
time duration

Major part of the discharge during a storm is a consequence of a small fraction

of the highest waves. Therefore, locally it can be said that:

 qmax = 100q

For the individual overtopping volumes Vi the distribution is available.

We have an overconservative design, We have an overconservative design,

because we are looking to an unit Lot of spray, that is called green
length. While we use this design value water, which is not easy to catch
for the full length of the cross- during experiments to define a
section. In other words, we are volume of water.
finding a value for short-crested and Between experiments and reality,
apply it on a long-crested. there is also the influence of scaling
factors. For example density of
water and gravity remains the same
between design and experiments,
while other parameters change.

Spray = water that is overtopping the structure, but due to waves that are
breaking at the seaward (slope) of the structure.

Green water = water that is overtopping the structure, due to waves that are
running up the structure.

Tolerable average overtopping discharge

The tolerable average overtipping discharge q is given in tables. Here, the
designer needs to look for an acceptable conditions for safety of pedestrians,
traffic, vehicles and structures.

Vb. dia 23

The maximum, average tolerable overtopping discharge q for the safety of

traffic is equal to 1 L/s per m. Mostly a value of 10 L/s per m are applied.
Wave overtopping discharge (EurOtop)
Wave overtopping is the average discharge q per linear meter of width, for
example in m³/s per m or in L/s per m.

In reality, there is no constant discharge over the crest of a structure during

overtopping. The process of wave overtopping is very random in time, space and
volume. The highest waves will push a large volume of water over the crest in a
short period of time, whereas lower waves may not produce any overtopping.

An example of wave overtopping measurements is shown in the figure. The left

graph (flow depths) shows the irregularity of wave overtopping, where in this
case most waves overtop the crest. The right graph gives the cumulative
overtopping as it was measured in the overtopping tank by a load cell.

The graph shows some irregularities due to the dynamic behaviour of

overtopping wave volumes that fall into the overtopping tank. That means that
individual overtopping volumes cannot easily be distinguished in this case, as
some overtopping waves come in one wave group. The graphs show that at least 9
waves gave overtopping and the total overtopping volume was about 15L.

In order to calculate the average wave overtopping discharge, one should take
into account the duration of the measurements and the width of the chute that
catch the overtopping water:
mean volume of water during a certain time of period
 q=
time duration

Wave overtopping volumes (EurOtop)

A mean overtopping discharge does not yet describe how many waves will
overtop and how much water will be overtopped by each wave. The overtopping
wave volume V, coming over the crest, is given in m³ per wave per m width.
Generally, most of the overtopping waves are fairly small, but a small number
gives significantly larger overtopping volumes.
Maximum volume overtopped in a sea state depends on the mean discharge q, on
storm duration and the percentage of overtopping waves. In this EurOtop, a
method is given to find the distribution of overtopping wave volumes for certain
wave condition and average overtopping discharge.

A longer storm duration gives more overtopping waves, but statistically, also a
larger maximum volume. Many small overtopping waves (like for river dikes or
embankments) can create the same mean overtopping discharge as a few large
waves for rough sea conditions. The maximum overtopping wave volume will,
however, be much larger for rough sea conditions with large waves.

Applications
Vb. dia 24 – 28

Geometry
So, our goal is to predict the required crest freeboard based on the following 2
parameters, found depending on the situation:

- Overtopping volume per wave Vi [m³/m]

- Average overtopping discharge q [m³/s/m]

Following figure gives an indication of the different geometrical parameters

that are used during the calculation of the wave overtopping.

Wave characteristics
(see summary)
Overtopping parameters
The question is which parameters are significant for the wave overtopping. We
know that the average overtopping discharge q is function of:

- Waves, described with a wave height H and period T

- SWL, described with a water depth d
- Geometry, described with α, Rc and a berm (with B and dh)
- Type of structure, described with a certain roughness (smooth or rough)
and a permeability

Besides that, also the fact that the waves are breaking or non-breaking need to
be taken into account. This is done by using the breaker parameter ξ 0 .

(see summary)

Type of mathematical formulation

The general formulation is given by using the following exponential function:

In this formula:

- Q = dimensionless average overtopping discharge per meter crest

- R = dimensionless freeboard (= freeboard / incident wave height)
- a, b = empirically obtained coefficients

This give an exponential relationship between the dimensionless average

overtopping discharge Q and the dimensionless freeboard R. This is again a semi-
empirical formula, because 2 empirical coefficients (a,b) are used. In this
formula, also a “-“ is used, because if the crest height increase, than the wave
overtopping should be smaller.

Figure shows the graphical illustration of that formula, with a logarithmic,

vertical axis. For that reason, the value of a can be found as the intersection
point of the curve with that vertical axis. Value of B is the slope of the curve.
 This general formula is further worked in 2 types: Van der Meer or Owen.
Value of a

Logarithmic scale
Slope = value of b

Van der Meer formula

(see summary)

The 5% exceedance and 5% non- Lines look to go to the same point

exceedance values are giving the if Rc goes to a value of 0.
band where 90% of the data will
For design, we make again use of
be present. On this graph, it can
deterministic calculations (due to
be seen that if the overtopping
safety reasons), which means that
discharge is decreasing, the
we increase the probabilistic (=
uncertainty increase. The 5% is
mean) with 1 deviation σ.
corresponding with 1.64 σ

Probabilistic approach is the best

exponential fit curve of the
Logarithmic scale
formula on the experimental data.
In that way, a and b can be found.
Hereby, 50% of the data is above
and 50% below the line (= mean).
Recommended for deterministic design

It is recommended to make use of the deterministic design. This is found, based

on the mean for probabilistic design approach:

- Coefficient b is normal distributed with a µ and σ:

o Breaking: µ = 4.75 ; σ = 0.5
o Non-breaking: µ = 2.60 ; σ = 0.35

The recommended line to use during the design (= deterministic), is a

conservative line with coefficient b founded with “µ - σ”:

- Breaking: µ = 4.30 (= 4.75 – 0.5)

- Non-breaking: µ = 2.30 (= 2.60 – 0.35)
A

The reason why a deterministic design is required, is due to the lot of scatter
was observed with the used experimental data sets to define a and b. In point A,
we have small amount of scatter, resulting in less uncertainty and as a
consequence both lines are coming together.

In point B, there is much more scatter, and therefore a large uncertainty. For
that reason, both lines (deterministic and probability) are quit far away from
each other compared to small values of R (or high values of overtopping Q).
There is more scatter obtained in point B, because ore overtopping needs to
happen due to the higher relative freeboard.
Reduction factors

The different reduction factors γ are:

- γb which takes the influence of the berm into account

- γf which takes the influence of the roughness of the slope into account
- γβ which takes the angle of direction and directional spreading in short
crested waves into account. In case of head-on waves, this factor γ β is
equal to 1 (β = 0°).
- γv which takes the influence of a vertical wall on a slope into account. This
parameter will results in large reduction on the wave run-up (but also on
the overtopping volumes).

In general, if the reduction factor has no influence, then γ = 1. But if the factor
has an influence, than it is always a value smaller than 1. In that way, a reduction
of the average overtopping discharge is obtained by increasing R c/Hm0 by dividing
by γ.

Reduction factor γf

The reduction for the slope surface roughness γf is described in tables based on
the used material. For some general materials (ex. asphalt and concrete) a value
of 1 is founded, because no reduction is required. Due to the fact that the
formula is derived for these cases.

Further it can be seen that by applying extra roughness, the value of γ f becomes
smaller than 1 resulting in a cheaper design (because the crest height can be
reduced for the same overtopping in case we increase the roughness).

Vb. dia 39

Reduction factor γb

With the reduction factor γb we take the berm into account. This is done with
the following formula:

Hereby, we can say that:

- rB = B/Lberm , berm width (is 0 if no berm is installed)

- rdh = berm level (is 0 if the berm lies on SWL)
 So, in general, the following is valid:

- If rdh = 0, then the berm is applied at SWL, which means that γ b becomes
smaller than 1. Reason is that this is not the standard case for which the
formula is derived.
- If rdh  0, then parameter rb is multiplied by a number smaller than 0,
therefore the reduction factor γb becomes larger again.

Limited to a value of 1,
The berm is most effective if
because a reduction factor
the crest height is equal to the
can’t be larger than 1.
free water surface. Because
So berm lying below 2Hm0 or then rdh = 0, resulting in the
above Ru2% has no influence lowest value of γb .
on wave run-up and wave
Effective = small γb
overtopping.

“+” means below free surface,

“-“ means above free surface

Owen formula
Is another option than the Van der Meer equation. The main differences are:

- Other representative wave characteristics are used (here H 1/3 and Tm

while Van der Meer makes use of Hm0 and ξ0)
- Mathematically easier to apply
- No probabilistic approach possible
- Regional usage (Anglo-Saxon)

The parameters a0 and b0 are empirically obtained values (with experimental

data and statistical fitting). These parameters are a function of “tan(α)”, with α
the slope of the structure, but also takes the influence of the berm into
account. Such a berm is described with a geometry B and d h .

(see summary)
Exercise
During the design, 2 options are possible:

- A tolerable volume of q is given at we search a value of R c = design

- A freeboard value of Rc is given (for an existing situation) and we check it
= verification of the design/situation

Vb. dia 42 - 44

Wave overtopping at vertical walls

The wave overtopping at vertical walls is described in the EurOtop manual
chapter 7. But there is an update of the EurOtop manual based on the paper van
van der Meer and Bruce (2014). Therefore, we kicked out the formulas for
vertical and steep seawalls and integrate them in the formula of van der Meer,
where we just change the empirical factors.

When looking to wave overtopping at vertical walls, then we have 2 options:

- Non-impulsive (top figure), here wave-overtopping is non-impulsive

because wave is running over the vertical wall and not breaking before the
vertical construction. So, this is the non-breaking case.

- Impulsive (bottom figure), here wave-overtopping is impulsive because we

have wave breaking before the vertical wall. So, this is the breaking case,
which needs to be avoided in practical projects.
EurOtop manual
Before the EurOtop manual we had 3 manuals, one of Holland, Germany and one
of the United Kingdom. But research and updates where needed/performed due
the constructions of breakwaters was applied in deep water, and not longer in
shallow water. There the waves are behaving differently and was not covered in
these 3 manuals. For that reason, Europa developed a general manual the
EurOtop manual. The first one was published in 2007, where an updated version
in 2016.

Overtopping manual (2016)

The following supporting material is available for the user:

- Web-based programmes for the calculation of overtopping discharge and

design details.
- Annex with photo and video visualisations of overtopping processes.
- Case studies and example calculations

The available engineering tools are:

- Empirical prediction formula and PC overtopping software

- Experimental and numerical modelling
- EurOtop Artificial Neural Network (ANN) and EurOtop database

Modifications compared to version of 2007

Due to the application of construction in deep water, a lot of research was done,
because the results where deviating from the expected values obtained for
shallow water conditions. Resulted in the modification of the EurOtop manual.

How do we describe overtopping

Wave overtopping is described with the following expressions:

 Run-up Ru2% > freeboard Rc

To make a design possible, we will make use of an averaged value, the so-called
mean overtopping discharge q (m³/s per m crest width). Hereby, we have seen
previously, the difference between the “mean overtopping discharge q” and the
“individual overtopping volume per wave Vi”.
The average overtopping discharge q, has been the main parameter to describe
wave overtopping. It is simply the measure the amount of water that comes over
the crest, divided by the time that the amount of water was measured. For
example, if 5 m³ water has been measured per m width and over 1 hour, the
average overtopping discharge becomes q = 1.39 L/s per m.

But the actual behaviour at the structure is completely different. The figure
gives an idea of the irregular behaviour of wave overtopping in time. The red line
shows the random waves that have been generated in a physical model. Only
some of the largest waves will reach the crest and generate wave overtopping.
The black line gives the height measurement of a wave gauge placed on top of
the crest and gives the flow thickness of overtopping wave volumes.

The graph shows that overtopping wave volumes are irregular, all different and
that there are less overtopping wave volumes than incident waves. Only a certain
percentage of the incident waves reach the crest and cause wave overtopping.

So, the waves have a stochastic behaviour. Hereby, it’s possible to measure the
distribution of the individual overtopping wave volumes V i to define a maximum
volume Vmax .

Tolerable wave overtopping

The tolerable wave overtopping has been tackled in a different way between
version 1 and 2 of the EuroOtop:

- In EurOtop 1, we make use of 4 tables with estimated tolerable wave

overtopping for pedestrians, vehicles, property behind the structure and
structural damage.
- In EurOtop 2, it was known that the tolerable overtopping depends
strongly on Vmax , and therefore on the incident wave height. Reason is
that for a given mean overtopping discharge q there are 2 options:
o Small waves resulting in a small Vmax
o Large waves resulting in a large Vmax
Therefore, we are coupling the tolerable overtopping discharge to
wave height, and assuming therefore that Vmax is a good indicator to
describe impact of overtopping.
Influence of wave height H on same discharge
The maximum overtopping volume Vmax can now be calculated for some structures
of simple geometry with reasonable accuracy. Of course it will not only be the
maximum volume that can cause damage, but all overtopping volumes that are
close to this maximum overtopping volume. Nevertheless, V max is a good
parameter to describe how heavy the wave overtopping is or can be.

The statistical distribution of overtopping wave volumes depends on structure

type, incident wave conditions (wave height and period), freeboard, duration of
wave overtopping and the mean overtopping discharge.

The graphs give the maximum overtopping wave volume V max, as well as the
number of overtopping waves. The graph shows that:

- Waves with a high wave height, only a small amount is overtopping the
crest, but with large volumes
- Waves with a smaller wave height are overtopping the structure with
much more individual waves but with smaller volume.

So, mainly the wave height has a large influence on the maximum overtopping
volume. It’s likely that most damage is caused by largest overtopping volumes, so
tolerable limits should be based on these volumes and not only on tolerable mean
discharges. A maximum tolerable overtopping volume, however, can be given by
different wave heights combined with different mean discharges.
Tolerable wave overtopping limits
The EurOtop gives information for the structural design of wave overtopping
for breakwaters, seawall, dikes and dams. Hereby, the design will be safe for
people, vehicles and property behind the defence.

In the EurOtop makes use of 3 wave height classes:

- Hmo ≤ 1m : rivers, wide canals and small lakes. Often embankments covered
with grass.
- Hm0 = 1-3m : sheltered seashores and large lakes. Embankments, seawalls
with the wave attack zone protected by rock, concrete units or block
revetments. Grass covered crest or protected promenades/boulevard.
- Hm0 ≥ 3-5m : high waves and large water depths (> 10 m) near the
structure. Breakwaters, seawalls at land reclamations.

In the EurOtop manual, a table is given which gives for each class and reason a
mean discharge q [L/s per m] and maximum volume V max [L/m].

Vb. dia 20

More insight in overtopping performance of dikes and vertical walls

In the EurOtop 1, the overtopping was based on the Owen formula. That is an
exponential function with 2 empirical coefficients a and b:

In the EurOtop 2, the overtopping is described with a modified formula based

on the UGent10 dataset which was focused on steep low-crested structures.

For sloping structures, overtopping at low and zero freeboard conditions has
often been overlooked in physical model studies but they represent important
situations. It is clear that familiar, exponential-type formulae work poorly in
these regions as they over-estimate the overtopping discharge.

Analysis has therefore been performed to bring together the classic

exponential formulae with the UGent10 dataset. This resulted in a Weibull-type
function with now 3 coefficients a, b and c (= 1.3):


 So, by using a additional experimental coefficient c with a fixed value of 1.3 we
have a new fitting of the formula on experimental datasets. This makes it
possible to use the formula for different types of structures with acceptable
accuracy.

Graph: the value of C = 1.3 gives a slightly curved line on a log-linear graph,
where the classic exponential distribution gives a straight line. The differences
between the formulas are small, but the largest for the area R c/Hm0 < 0.5 (= mild
slopes and low freeboards) where the EurOtop 1 formulae is not valid anymore.
The formula of EurOtop 1 is only valid in the range 0.5-1.8m for R c/Hm0 , when
going out of that range (ex. due to seawater level increase) it can’t be applied
anymore and the new formula of EurOtop2 can be used.

Logarithmic scale

Graph: illustrate the formula that give a prediction of the average discharge or
mean value approach (new name of the probabilistic design) with the formula of
EurOtop 2 (c = 1.3), and the value for a deterministic design. That means that
one time a standard deviation σ is added to have a safe design. Also the 5%
lower and upper confidence limits have been plotted (1.64σ).

Logarithmic scale
Graph: shows the difference for steep sloping structures with non-breaking
waves and a vertical wall. It is clear that if the slopes become very steep, up to
vertical, the overtopping discharge should decrease for similar wave conditions.
Question hereby is the difference between both lines.

Another difference between the formula of EurOtop 1 and 2 is that the

parameters a and b are not longer described with algebraic values, but with
formulas which depends on the slope α of the structure. In that way different
types of slope cases can be designed with 1 formula. This was not possible with
the old formula of EurOtop1.

This formula will be applicable for Rc ≥ 0, γf = 1 and cot(α) ≤ 2.

(see summary)
Remark
We would think that due to the exponent c = 1.3, the effect of the influence
factors would change, as the part within the exponential function becomes
different for c = 1 (= the old formula) and c = 1.3 (= the new formula). This is
indeed true for the same (large) value of Rc/Hm0, but the effect of an influence
factor is that the curve shifts to the left and remains in the same overtopping
range. The actual difference between the old and new formulae is similar with
and without an influence factor different from 1.0 .