Hydraulic Behaviour of Submerged Breakwaters: a Case Study

Christos Makris
Department of Civil Engineering, Aristotle University of Thessaloniki, Thessaloniki, Hellas
Ioannis Avgeris
Department of Civil Engineering, Aristotle University of Thessaloniki, Thessaloniki, Hellas
Constantine Memos
School of Civil Engineering, National Technical University of Athens, Athens, Hellas

Abstract
New forms of coastal structures are being investigated nowadays, catering also for the
aesthetic value of the nearshore landscape. Among those structures, the submerged
breakwater is becoming attractive for obvious reasons. The wave transmission coefficient
associated with the latter has been studied extensively in the past. However, an aspect not
thoroughly investigated so far is the effect of the structure porosity on the above coefficient.
In this paper a review of the transmission coefficient over submerged breakwaters is given,
based on application of empirical formulas and numerical models to a case study. Apart from
the porosity, a second parameter was investigated, namely the friction along the breakwater
perimeter. It was found that porosity can have a significant effect on the transmission
coefficient, and that it can be adequately described by one of the wave models tested and by
an empirical formula. The bed friction was found to have a smaller effect on the wave
transmission coefficient than permeability has.

Introduction
The design of non-conventional types of coastal protection structures is increasingly
becoming a field, where environmental issues can put stringent criteria regarding the layout of
the structure, the materials used, etc. A major environmental consideration refers to the
restricted view to the horizon associated with conventional surface piercing breakwaters. The
submerged breakwater is, therefore, widely investigated, offering a major aesthetic advantage,
since no part of the structure is visible from the shore. A key factor measuring the
effectiveness of such a structure is the transmission coefficient Kt, i.e. the ratio of the
transmitted to the incident wave height. In a recent paper (Makris and Memos, 2007 denoted
in the following by MM) it was shown that the wave transmission is often deduced
satisfactorily by semi-empirical formulas or by models based on a parabolic approximation to
the mild-slope equation. The principal factor controlling the transmission coefficient is
associated with the description of the wave breaking at the breakwater. Various breaking
formulations have been examined in this context and the main ones are tested in the
following. Of the geometric characteristics the two most important in shaping Kt are the
freeboard F and the crest width.
Following in significance appears to be the flow allowed through the pores of the submerged
structure, usually made of rubble. The role of the porosity on Kt is investigated in this paper
and its significance assessed through applications of wave models and formulas to a real-life
problem. Also, the effect that the friction along the perimeter of the structure has on the wave
transmission is studied herebelow.

Wave Models Used

Three wave propagation models were used for the estimation of the transmission coefficient.
Two of those, PMS and BW1D modules of MIKE 21 (DHI, 2005), were briefly presented in
the companion paper MM. The third model, denoted by BWA, is a Boussinesq wave model
for porous submerged structures (Avgeris et al., 2004). This is a higher-order model, with
improved linear dispersion characteristics incorporating extra terms that account for the
interaction between the waves and the flow within the porous structure. The governing
equations are coupled in the region of the structure with a depth-averaged Darcy-Forchheimer
(momentum) equation that describes the porous flow.
Wave Breaking Formulations
Energy dissipation due to wave breaking is the dominant factor for correctly tuning wave
propagation models in shallow waters. For PMS module four wave breaking formulations
were checked, namely the basic formulation due to Battjes and Janssen (BJ, 1978), the
modification due to Battjes and Stive (BS, 1985), the one due to Nelson (1987), and that by
Johnson (2006). The first three formulations have been developed for wave breaking on a
beach, while the latter for wave breaking over submerged structures with steep slopes. The
expressions of the above modules, which can be found in MM, were applied in this study
externally to the PMS wave model.
In MIKE 21 BW module the surface roller concept has been used, as presented in MM. In
BWA model the eddy viscosity method (Kennedy et al., 2000) is used to model wave
breaking. The breaking dissipation terms, added in the momentum equation, depend on the
eddy viscosity coefficient, which is a function of both time and space. The empirical value of
the parameter t(I) that controls the initiation of the breaking event is set to 0.35(gh)1/2, h water
depth, as proposed by Kirby et al. (1998) in the case of submerged breakwaters.
Energy Dissipation due to Bottom Friction
In PMS module the rate of energy dissipation due to bottom friction is formulated by
introducing a dissipative term in the governing momentum equation. In BW module the
quadratic friction law is used to express bottom shear stress. Details on both formulations are
given in MM. The friction coefficient fw along the breakwater perimeter was calculated
through the expressions of Madsen and White (1975) and Van Gent (1995).

Empirical relations for wave transmission

In the present investigation four expressions of those presented in MM were applied to a case
study and the consistency of their results was checked with reference mainly to the porosity of
the structure. These formulas are the one by VdMeer and dAngremond (1991) referred to in
CEM (2004), that by DAngremond et al. (1996), by Seabrook and Hall (SH, 1998), and by
Friebel and Harris (2003).
Submerged Breakwater Stability
Some of the applied empirical relations for wave transmission require as input the nominal
armour rock diameter Dn50. This was estimated through the following two procedures:
VdMeer and Pilarczyks (1991) expression
Use is made of the following relation applicable to statically stable submerged breakwaters:
1
F
h
(
= (2.1 + 0.1S )exp 0.14 N s* , valid for slopes 1/1.5~1/2.5 ) (1)

where, S a damage index (S=0 no damage, S=8 complete failure), N s* = H i / Dn50 S 1p/ 3 ,
=(a/w)-1, Sp=Hi/Lp, Hi the significant incident wave height, Lp the local wave length at the
spectral peak, h the water depth at the structure toe.
Rule of thumb (RoT) selection of Dn50
Following Burcharth et al. (2006) a quick estimate of Dn50 can be obtained through the
expression Dn500.29(h-F).

Application to a case study

The study area
The project under study is developed around a water expanse comprising a man-made lagoon
and occupying an area of about 6.2 hectares on the shores of north Red Sea. The lagoon
waters will be used mainly for swimming and related activities. Figure 1 shows the general
layout, containing two submerged breakwaters, the principal role of which is the protection
from wave attack of the bungalows to be built on piles at the shore.

y
-11

-4

-2

500
-11

450
-4

-1

400
-11

-4

-1

350 Palette
-11

-1
-4

-1
(Units in meter)

.5

Above 0
300 -2 -0.5 - 0
-11

.5
-4

-1 - -0.5
-1.5 - -1
250 -2 - -1.5
-11

-2.5

-1

-2
-4

-1
0

-2.5 - -2
.5

.5

200 -3 - -2.5
-2.5
-0.5

-4 - -3
-11

-4

-2

-5 - -4
150 -6 - -5
-0.5
-11

-4

-7 - -6
0
-2

100 -9 - -7
-11 - -9
-2.5
-11

-4

-13 - -11
50 -15 - -13
Below -15
-2.5
-11

-4

0 Undefined
0 100 200 300 400 500 600 700
(Units in meter)

Figure 1. Lagoon reference plan

Input conditions
The main input conditions are given in MM. The value of Dn50 for the breakwaters was
calculated in the range 0.52m to 0.56m for the various hydrografic conditions tested, through
VdMeer and Pilarczyks (1991) expression. The rule of thumb presented previously gave for
the same as above conditions Dn50=0.51m to 0.55m. These values were used as input to the
empirical expressions giving the wave transmission coefficient.
Additional input to the MIKE (PMS & BW1DH) models were as follows.
Bed friction along the breakwater skin: fw=0.15 following Van Gents (1995) proposal.
For the sea bed a value of the Nikuradse roughness parameter kN=0.3mm was used.
Wave breaking formulation due to Johnson (2006) was applied with 2=1.262 for the
10-yr and with 2=1.355 for the 50-yr conditions.
Input to BWA model was as follows: Dn50=0.52m, fw=0.15 at structure, fw=0.006 at sea
bottom, porosity =0.5, =1100, =1.2. The selected value of is representative of single
layer submerged structures in physical studies and field projects. The values of the porous
resistance coefficients and are in the range proposed by Van Gent (1995) and have been
previously used with success in similar tested cases.

Results and Discussion

Wave Transmission: Formulas
The wave transmission coefficient Kt calculated by the four empirical expressions presented
previously are given in Table 1, for four incident wave conditions. Sea level is at mean
position except where LAT (lowest astronomical tide) is noted.
TABLE 1
WAVE TRANSMISSION COEFFICIENT BY FORMULAS

Formula Hs 10-yr Hmax 10-yr Hs 50-yr Hmax 50-yr

CEM 0.695 0.631 0.677 0.652
D'Angr et al. 0.342 0.425 0.503 0.474
10yr, 50yr 0.607 0.508 0.544 0.517
Sbrk+Hall
LAT 0.587 0.494 0.529 0.505
Sbrk+Hall 10yr, 50yr 0.551 0.482 0.540 0.518
Dn50 by RoT LAT 0.581 0.511 0.581 0.559
Friebel+Harris 0.472 0.417 0.526 0.508

The relation by SH involves Dn50 as a parameter and it gives results in the mid-range for all
four wave conditions tested. Thus this formulation is retained in the following as the most
suitable one for comparison with the model results.
Wave Transmission: Models
The numerical models used in this application produced Kt values that vary, for PMS models,
with respect to the wave breaking formulation employed. Results were taken at a typical
cross-section of the southern breakwater at the middle of its length. The relevant values of Kt
are presented in Table 2.
TABLE 2
WAVE TRANSMISSION COEFFICIENT BY MODELS

Model Hs 10-yr Hmax 10-yr Hs 50-yr Hmax 50-yr

Default (BJ) 0.550 0.393 0.598 0.466
BS85 0.539 0.399 0.595 0.472
PMS1DH
NEL87 0.443 0.392 0.484 0.467
JOHNS06 0.688 0.448 0.725 0.464
Default (BJ) 0.464 0.369 0.431 0.525
BS85 0.460 0.373 0.431 0.531
PMS2DH
NEL87 0.384 0.317 0.337 0.573
JOHNS06 0.549 0.396 0.484 0.548
BW MIKE21 0.367 0.690 0.480 0.678
BWA 0.580 0.530 0.620 0.580
Wave height profiles
In this subsection some profiles are given providing the significant wave height along the
considered typical breakwater cross-section. In Figure 2 results are shown for the 50-yr
Jonswap waves using the PMS1D model associated with the wave breaking formulations
cited above. It can be seen that the transmitted wave and hence Kt, associated with Johnsons
breaking criterion is notably higher than the transmitted wave produced by the rest breaking
formulations.
Default Hmo [m] WaterDepth [m]
JOHNSON06 Hmo [m]
NELSON87 Hmo [m]
BS85 Hmo [m]

Profile1DResJonswapRanWav50yr
5.0 60

55
4.5

50
4.0
45
3.5
40

3.0
35
Significant Wave Height Hmo (m)

2.5 30

Elevation (m)
25
2.0

20
1.5
15
1.0
10

0.5
5

0.0 0

-5
-0.5

-10
-1.0
-15
-1.5
0 50 100 150 200 250 300
Horizontal Distance (m)

Figure 2. Significant wave height along a cross section, by PMS1D

A similar graph of the significant wave height as produced by MIKE 21 BW model can be
seen in Figure 3. The same as previously wave conditions apply.
Finally, a graph is given in Figure 4, where results by the model BWA are presented. Here the
free surface elevation is reproduced along the same cross-section for the final time step, where
stable conditions have been achieved.
Effect of permeability- Comparison between models & formulas
A comparison between Kt results obtained by the models and those by applying SH formula is
presented in Figure 5. As mentioned earlier this formula is assumed in the present context as
the most suitable to compare model results to. In this figure two lines have been drawn
denoting a band of acceptable deviation of 5% from the true values of Kt provided by the
formula.
Wave height H [m] Bathymetry [m]

Significant Wave Height Hmo BWJonswap50yr

5.0 60

55
4.5

50
4.0
45
3.5
40

3.0
35
Significant Wave Height Hmo (m)

2.5 30

Elevation (m)
25
2.0

20
1.5
15
1.0
10

0.5
5

0.0 0

-5
-0.5

-10
-1.0
-15
-1.5
0 100 200 300 400 500
Horizontal Distance (m)

Figure 3. Significant wave height along a cross section, by BW

-1

-2

-3
H=2.72 m
-4 T=7.14 sec
d (m)

-5

-6

-7

-8

-9

-10

-11

-12

350 375 400 425 450 475 500 525 550

x (m)

Figure 4. Free surface elevation, by BWA

 0.75
BW1D
MIKE
0.70
BWA
Kt from 'Seabrook & Hall' FORMULA
0.65
PMS2D
J06
0.60
PMS1D
BJ
0.55
PMS1D
BS85
0.50
DIAG

0.45
-5%

0.40
+5%

0.35
0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75
Kt from MODELS

Figure 5. Comparison of models with SH results

It can be noted that BWA model behaves consistently and quite close to the values obtained
by the SH formula. It is reminded that the said model caters for the porosity of the submerged
breakwater. In contrast, MIKE BW1D model gives results all of which fall outside the 5%
band. The modules of PMS model perform somehow in the mid-range between the above two
extremes. Referring to PMS1D it can be said that since both wave breaking formulations
tested, i.e. BJ and BS modules, have been developed initially for wave breaking over mild
slopes, they should overestimate wave breaking at the relatively steep slopes of the structure.
At the same time PMS1D was found (MM) to overestimate Kt with respect to PMS2D, when
no wave breaking is assumed. Thus it seems plausible that the two opposite effects cancel
each other out and the net results of PMS1D fall in good agreement with SHs values. The
wave breaking module due to Johnson (2006) was found in MM that in general
underestimates the amount of breaking at submerged structures, except when associated with
the wave propagation model PMS2D, for which it had been actually calibrated.
Effect of bottom friction
Apart from the porosity effect on wave transmission a second parameter investigated in this
study was the friction along the outline of the cross-sectional area of the submerged
breakwater. In PMS module two sets of roughness value were tested as follows:
(a) kN=0.3mm (sea bed), kN=12.5mm (structure)
(b) kN=0.3mm (sea bed), fw=0.15 (structure)
The Nikuradse roughness parameter kN on the structure of case (a) corresponds roughly to
fw=0.02 through Svendsen and Jonssons (1980) formula.
If we denote by Kt1, Kt2 the transmission coefficient factor related to the values fw=0.15,
fw=0.02 respectively, then the ratio Kt2/Kt1 produced by PMS1D model is presented in Figure
6 for various wave conditions and wave breaking formulations.
As expected, the above-mentioned ratio is greater than one for all cases checked. The wave
conditions referred to along the horizontal axis of the previous graph can be decoded through
Table 3.
1.4

PMS1D
1.3 BJ

1.2 PMS1D
BS85
Kt2 /Kt1

1.1

PMS1D
1.0 N87

0.9 PMS1D
J06

0.8
0 1 2 3 4 5 6 7
Wave Cases

Figure 6. Change in Kt due to frictional variation, PMS1D model

TABLE 3
WAVE CONDITIONS USED IN COMPARISONS

Hs 10yr Hmax 10yr Hs 50yr Hmax 50yr

Kt2/Kt1
JONSWAP TMA MaxReg JONSWAP TMA MaxReg
Cases 1 2 3 4 5 6

Figure 7 depicts similar comparisons of results given by PMS2D model. It can be seen that
for most wave conditions tested the ratio Kt2/Kt1 does not fall below 1.0, which is quite
plausible. Finally Figure 8 gives the same ratio for the BW model for various wave conditions
as shown. Again the said ratio is always larger than one for all wave conditions checked but it
does not exceed 1.1 in contrast with the corresponding values around 1.2 and 1.3 for models
PMS1D and PMS2D respectively.
1.4

PMS2D
1.3 BJ

1.2 PMS2D
BS85
Kt2 /Kt1

1.1

PMS2D
1.0 N87

0.9 PMS2D
J06

0.8
0 1 2 3 4 5 6 7
Wave Cases

Figure 7. Change in Kt due to frictional variation, PMS2D model

 1.150
BW1D Jonswap
10yr
1.125
BW1D TMA 10yr

1.100
BW1D MaxReg
10yr
Kt2/Kt1

1.075
BW1D Jonswap
50yr
1.050
BW1D TMA 50yr

1.025
BW1D MaxReg
50yr
1.000
0 1 2 3 4 5 6 7
Wave Cases

Figure 8. Change in Kt due to frictional variation, BW1D model

In any case it can be noted that the rate of increase of the wave transmission is far less than
the corresponding decrease in fw, as it is easily verified through inspection of the three
previously mentioned graphs.

Conclusions
In this study the effects on the wave transmission coefficient of the structure porosity and
bottom friction along the skin of a submerged breakwater were investigated through
application to a case study. Some widely accepted empirical formulas and wave models
associated with wave breaking formulations were examined and compared. The following
results were obtained:
(a) The porosity of the breakwater has a significant effect on the value of the wave
transmission coefficient Kt.
(b) In this respect the empirical formula by Seabrook and Hall (1998) gives satisfactory
results and can be used with some confidence in predicting Kt in the presence of
porous structures.
(c) The wave model, among those tested, best suited in describing the process of wave
transmission through porous breakwaters is the one-dimensional Boussinesq model
developed by Avgeris et al. (2004).
(d) The wave model MIKE PMS behaves adequately, especially the two-dimensional
one equipped with Johnsons breaking formulation. The one-dimensional MIKE
BW model predicts rather poorly the wave transmission under the conditions tested.
(e) The bed friction along the outline of the breakwater cross-section is a less crucial
factor than porosity in shaping the wave transmission coefficient.

References
Avgeris, I, Karambas, ThV and Prinos, P, (2004), Boussinesq Modeling of Wave Interaction
with Porous Submerged Breakwaters, Proc. of the 29th Int. Conf. on Coastal Engineering,
ASCE, pp 604-616.
Battjes, JA, and Janssen, JPFM, (1978), Energy Loss and Set-up due to Breaking of
Random Waves, Proc. 16th Int. Conf. on Coastal Engineering, Hamburg, Germany,
pp 569-587.
Battjes, JA, and Stive, MJF, (1985), Calibration and Verification of a Dissipation
Model for Random Breaking Waves, J. Geophysical Research, Vol 90 (C5), pp
9159-9167.
Burcharth, HF, Kramer, M, Lamberti, A, and Zanuttigh B, (2006), Structural Stability of
Detached Low Crested Breakwaters, Coastal Engineering, 53, Elsevier, pp 381-394.
Coastal Engineering Manual (2004), CEM 2.01 Professional Edition, US Army
Engineer Research and Development Center, Veri-Tech, Incorporated, Vicksburg,
USA.
D'Angremond, K, Van der Meer, JW, and De Jong, RJ, (1996), Wave Transmission at
Low-crested Structures, Proc. 25th Int. Conf. on Coastal Engineering, Orlando, Florida, pp
2418-2426.
DHI (2005), MIKE21 User Guide and Reference Manual, Danish Hydraulic Institute,
Water and Environment, Denmark.
Friebel, HC, and Harris, LE, (2003), Re-evaluation of Wave Transmission Coefficient
Formulae from Submerged Breakwater PhysicalModels, Index paper, Internet version.
Johnson, HK, (2006), Wave Modelling in the Vicinity of Submerged Breakwaters,
Coastal Engineering, Vol 53, pp 39-48.
Kennedy, AB, Chen, Q, Kirby, JT and Dalrymple, RA, (2000), Boussinesq Modeling of
Wave Transformation, Breaking, and Runup. I: 1D., Journal of Waterway, Port, Coastal
and Ocean Engineering, Vol. 126, No. 1, pp 39-47.
Kirby, JT, Wei ,G, Chen, Q, Kennedy, A B, and Dalrymple, R A, (1998), FUNWAVE 1.0,
Fully nonlinear Boussinesq wave model, Documentation and users manual. Research
report no. CACR-98-06, University of Delaware, pp 1-80.
Madsen, OS, and White, SM, (1975), Reflection and transmission characteristics of
porous rubble mound breakwaters, Report No. 207, RM Parsons Lab, Dept of Civil
Eng, Massachusetts Institute of Technology, Cambridge, Mass.
Makris, CV and Memos, CD, (2007), Wave Transmission over Submerged
Breakwaters: Performance of Formulae and Models, Proc. 17th ISOPE Conference,
ISOPE, Lisbon, Portugal, pp 2613-2620.
Nelson, RC, (1987), Design Wave Heights on Very Mild Slopes: an Experimental
Study, Civil Eng. Trans., Inst. Eng. Australia, Vol. 29, pp 157-161.
Seabrook, SR, and Hall, KR, (1998), Wave Transmission at Submerged Rubble Mound
Breakwaters, Proc 26th Int Conf on Coastal Engineering, ASCE, pp 2000-2013.
Svendsen, LA, and Jonsson, IG, (1980), Hydrodynamics of Coastal Regions,
Technical University of Denmark.
Van der Meer, JW, and d'Angremond, K, (1991), Wave transmission at low-crested
structures, Coastal structures and breakwaters, Thomas Telford, London, England,
pp 25-42.
Van der Meer, JW and Pilarczyk, KW, (1991), Stability of Low Crested and Reef
Breakwaters, Proc. 22th Int. Conf. on Coastal Eng., ASCE, New York, USA.
Van Gent, MRA, (1995), Wave interaction with permeable coastal structures, PhD Thesis,
Delft University, Delft, The Netherlands.