TH File Compressed Main File 2mb

Gibraltar Harbour
Coaling Island Reclamation Study

Gibraltar Harbour Coaling Island
Reclamation Study
Bas Reijmerink
Maria Georgiou
Roderik Hoekstra
Sofia Caires
© Deltares, 2019
Title
Gibraltar Harbour Coaling Island Reclamation Study
Client Project Attribute Pages

Lievense 11204287-000 11204287-000-HYE-0001 63
Keywords
Gibraltar, Harbour, Coaling Island, land reclamation, extreme wave conditions, normal wave conditions,
hydrodynamics, flushing, sediment plume, SWAN, Delft3D-FLOW.
Summary
The government of Gibraltar is currently planning to build a new reclamation within the Gibraltar Harbour basin
in front of Coaling Island; this development is known to the public as the Victoria Keys. This reclamation will
have a certain impact on the local wave climate, hydrodynamics and ecology within the Gibraltar Harbour. This
impact needs to be described within an Environmental Impact Assessment (EIA). Lievense is contributing to
this and has commissioned Deltares to provide wave and hydrodynamic studies as input to this assessment.
In this study, two situations have been considered, being: (1) the present-day situation and (2) the situation
including the Coaling Island reclamation. The model results show that:
Wave conditions
• the easterly extreme conditions, EAST95 or EAST90, generally result in the severest wave conditions in
the Gibraltar Port waters;
• with the new land reclamation in place, wave energy entering through the southern port entrance is
partially dissipated on the rubble mound slopes, leading to an overall decrease of wave energy within
the port basins;
• only in a limited area along the inner side of the detached port breakwater, a small increase of wave
energy is expected due to reflection of wave energy off the planned vertical caisson head wall of the
Coaling Island reclamation;
• the Coaling Island reclamation provides a significant amount of sheltering of wave energy for the
entrances of the Cormorant Camber Marina and the Queensway Quay Marina, leading to milder wave
conditions in those areas
Flow conditions
• The flow in the project area is mainly driven by the tide and wind and reaches velocities in the order of
5cm/s under normal wind conditions. There is a minimal impact of the reclamation on the flow velocities,
showing local increase and decreases in the order of 1cm/s.
Flushing
• The flushing times in the project area are relatively short. The Small Boats Marina shows the highest
flushing times in the order of 6-7 days, mainly due to the limited connection to the open sea. The two
southern basins have slightly lower flushing times in the order of 5 days.
• The reclamation has a limited impact on the flushing times. The flushing times in the SBM are not
impacted and increase in the order of 1 day in the southern basins.
• The direction of normal wind conditions has no effect on the flushing times of the project area
Sediment dispersion
• The impact of the backhoe dredging activities is expected to stay relatively close to the location of the
dredge activity. Maximum sediment concentrations reach up to a few hundred mg/L, but quickly decrease
to typical background values 100-200m away.
• The sediment deposition as a result of the backhoe dredging is in the order of a few centimetres for the
short dredging operations (Scenario 1 and 2) and reaches up to a few decimetres for the long dredging
operation (Scenario 4). However, in line with the extent of the suspended sediment plumes, the impact
remains close to the location of the dredge activity.

Title
Client Project Attribute Pages

Lievense 11204287-000 11204287-000-HYE-0001 63
References
Offer: 11204110-000-HYE-0001, dated 15 April 2019.
Project award: MGRC-3-U-003, dated 25 April 2019.
Version Date Author Initials Review Initials Approval Initials

1.1 July 2019 S.P. Reijmerink M.P.C. de Jong J. Schouten
M. Georgiou
R. Hoekstra
S. Caires
1.2 August 2019 S.P. Reijmerink M.P.C. de Jong M.R.A. van Gent
M. Georgiou
R. Hoekstra
S. Caires
Status
final

11204287-000-HYE-0001, Version 1.2, August 7, 2019, final
Contents
1 Introduction 1
1.1 Background 1
1.2 Objectives 1
1.3 Approach 1
1.4 Project team 2
1.5 Outline of this report 2
2 Wave conditions 3
2.1 Introduction 3
2.2 Metocean data used for model input 3
2.2.1 200-yr Offshore wave and wind conditions 3
2.2.2 Normal offshore wave and wind conditions 4
2.2.3 Water levels 7
2.3 Wave modelling 7
2.3.1 Introduction 7
2.3.2 Model domains 8
2.3.3 Layout, bathymetry and structure schematizations 10
2.3.4 Output locations 12
2.3.5 Numerical and physics parameter settings 13
2.3.6 Boundary and input conditions 14
2.3.7 Run definitions 15
2.4 Modelling results for 200-yr wave conditions 16
2.4.1 Significant wave height and peak wave direction surface plots 16
2.4.2 Extreme wave conditions at the output locations 18
2.4.3 Influence of the land reclamation on the inner port wave conditions 19
2.4.4 Sensitivity of results to reflection levels Small Boats Marina caissons 22
2.5 Modelling results for normal wave conditions 23
2.6 Concluding remarks 27
3 Hydrodynamic and flushing conditions 28

3.1 Introduction 28
3.2 Hydrodynamic model setup 28
3.2.1 Computational grid 28
3.2.2 Bathymetry and other hydraulic-modelling features 30
3.2.3 Boundary conditions and forcing 32
3.2.4 Model sensitivity analysis 32
3.3 Model scenarios 32
3.4 Impact of reclamation on hydrodynamics 33
3.5 Impact of reclamation on flushing times 37
4 Sediment plume dispersion 44

4.1 Introduction 44
4.2 Model approach 44
4.2.1 Model starting points 44
4.2.2 Schematization of sediment spill terms 45
4.2.3 Dredge scenarios 46
Gibraltar Harbour Coaling Island Reclamation Study i

4.2.4 Evaluation criteria 49

4.3 Results of the sediment plume dispersion 50
4.3.1 General behaviour of sediment plume dispersion 50
4.3.2 Computed sediment plume dispersion 50
4.3.3 Suspended sediment concentrations near the sensitive receivers 55
4.3.4 Computed sediment deposition 56
4.5 Discussion 59
5 Key findings 61
References 62
Appendices
A Figures/Tables of wave study tasks A-1

A.1 Bathymetric sample data A-1
A.2 Interpolated bathymetry of numerical domains A-4
A.3 Reflection coefficients A-7
A.4 Extreme wave results A-10
A.4.1 Surface plots of the extreme significant wave heights A-10
A.4.2 Zoomed plots of extreme significant wave heights (Layout 1) A-34
A.4.3 Tables of wave results A-38
A.5 Normal wave results A-47
A.5.1 Zoomed plots of normal wave scenarios (Layout 1) A-47
A.5.2 Joint occurrence tables A-50
B Background on computation of flushing times B-1
C Results of hydrodynamics and flushing simulations C-1
D Turbidity assessment of backhoe dredging D-7
E Dredge plume simulations E-1

E.1 Input to simulations E-1
E.2 Simulation results E-5
Gibraltar Harbour Coaling Island Reclamation Study ii

1 Introduction
1.1 Background
The government of Gibraltar is currently planning to build a new reclamation in front of Coaling
Island within the Gibraltar Harbour basin (see Figure 1.1). This reclamation will have a certain
impact on the local wave climate, hydrodynamics and ecology within the Gibraltar Harbour.
Lievense is involved in working out the Environmental Impact Assessment (EIA) of the Coaling
Island reclamation and has commissioned Deltares to contribute with wave and hydrodynamic
studies to support this assessment.
1.2 Objectives
The study comprises the following tasks:
• Task 1: derive local estimates of 200-yr wave conditions for design purposes;
• Task 2: determine impacts of the reclamation on the normal wave conditions in Gibraltar
Harbour (EIA.4);
• Task 3: determine the impacts of proposed reclamation on flow conditions in Gibraltar
Harbour (EIA.5);
• Task 4: determine the impacts of proposed reclamation on residence times in the
Gibraltar Harbour (EIA.6); and
• Task 5: determine the dispersion of sediments during construction of perimeter
structure (EIA.7).
1.3 Approach
Deltares has since 2006 performed a number of studies involving the modelling of the waves
and hydrodynamics in Gibraltar waters. This study uses as basis the in-house regional models
developed in those studies. Particularly the studies in which offshore mean (WL|Delft
Hydraulics, 2006a and 2006b) and extreme (WL|Delft Hydraulics, 2006c and 2006d) wave
conditions have been defined and overall models developed. Also, studies focusing on
developments on the west side waters (Deltares, 2008, 2009a and 2009b) have been used, in
particular those supporting the development of the Small Boats Marina (Deltares, 2012, 2015a
and 2015b) and the most recent project for Lievense on the eastern side waters but in which
the bathymetry of the overall models domains has been updated (Deltares, 2019). These
models have been updated for the present study using bathymetry data and structure and
breakwater information provided to Deltares for this study by Lievense (e.g. Figure 1.1).
Gibraltar Harbour Coaling Island Reclamation Study 1 of 63

Figure 1.1 Layout of the new-to-be-constructed Coaling Island reclamation (detail of drawing MGRC-03-D-RD-SIT-
301 rev0b.dwg)
1.4 Project team

The project team consists of Bas Reijmerink (main reporter and wave modeler), Sofia Caires
(project manager and metocean expert), Robin Morelissen (flow, flushing and sedimentation
expert), Roderik Hoekstra (reporter of flow, flushing and sedimentation), Maria Georgiou (flow,
flushing and sedimentation modeler), Dick Mastbergen (spill rate expert and author of
Appendix D), Martijn de Jong (project initiator and reviewer) and Jan-Joost Schouten (final
quality assurer).
1.5 Outline of this report

In the next chapter the wave modelling and the determined extreme and normal wave
conditions (Tasks 1 and 2) are described. In Chapter 3 the impact of the Coaling Island
reclamation on the hydrodynamic and flushing conditions within the Gibraltar Harbour is
discussed (Tasks 3 and 4). Chapter 4 discusses the dispersion of the sediment plumes during
the different construction phases (Task 5). The report ends with a summary of the key findings
in Chapter 5.

2 Wave conditions
2.1 Introduction
This chapter describes the data and models used to derive the normal and extreme wave
conditions at the project site. In the next section the metocean data used as input and to force
the models is described. In Section 2.3 the setup of the wave model including considered
layouts, schematizations and inputs are described. In sections 2.4 and 2.5 the extreme (Task
1) and normal (Task 2) conditions are described, respectively. The chapter ends with
concluding remarks in Section 2.6.
2.2 Metocean data used for model input
2.2.1 200-yr Offshore wave and wind conditions

The extreme offshore wave and wind conditions determining the resulting conditions in
Gibraltar Harbour have been derived in previous studies using data from the locations given in
Figure 2.1. Gibraltar Harbour is affected by extreme wave conditions caused mainly by easterly
and westerly storms (e.g. Deltares, 2015b):
• The extreme easterly wave conditions were originally determined to support the Eastside
Development project (WL | Delft Hydraulics, 2006c). These conditions are taken from
operational wave model data from the European Centre for Medium-range Weather
Forecasts (ECMWF) for locations 3 and 4 in Figure 2.1. They have been calibrated using
Mar de Alboran (east of Gibraltar) buoy observations, represented as ‘Location 2’ in
Figure 2.1. The easterly wind conditions were determined in WL | Delft Hydraulics (2006c)
based on spatially averaged operational wind model data (HIRLAM) for Location 5a-d in
Figure 2.1. Those data were obtained via Puertos del Estado.
• The extreme westerly wave conditions were determined in WL | Delft Hydraulics (2006d).
These conditions are based on operational wave model (WANA) data of Puertos del
Estado, for Location 7 in Figure 2.1, calibrated using Cadiz (west of the Strait of Gibraltar)
wave buoy data and nearby WANA data, for Locations 1 and 6 in Figure 2.1. The westerly
wind conditions were determined in WL | Delft Hydraulics (2006d) based on operational
wind model data (HIRLAM), from Puertos del Estado for Location 7 in Figure 2.1.
The considered data sources covered the following periods:

• Mar de Alboran buoy: 1997 to 2005,
• the Puertos del Estado data for the western locations (Location 6 and 7): 1996 to 2005,
• the Puertos del Estado data for the eastern locations (Location 5 a-d): 1999 to 2005, and
• the ECMWF data: 1999 to 2005.
The above-mentioned studies considered conditions for return periods up to 100 years. In
WL | Delft Hydraulics (2009a) the corresponding 200-yr estimates for all conditions have been
derived. These are reproduced here in Table 2.12. Note that for both the easterly and westerly
conditions two mean wave and wind directions are considered and are applied at the relevant
incoming boundaries (either east of west) based on nearby input locations for those boundaries
(Figure 2.1)
2
The labels of these conditions have been taken from the original study. They indicate that a condition represents an east
or west condition and its mean direction (‘coming from’).

Figure 2.1 Overview of the locations of the data used in previous studies to derive the offshore wave and wind
conditions that determine the conditions inside the Gibraltar Harbour.
Table 2.1 200-yr offshore wave and wind conditions

Offshore wave conditions Wind conditions
Condition
Data point Hs [m] Tp [s] MWD [°N] U10 [m/s] Udir [°N]
Northeast 7.8 95
E090 90
Southeast 9.1 90
10.8 25.8
Northeast 7.8 100
E095 95
Southeast 9.1 95
W240 240 240
Atlantic 7.1 12.5 22.1
W270 270 270
2.2.2 Normal offshore wave and wind conditions

Normal offshore wave conditions were originally determined to support the Environmental
Impact Assessment of the East Side Development project (WL|Delft Hydraulics, 2006b). These
conditions are based on calibrated ECMWF Northeast (Location 3 in Figure 2.1) and Southeast
(Location 4) wave model data and raw operational wave model data (ECMWF) from location 8
(West). All datasets covered the same period (1999-2005). Figure 2.2 depicts the mean wave
climate roses as described by the ECMWF Northeast, Southeast and West data, respectively
(note the difference in scale for the West data).

NE
SE
W
Figure 2.2 Wave height roses ECMWF of respectively (top to bottom) Northeast, Southeast and West data (1999-
2005, WL|Delft Hydraulics, 2006b).
As input to the normal wave modelling task, a total of 66 climate scenarios (wind and waves)
were defined using the Northeast, Southeast and West wave data and the corresponding
spatially averaged HIRLAM wind data of locations 5a-d in Figure 2.1. An overview of these
scenarios is given in Table 2.2. Each wave scenario has an individual associated percentage
of occurrence per year (indicated in the second column of that table) and together they
represent in total of 98% of the year. With that they provide a suitable representation of the
normal wave climate.

Table 2.2 Overview of the 66 offshore wave and wind scenarios representing the mean wave climate.
Offshore wave conditions Wind
Occurrence
Northeast Southeast West Condition
Scenario per year
[%] Hs MWD Hs MWD Hs MWD U10 Udir
Tp [s] Tp [s] Tp [s]
[m] [°N] [m] [°N] [m] [°N] [m/s] [°N]
1 1.199 0.30 8.18 62.3 0.65 10.84 293.8 0.65 11.36 289.9 3.99 320.1
2 5.328 0.27 6.39 75.8 0.52 9.82 305.4 0.55 10.41 292.9 3.57 357.2
3 2.346 0.28 4.80 88.2 0.47 9.03 314.4 0.55 10.05 292.9 3.40 95.7
4 6.135 0.53 4.80 108.5 0.78 7.03 117.9 0.79 8.86 9.6 4.48 88.7
5 9.435 0.51 5.27 119.6 0.72 7.59 94.3 0.70 9.04 336.2 4.06 90.7
6 5.519 0.47 5.97 135.2 0.68 8.26 2.4 0.63 9.18 320.4 3.87 75.3
7 1.284 0.43 6.49 146.8 0.71 9.43 332.2 0.67 9.86 306.6 4.29 288.5
8 1.114 0.44 7.74 164.3 0.89 10.25 293.9 0.88 10.73 289.4 4.47 297.5
9 2.027 0.42 7.35 182.8 0.69 9.12 321.3 0.66 9.84 301.0 4.10 301.6
10 1.030 0.44 9.89 197.7 1.25 11.83 276.6 1.35 12.61 279.2 5.03 283.8
11 1.295 0.50 5.47 214.3 0.97 7.76 273.8 0.87 9.72 280.1 5.47 274.1
12 9.159 0.52 5.02 225.9 0.97 6.98 272.1 0.83 9.31 279.2 5.67 272.3
13 5.763 0.47 5.54 237.2 0.93 8.01 274.3 0.85 9.80 280.1 5.20 277.2
14 9.510 0.31 8.14 258.0 0.85 10.82 279.1 0.91 11.28 281.6 4.30 288.2
15 5.275 0.40 6.15 267.6 0.76 10.52 281.9 0.75 10.84 283.6 5.99 286.0
16 1.093 0.42 5.35 284.3 0.74 10.89 283.5 0.71 11.20 284.5 6.58 291.8
17 0.021 0.79 7.28 62.0 0.98 5.49 54.8 0.48 9.90 18.2 5.13 33.5
18 0.106 0.89 10.28 73.6 1.16 10.62 29.1 0.95 11.66 349.9 6.22 356.8
19 1.051 1.17 5.91 93.4 1.48 6.61 87.1 1.24 6.98 85.7 6.97 90.9
20 5.084 0.94 5.44 108.3 1.26 6.40 119.4 1.11 7.44 73.5 6.42 86.5
21 4.171 0.91 5.88 118.9 1.18 6.91 124.9 1.02 7.68 64.2 5.52 86.9
22 0.722 0.86 7.14 134.7 1.07 8.01 114.9 0.92 8.54 40.4 4.21 96.7
23 0.180 0.90 7.22 148.5 1.13 8.48 53.0 0.95 9.53 15.3 5.31 75.6
24 0.138 0.93 6.14 165.7 1.23 8.72 284.1 1.02 10.29 281.0 7.56 279.6
25 0.127 0.91 8.80 181.6 1.28 10.47 341.6 1.23 11.09 314.3 5.28 102.6
26 0.138 0.99 5.22 197.1 1.45 9.21 272.1 1.37 11.12 276.3 7.65 259.0
27 0.435 0.98 5.46 214.9 1.47 6.58 271.6 1.11 8.77 278.6 7.88 270.2
28 2.802 0.99 5.38 226.0 1.46 6.72 268.1 1.13 8.62 275.7 7.48 266.6
29 1.826 1.03 5.37 236.6 1.52 7.03 268.2 1.24 8.72 274.8 7.70 268.0
30 1.263 1.14 5.18 255.8 1.63 7.64 274.5 1.33 9.24 279.8 8.65 273.5
31 0.902 1.05 4.23 268.3 1.30 7.21 281.0 1.07 8.67 283.0 9.54 281.0
32 0.212 1.04 4.30 283.6 1.23 8.00 286.8 1.07 8.64 286.5 10.16 288.2
33 0.074 1.41 9.62 78.3 1.79 9.70 70.8 1.16 11.56 329.9 4.18 112.5
34 2.834 1.46 6.36 93.4 1.79 6.76 83.4 1.37 7.23 84.1 7.95 92.1
35 1.719 1.49 5.83 100.4 1.97 6.50 90.1 1.61 6.95 86.9 8.51 89.7
36 0.042 1.43 5.29 116.9 2.15 10.23 262.3 1.83 10.23 261.5 7.85 109.4
37 0.011 1.26 11.17 137.0 2.24 14.86 266.1 1.96 14.86 263.9 9.40 79.0
38 0.011 1.30 9.23 152.6 1.94 9.23 261.8 1.76 9.23 250.0 7.40 275.4
39 0.011 1.30 5.21 160.8 1.52 5.30 222.3 1.11 6.30 250.2 4.04 263.2
40 0.011 1.65 11.17 190.9 2.08 11.17 259.5 2.58 11.17 254.8 7.36 248.6
41 0.011 1.26 3.91 212.4 1.62 4.31 238.0 1.28 6.30 249.3 10.80 223.6
42 0.011 1.38 8.39 226.4 1.54 5.02 266.5 0.99 4.31 261.0 7.76 260.6
43 0.817 1.51 6.74 243.3 2.05 8.62 263.6 2.15 9.80 266.4 9.30 248.5
44 1.762 1.47 5.97 253.2 2.03 7.92 270.7 1.84 8.86 274.4 9.97 269.2
45 0.340 1.44 5.13 268.3 1.77 7.33 279.4 1.53 8.50 281.1 11.36 281.4
46 0.064 1.32 4.82 279.7 1.68 7.43 285.1 1.54 7.64 286.9 10.80 291.2
47 0.021 2.08 12.90 77.3 2.34 12.90 72.8 1.60 12.90 73.6 2.13 293.4
48 0.955 2.02 6.97 93.3 2.34 7.27 86.0 1.71 7.66 84.8 9.45 91.0
49 0.722 1.99 6.45 100.0 2.49 6.88 92.5 1.87 7.26 88.3 9.68 90.1
50 0.127 2.03 6.93 242.5 2.79 9.00 258.8 2.79 9.46 260.7 11.94 247.1
51 0.021 1.80 6.02 255.1 2.50 6.57 270.5 2.00 6.93 276.7 13.91 271.9
52 0.658 2.47 7.26 93.5 2.84 7.41 86.6 2.02 7.44 87.5 11.27 90.4
53 0.297 2.46 6.95 100.2 2.99 7.05 92.1 2.19 8.32 89.2 11.19 93.7
54 0.021 2.43 6.07 117.0 3.09 11.17 216.2 2.85 11.17 246.0 8.83 86.7
55 0.042 2.43 7.68 239.5 3.34 10.22 254.7 3.30 11.19 256.2 10.88 242.3
56 0.011 2.88 12.28 82.1 3.63 10.32 75.2 2.11 12.28 78.9 3.00 122.3
57 0.297 2.99 7.84 93.0 3.64 7.95 86.8 2.35 7.75 89.0 11.72 94.7
58 0.117 2.89 7.26 99.7 3.56 7.33 92.2 2.51 8.13 93.9 12.43 97.2
59 0.011 2.97 6.30 115.3 3.72 11.17 192.1 3.18 11.17 242.2 7.90 142.8
60 0.149 3.59 8.29 92.9 4.21 8.29 85.7 2.40 8.12 92.1 12.42 90.1
61 0.011 3.55 7.63 98.0 4.24 7.63 90.5 2.37 6.30 93.5 10.50 82.4
62 0.011 3.31 7.63 240.5 4.04 8.39 255.8 3.31 9.23 260.9 12.78 279.6
63 0.149 4.12 8.34 91.9 4.57 8.35 85.3 2.85 8.33 86.7 14.89 88.7
64 0.074 4.42 8.75 93.9 5.22 8.75 87.4 3.04 8.57 90.6 14.27 90.8
65 0.053 4.91 9.23 93.6 6.09 9.24 87.5 3.41 9.41 89.7 15.76 89.8
66 0.011 5.92 9.23 94.1 6.31 9.23 89.0 4.18 9.23 90.1 20.86 85.4

2.2.3 Water levels
Extreme conditions
Extreme water level statisctics are available from WL | Delft Hydraulics (2006a) and
Deltares (2008). They are based on numerical model results by Deltares and extreme value
analysis of measurements in the Port of Gibraltar for the period of 1967-1982. As agreed upon
with Lievense, the 200-yr wave conditions were computed considering a sea level rise of 50
cm. This figure is taken as a projection of the sea level rise 100 years from now. Furthermore,
for the extreme conditions the scope of work also included computations without sea level rise
to assess the influence of sea level rise on the nearshore extreme wave values. The considered
water levels for the computations of extreme wave conditions are presented in Table 2.3.
Table 2.3 200-yr return value estimates of the water level based on measurements in the West Port of Gibraltar
in- and excluding sea level rise in 100 years (water levels relative to Gibraltar Ordnance Datum (GOD)).
Tidal component Total still water level
Surge component [m] Sea level rise [m]
[mGOD] [mGOD]
0.50 1.69
0.93 (MHWS) 0.28
- 1.19
Normal conditions
For the normal wave conditions a constant water level at Mean Sea Level (MSL) was
considered for all scenarios: MSL = GOD +0.43 m (GOD = Gibraltar Ordnance Datum).
2.3 Wave modelling
2.3.1 Introduction
Numerical wave modelling has been performed to determine, based on the offshore wave, wind
and water level conditions described in the previous section, the normal and extreme wave
conditions at the new-to-be-constructed Coaling Island reclamation. Separate calculations
were made for the present-day situation and for the future layout.
The third-generation shallow water wave model SWAN (Simulating WAves Nearshore;
Booij et al, 1999, version 41.20) is used as basis for the modelling. SWAN has been developed
at the Delft University of Technology (e.g., Van der Westhuysen, 2010 and Zijlema, 2010) with
contributions of Deltares. It computes wave propagation and energy evolution very efficiently
and it includes a number of non-linear effects via parameterised formulations. More specifically,
SWAN can account for:
• Wave propagation in time and space (from laboratory up to global scales), shoaling,
refraction due to current and depth, frequency shifting due to currents and non-uniform
depth;
• Wave generation by wind;
• Three- and four-wave interactions;
• White-capping, bottom friction and depth-induced breaking;
• Wave energy dissipation due to vegetation;
• Wave-induced set-up;
• Transmission through and reflection (specular and diffuse) from obstacles; and
• Diffraction.
SWAN is widely used for nearshore wave modelling and has been successfully validated under
a variety of field cases. The software is continually undergoing further development; see
www.swan.tudelft.nl for more information. For this study we have used the latest operational
version that includes the most recent insights and model developments.

2.3.2 Model domains

SWAN requires the specification of three types of grids:
1 computational grid which defines the 2D geographical locations of the points in the
calculation grid;
2 directional grid which defines the directional range (usually 360°) and resolution;
3 spectral grid which defines the range and resolution of the computations in frequency
space.
In line with the computational domains used in previous studies by Deltares (Deltares, 2012,
2015a and 2015b), the following three computational domains were used in this study:
• Grid 0: an overall domain (250 m x 250 m, green),

• Grid 1: an intermediate domain (50 m x 50 m, red), and
• Grid 2: a detail domain (10 m x 10 m, blue).
The three domains are presented in Figure 2.3 and Figure 2.4. Each domain provides input for
boundary conditions for the next, smaller and finer, calculation grid (nesting). Please note that,
in accordance with the main directions of the wave condition considered, for the easterly
extreme conditions the overall domain is smaller than that for the westerly extreme conditions
(indicated by the dashed white line in Figure 2.3). This approach was taken in view of efficient
calculation times, by leaving out sections of the calculation grid not influencing the project site
under those particular conditions. For the normal conditions, the smaller overall domain was
also used as these conditions are not expected to be influenced by wave conditions generated
on the Atlantic Ocean.
The directional space in SWAN covers the full circle (360°). The number of directional bins was
set to 36, resulting in a directional resolution of 10°.
The spectral space covers a frequency range from 0.03 Hz to 2.5 Hz, allowing for wave period
of 0.4 – 33.3 seconds. The distribution of the frequencies, f, is logarithmic with a constant
relative resolution, Δf/f, close to 0.1 given that the number of frequency bins is set to 47.

Figure 2.3 Computational SWAN domains. The dashed white line indicates the edge of the western area cut-off from
the overall domain (Grid 0) for all normal and for the easterly extreme conditions..
Figure 2.4 Zoom of intermediate (Grid 1) and detail (Grid 2) computational SWAN domains.

2.3.3 Layout, bathymetry and structure schematizations

The modelling has carried out for two layouts (geometries) and bottom schematizations:
• Layout 0 - The current situation including the Small Boats Marina.

• Layout 1 - Taking into account the future land reclamation, with the layout derived from
drawing “MGRC-03-D-RD-SIT-301.dwg” as provided by Lievense.
These have been agreed upon with Lievense, who have revised previously applied layouts
using collected local information, before sharing the layout information with Deltares.
The depth schematisation in the area of interest is based on the following bathymetric datasets
with increasing level of detail (but with smaller spatial extents):
1 the GEBCO 2019 dataset (General Bathymetric Chart of the Oceans3) with a resolution
of 15 arcseconds (~460 m near Gibraltar, Figure A.1 in Appendix A.1),
2 digitized navigational charts with varying resolution (10-250 m, Figure A.2 in
Appendix A.1),
3 in-house depth schematization of inner port waters of the most recent previously applied
wave model (Deltares, 2015), and
4 isolines of a recent local depth survey provided by Lievense (Figure A.3 in Appendix A.1).
These data were interpolated to the computational grids of the SWAN model, starting from the
most detailed data (depth survey) on the finest grid and working towards the least detailed data
(GEBCO) on the coarsest grid. Consistency between the different datasets (including vertical
and horizontal references used)4 was checked and ensured as part of that process. The
resulting bathymetry as applied in the numerical model is presented in Figure A.4 to Figure A.6
in Appendix A.2 for respectively the overall, intermediate and detail (Layout 0) domain.
Figure 2.5 shows the bathymetry used in the detail domain of Layout 1. Please note that the
future situation, i.e. the layout of the Coaling Island reclamation of drawing MGRC-03-D-RD-
SIT-301.dwg (cf. Figure 1.1), has been implemented in the different computational domains to
the level of detail that each grid allows.
Breakwaters and structures are schematized in the SWAN computations by means of obstacles
with structure-specific reflection coefficients. Therefore, reflection coefficients need to be
specified for all edge structures in the project area. For the Gibraltar Bay area, we apply the
reflection coefficients used in our previous studies as shown in Figure A.7 for Layout 15 in
Appendix A. The computations are made under the assumption that there is no transmission
(overtopping) of wave energy over these structures.
For the inner Gibraltar port domain, the following assumptions hold:
• Inner port/marina mooring structures (jetties) have no significant influence on the overall
wave patterns and are ignored.
• The outer walls of the southwest facing caissons of the Small Boats Marina are perforated
and contain wave chambers in which wave energy is dissipated. The effectiveness of
such wave chambers depends on the local incoming wave conditions (wave length/period
and incoming angle) and thus varies over time. In the modelling only one reflection
3
https://www.gebco.net
4
Horizontal coordinate system: ETRS89 / UTM zone 30N (EPSG: 25830), vertical reference: Gibraltar Ordnance Datum.
5
The differences in reflection boundaries between Layout 0 and Layout 1 are only shown for the detail wave domains as
they are better visible at that zoom level.

coefficient value representing these structures has been considered for all wave
conditions. In consultation with the Lievense, this reflection coefficient was set to 70%6.
All other vertical (caisson) walls have a reflection coefficient of 95%.
• The caisson walls of the Coaling Island reclamation are schematized with a reflection
coefficient of 95% and the rubble mound slopes schematized with a reflection coefficient
of 40%.
• The rubble mound spur at the entrance of the Cormorant Camber Marina and Queensway
Quay Marina is schematized in the present situation (Layout 0) with a reflection coefficient
of 40%. In the future situation (Layout 1) it has been removed entirely.
The reflection coefficients as applied in the detail domains of Layout 0 and Layout 1 are shown
in respectively Figure A.8 and Figure A.9 of Appendix A.3.
6
To check the sensitivity of the wave model results to this value, an additional sensitivity run was performed for Layout 0
in which the reflection coefficient of the perforated wall was lowered to 50%. The result of this test is discussed in
Section 2.4.4.

Figure 2.5 Bathymetry of the SWAN detail wave domain for the future situation (Layout 1). The dashed blue line
outlines the detail model domain. The crosses indicate the output locations requested by the Client.
2.3.4 Output locations

Output is provided at a total of 33 locations distributed over the Gibraltar Port area. The output
locations are shown in Figure 2.5. They have been provided to Deltares by the Client. The
coordinates of the output locations are listed in Table 2.4. This table also mentions the bed
level at each output location relative to the local datum according to the bathymetry
schematization of the model.

Table 2.4 Coordinates of output locations (ETRS89 / UTM zone 30N, EPSG: 25830) and the local bed level (relative
to GOD).
Easting Bed level Easting Northing Bed level
Point Northing [m] Point
[m] [mGOD] [m] [m] [mGOD]
P1 287795.3 4000716.2 -11.09 P18 287579.5 4001995.6 -9.72
P2 287588.3 4000997.2 -11.44 P19 287639.1 4001823.7 -9.82
P3 287345.3 4001318.2 -13.41 P20 287485.7 4001772.0 -10.09
P4 287279.3 4001655.2 -13.09 P21 287551.2 4001573.9 -10.69
P5 287164.3 4001963.2 -12.29 P22 287637.4 4001433.2 -11.13
P6 287041.3 4002309.2 -10.88 P23 287795.7 4001480.0 -9.13
P7 287192.3 4002553.2 -10.31 P24 287836.3 4001626.2 -9.02
P8 287200.3 4002862.2 -9.49 P25 287911.3 4001343.2 -11.31
P9 287441.3 4002828.2 -9.44 P26 287962.3 4001185.2 -8.81
P10 287578.9 4002826.2 -7.61 P27 288015.0 4001027.7 -8.61
P11 287469.3 4002617.2 -9.86 P28 287728.7 4001173.0 -12.38
P12 287476.3 4002399.2 -9.76 P29 287270.3 4001481.2 -16.27
P13 287689.3 4002236.2 -4.53 P30 287410.8 4001527.5 -13.06
P14 287782.3 4002034.2 -5.16 P31 287363.3 4002165.2 -11.34
P15 287697.3 4001964.2 -9.23 P32 287030.3 4002456.2 -20.68
P16 287640.3 4002144.2 -9.00 P33 287289.3 4002437.2 -11.74
P17 287519.9 4002167.5 -10.13
2.3.5 Numerical and physics parameter settings

This section describes detailed settings for numerical and physics parameters within the SWAN
model. It is primarily included here for recording purposes, e.g. for possible future interpretation
or reproduction of results.
The modelling was carried out using SWAN, version 41.20, in stationary mode. The most
relevant applied wave physics settings are:
• Dissipation of wave energy by bottom friction and wave breaking (wave steepness-
induced and depth-induced) have both been activated in the SWAN computations.
– For dissipation by bottom friction the JONSWAP formulation (Hasselmann et al.,
1973) with a friction coefficient of 0.038 m2s-3 (Zijlema et al., 2012) has been applied.
– For dissipation by depth-induced wave breaking the Battjes-Janssen formulation
(Battjes and Janssen, 1978) with a proportionality coefficient of 0.73 has been
applied.
• For representing the effects of white-capping, the formulations by Rogers et al. (2003),
default since SWAN version 40.91, have been applied.
• For the wind drag the default Wu (1982) approximation of the Charnock relation is set,
which describes the effect of the wind over the disturbed water surface.
• The approximation of diffraction available in SWAN was activated in all runs.
The applied criteria for numerical accuracy were set as follows:
• the computation is finished in case of changes of less than 1% in Hs and Tm0,1 at 99.9%
of the grid points relatively to the previous computational iteration rounds, and
• a maximal number of 80 iterations
All runs have been verified to have converged within 80 iterations.

2.3.6 Boundary and input conditions
Boundary conditions
The SWAN model was forced at the outer boundaries of the overall domain (Grid 0) with
parameterized wave spectra described by five wave parameters following from Section 2.2.1
(extreme wave conditions) and Section 2.2.2 (normal wave conditions):
• Significant wave height, Hs
• Peak wave period, Tp
• Mean wave direction (coming from), MWD
• Directional spreading, σ/m
• Spectral shape, γ
The spectral shape, γ, was assumed to be constant for all computations, being a JONSWAP
shape (Hasselmann et al., 1973) with a value of γ = 3.3. The amount of directional spreading
present at the incoming boundaries was also assumed to be constant along the boundaries (σ
= 31.5°/ m = 2).
Extreme conditions
In case of the easterly extreme conditions, there are no (incoming) boundary conditions
imposed on the Atlantic side of the overall grid (i.e. Grid 0, indicated in green grid in Figure 2.3)
and the conditions of the Northeast and Southeast data points (Table 2.1) are applied along
the Mediterranean boundaries of the overall grid as follows:
• the conditions of the Northeast data point are applied along the eastern boundary of the
overall grid from the coast in the north down to halfway the Northeast and Southeast data
point locations (cf. Figure 2.1);
• further southward of that point the conditions of the Northeast data point are scaled
linearly up to the conditions of the Southeast data point at the southeast edge of the
overall grid; finally
• the conditions of the Southeast data point (Table 2.1) are applied uniformly along the
Mediterranean extent of the southern boundary of the overall grid (cf. Figure 2.3).
In case of westerly extreme conditions, the conditions of the Western data point are applied
uniformly along the Atlantic boundary the overall grid and there are no (incoming) boundary
conditions are imposed on Mediterranean side of the overall grid.
Normal conditions
For the normal conditions the wave boundary conditions of Table 2.2 are applied at all of the
boundaries of the largest computational domain. At the eastern boundary of the overall grid the
same variations along the boundary locations are applied as for the easterly extreme conditions
described above. At the western boundary, the wave conditions are applied uniformly along the
boundary. The SWAN modelling software considers which wave components enter the domain
and which are directed outwards at a given model boundary.
Wind input
The SWAN model was forced using uniform wind fields corresponding to the wind speed and
wind directions as reported in Sections 2.2.1 and 2.2.2 for respectively the extreme and normal
conditions. This is an often-used and proven approach to including the effect of wind in such
computations.

Water level input

The water levels of Section 2.2.3 were applied as uniform water levels over all computational
domains. Possible (small) local differences in these water levels are expected to have a minor
influence on the results.
Effect of currents
In principle, strong (sheers in) currents can influence wave patterns and wave heights.
According to a previous study by Deltares on flow conditions around the Gibraltar area
(WL | Delft Hydraulics, 2006a), the 100-yr magnitude of currents on the east side of Gibraltar
is a few decimeters per second. The expected 200-yr magnitude of currents will be less than
half a meter per second. Such current magnitudes will not have a strong influence on local
wave propagation and wave heights. Therefore, the effect of currents is not considered in the
SWAN modelling.
2.3.7 Run definitions
Extreme conditions
In total 17 SWAN computations have been carried out for the 200-yr conditions:
• eight runs representing combinations of two water levels and four directional settings for
the present-day situation (Layout 0),
• eight runs representing combinations of two water levels and four directional settings for
the future situation (Layout 1), and
• one sensitivity run considering a reflection coefficient of 50% for the outer edge of the
southwest facing caissons of the Small Boats Marina for the present-day situation
(Layout 0).
All combinations of input parameters as applied in the different computational runs for the
extreme conditions considered are summarized in Table 2.5. The run-IDs mentioned in
Table 2.5 are used in the next chapter to refer to specific runs/results. For consistency, they
have been based on the original labels for these conditions from previous studies (e.g.
WL | Delft Hydraulics, 2006c).

Table 2.5 Overview of boundary conditions for all extreme wave simulations (LO is layout, Hs is significant wave
height, Tp is peak wave period, MWD is mean wave direction, WL is still water level including/excluding 50cm
sea level rise, U10 is wind speed and Udir is wind direction)
Run LO Hs Tp MWD WL U10 Udir
Run-ID Data point
number [-] [m] [s] [°N] [mGOD] [m/s] [°N]
Northeast 7.8 95
01 E090_SLR_L0 90
Southeast 9.1 90
10.8 25.8
Northeast 7.8 100
02 E095_SLR_L0 1.69 95
Southeast 9.1 95
03 W240_SLR_L0 240 240
West 7.1 12.5 22.1
04 W270_SLR_L0 270 270
0
Northeast 7.8 95
05 E090_L0 90
Southeast 9.1 90
10.8 25.8
Northeast 7.8 100
06 E095_L0 1.19 95
Southeast 9.1 95
07 W240_L0 240 240
West 7.1 12.5 22.1
08 W270_L0 270 270
Northeast 7.8 95
09 E090_SLR_L1 90
Southeast 9.1 90
10.8 25.8
Northeast 7.8 100
10 E095_SLR_L1 1.69 95
Southeast 9.1 95
11 W240_SLR_L1 240 240
West 7.1 12.5 22.1
12 W270_SLR_L1 270 270
1
Northeast 7.8 95
13 E090_L1 90
Southeast 9.1 90
10.8 25.8
Northeast 7.8 100
14 E095_L1 1.19 95
Southeast 9.1 95
15 W240_L1 240 240
West 7.1 12.5 22.1
16 W270_L1 270 270
Northeast 7.8 100
17 0 E095_SLR_L0_R50 10.8 1.69 25.8 95
Southeast 9.1 95
Normal conditions
The wave and wind conditions of the 66 scenarios used to represent the normal wave
conditions were applied to force the model as described in Section 2.3.6.
For each of these scenarios the SWAN computations were carried out and the mean wave
climate at the locations of interest was obtained by combining results from the individual
computations. Deriving the local wave climate from the results of the scenario runs was based
on the relative occurrence of the different source wave conditions in the original offshore
dataset of ECMWF (i.e. by means of wave transformation).
2.4 Modelling results for 200-yr wave conditions

In this section the results of the SWAN computations for the 200-yr conditions are presented.
Section 2.4.1 describes the resulting wave fields by means of surface plots of significant wave
height and peak wave directions. Section 2.4.2 presents the wave parameter values at the
selected output locations for the total and the incoming part of the wave energy in tabular form.
The influence of the reclamation on the wave conditions in the Gibraltar Harbour area is
discussed in Section 2.4.3. Last, in Section 2.4.4 the sensitivity of the wave conditions inside
Gibraltar Harbour to the reflection levels of the Small Boats Marina caissons is discussed.
2.4.1 Significant wave height and peak wave direction surface plots
Figure A.10 to Figure A.33, distributed per offshore wave condition over the sub-sections of
Appendix A.4.1, present the surface plots of significant wave height and peak wave direction
for the first 16 runs (cf. Table 2.5). The surface plots (three per run) represent each of the three

domains. Each figure contains two panels: one panel representing the present-day situation
(Layout 0) and the other panel the future situation (Layout 1). In each panel, the significant
wave height is given in colour and the peak wave directions are presented as black arrows.
Each panel also presents the bathymetry contour lines as grey lines for specific bottom level
intervals appropriate for each of the domains. In addition, Figure A.34 to Figure A.41 in
Appendix A.4.2 present zoomed plots of significant wave height and peak wave direction for
only the future situation (Layout 1) focussing on the entrances of the Cormorant Camber Marina
and the Queensway Quay Marina. It is noted that the “staircases” visible along the edges of
some of the plots are related to the model resolution in combination with the plotting routines,
but do not influence the overall quality and accuracy of the model results.
From the figures it is concluded that easterly storms yield the most severe wave conditions at
the area of interest; the westerly conditions are not likely to be normative for the design of the
reclamation in terms of significant wave height and wave periods (not shown). Taking the
EAST95 storm as an example (considered further below), the wave field East of Gibraltar in the
overall domain is rather uniform with the waves being slightly higher in the southern part (see
Figure A.13). The wave height remains high until close to the coast due to the large water depth
and steep foreshore. As the waves reach the southern tip of Gibraltar (Europa Point), land
blocking and diffraction occurs, and the peak wave direction becomes south-easterly (see
Figure A.14). Large reflection patterns are observed where the waves reflect against the
harbour extensions of Algeciras. This is mainly due to the high reflection that is caused by those
structures and by their orientation from southwest to northeast. Smaller areas with reflected
wave energy are observed near the vertical wall sections in the northern part of the Bay and
just outside Gibraltar Harbour waters.
From the easterly conditions, the EAST095 storm condition yields the most severe wave
conditions within the port area and along the planned reclamation. Note that the variations in
wave parameters at the output locations inside the model domain between the EAST090 and
the EAST095 conditions are only minor (at most variations of 0.1 m in significant wave height,
0.2 s in wave periods and 10 degree in peak wave direction). It should also be noted that within
the port area, several wave systems can occur simultaneously (combination of locally
generated wind waves, swell wave propagating into the port and multiple reflections from the
obstacles). The presented peak wave direction is, therefore, in most of the cases a direction
that does not correspond to only one wave system. Care must, therefore, be taken when
interpreting the presented wave directions.
Due to the wave sheltering effect that the Iberian peninsula has on the bay of Gibraltar, only a
limited part of the westerly (lower frequency) wave energy, coming through the Strait of
Gibraltar from the Atlantic Ocean, arrives at the project site: a significant part of the wave energy
is locally generated by wind (higher frequency wave energy). This results in low values of the
considered mean wave periods, Tm0,1 and Tm0,2. Given that the westerly conditions comprise
more than one wave system, we recommend that the given low values of the Tm0,1 and Tm0,2
periods are not considered as basis for design, but the peak periods (or the spectral wave
periods, Tm-1,0) given that those correspond to the most energetic wave system in the sea state.
A comparison of the 200-yr return period runs including and excluding 0.5 m sea level rise
shows that along the edges of the land reclamation the differences in the wave parameters are
negligible due to the relatively large water depths present in the port area: the relative change
in water depth as result of sea level rise is small.

2.4.2 Extreme wave conditions at the output locations
Total wave spectrum

Table A.1 and Table A.2 in Appendix A.4.3.1 present the resulting wave parameters at the
selected output locations (cf. Table 2.4 and Figure 2.5) for Layout 0. Table A.3 and Table A.4
do the same for Layout 1. The printed values are the computed values rounded off to the
nearest 0.1 m (either down or up), 0.1 s or angular degree. This rounding-off effect is expected
to fall within practical accuracy of the SWAN model outcomes. The observations made for the
surface plots of Section 2.4.1 (wave heights and directions) are consistent with the output data
presented in these tables. The tables also include different representations of wave periods
derived from the computed wave spectra (Tm0,2, Tm0,1, Tm-1,0)7. These are therewith available to
serve as input to different design aspects, where relevant/applicable.
As can be seen from the output tables, the E095 and E090 storms are yielding more or less
the same significant wave heights and these are generally larger than wave heights resulting
from the W240 and W270 storms at most output locations8. The difference of significant wave
height between the wave heights for the E095 storm and E090 storm is maximum 0.1 m (E095
leading to the most severe conditions).
Incoming wave spectrum

Designs of structures and revetments are based on the incoming wave energy reaching that
particular structure. Since the total sea states within the port area are combinations of different
wave systems (locally generated wind waves, swell waves originating from distant storms and
reflected waves), the incoming parts of the total wave spectra, only make up a portion of the
total wave spectrum. Focussing only on incoming waves therefore results in more precise
estimates of the normative wave values to serve as input to design and generally lead to lower
values than otherwise would have been obtained by considering the total wave spectrum
irrespective of direction. Therefore, incoming wave parameters have been determined for the
output locations closest to the land reclamation (P19 to P24). The directional sector
representing incoming waves was determined for each of these locations separately and
depended on the particular shoreline orientation at that output location (see Table 2.6 and
Figure 2.6).
The resulting incoming wave parameters are presented in Table A.6 and Table A.7 of Appendix
A.4.3.2 for the runs considering Layout 0 (Run 01 to Run 08) and in Table A.8 and Table A.9
for the runs considering Layout 1 (Run 09 to Run 16).
From these tables it is found that for locations P19 and P20 the highest incoming waves are
found for the westerly storm conditions. However, the wave lengths are shorter than for the
easterly storm conditions. For locations P21 to P24 the highest incoming waves are found for
the easterly storm conditions.
7
The spectral wave period Tm0,2 corresponds to Tz, the mean wave period of a temporal wave-by-wave analysis, with
waves being separated by crossings of the mean water level line. Tm0,1 only represents the centre of gravity of the 1D
wave spectrum (computed by dividing the spectral zeroth order moment by the spectral first order moment: m0/m1),
it gives more weight to the longer waves than Tm0,2 ( √(zeroth order moment/second order moment): √(m0/m2)).
Tm-1,0 gives even more weight to the longer waves (m-1/m0).
8
Only at locations P18 to P20 (in front of the caissons of the SBM) and P25 to P27 (southeastern harbor area) slightly
higher wave heights are found for a few runs with westerly wave conditions, while the corresponding peak periods
are smaller (waves generated locally).

Table 2.6 Directional sectors containing incoming wave energy for output locations P19-P24 alongside the
Coaling Island reclamation.
Output location From [°N] To [°N] Sector size
P19 250 60 180
P20 160 60 270
P21 160 340 180
P22 110 290 180
P23 30 210 180
P24 70 190 120
Figure 2.6 Overview of incoming directional sectors for output locations P19-P24 (see Table 2.6).
2.4.3 Influence of the land reclamation on the inner port wave conditions
The construction of a structure or reclamation, like the Coaling Island reclamation, in confined
water bodies (e.g. ports) may result in a change of wave action within that area. To avoid that
problems arise within the Gibraltar Harbour area due to changes in wave loads it is important
to determine how strong the influence of the new reclamation on the wave conditions in the
area will be. For this reason, difference plots of significant wave height (Hs) between the future
and present-day situation have been made for each of the considered storm conditions, both
including sea level rise (Layout 1 – Layout 0). These plots are presented in Figure 2.7 for the
easterly storms and in Figure 2.8 for the westerly storms.

Figure 2.7 Surface plots of differences in significant wave height of detail domain including 5 m (>20 m) / 2 m interval
bathymetry contour lines (grey). Bottom panel: Run 09-Run 01 (E090), top panel: Run 10-Run 02 (E095).

Figure 2.8 Surface plots of differences in significant wave height of detail domain including 5 m (>20 m) / 2 m interval
bathymetry contour lines (grey). Bottom panel: Run 11-Run 03 (W240), top panel: Run 12-Run 04 (W270).

From the differences plot it follows that in general the construction of the land reclamation
results in a lowering of the significant wave heights within the port waters. This is because of
sheltering effects of the reclamation shape and because a part of the wave energy entering
through the southern port entrance is dissipated on the rubble mound slopes of the Coaling
Island reclamation, where before wave energy penetrated further into the harbour area or
reflected off other structures. The largest changes are observed (~-1 m wave height) near the
entrances of the existing Cormorant Camber Marina and the Queensway Quay Marina: the
previously open entrance will be sheltered by the land mass of the reclamation. This is observed
for both the easterly and westerly storm conditions. It is noted that the decrease of the wave
heights within the marinas itself is less than near the entrance. This can be explained by the
fact that the wave energy within the marinas itself originates mostly from local wave generation
rather than from wave penetration (especially for the Queensway Quay Marina) and by the fact
that the spur near the entrance is removed.
Another area where a significant lowering in wave energy is observed is the shoulder of the
SBM and the reclamation (where the caissons meet). This is especially the case for the E090
and W270 storm conditions and is due to sheltering caused by the reclamation, preventing
wave energy from reaching that port area after entering the harbour through the southern port
entrance.
An area of attention is the inner side of the detached port breakwater that is also used as
berthing location for large vessels. For the easterly storm conditions, the wave heights along
this breakwater are lowered after the construction of the land reclamation. Less wave energy
will reach this side of the breakwater due to an increase of dissipation of wave energy on the
rubble mound slopes of the land reclamation. For the westerly storm conditions however, the
wave heights will slightly increase (+0.05 - +0.15 m) along part of this breakwater. This increase
in wave height is caused by waves reflecting off the vertical head wall of the northerly caissons
(R = 95%). This increase is most clearly observed for the W240 storm condition at coordinate
(287200, 4001800) (bottom panel of Figure 2.8).
2.4.4 Sensitivity of results to reflection levels Small Boats Marina caissons

As mentioned in Section 2.3.3, the assumed effectiveness of the wave chambers within the
outer walls of the southwest-facing caissons of the SBM in dissipating wave energy for different
wave lengths was represented by a reflection coefficient of 70% in all computations. From
literature however, it is known that such wave chambers are more effective in dissipating wave
energy for certain wave lengths (typically short waves) than for others. To assess the sensitivity
of the modelling choices made on the computed wave conditions within the port, an additional
sensitivity run was performed for the E095 storm condition (leading to the highest wave
condition within the port) for Layout 0 in which the reflection coefficient of the southwest-facing
caissons of the SBM was lowered to 50% (i.e. the wave chambers are more effective). The
results of this run are presented as difference plot with respect to Run 02 in Figure 2.9 and as
wave parameters at the output locations in Table A.5 of Appendix A.4.3.1.
The results show that accounting for 20% less reflection of wave energy, leads to a decrease
in wave height of at most 0.20 m in front of the SBM caissons and at most 0.15 m in front of
the inner side of the detached port breakwater. As expected, the influence on the wave periods
is only minor (typically -0.1 s).

Figure 2.9 Surface plots of differences in significant wave height (Run 17: R50% - R02 R70%) of detail domain
including 5 m (>20 m) / 2 m interval bathymetry contour lines (grey).
2.5 Modelling results for normal wave conditions

For each of the 66 scenarios of Table 2.2 SWAN computations were carried out to obtain the
normal wave climates at the selected output locations. In Appendix A.5.1 examples of
significant wave height surface plots near the Coaling Island reclamation are given for six of
the most frequently occuring mean wave and wind scenarios (Scenarios 05, 12, 14, 28, 34 and
44).
The mean wave conditions at the selected output locations are obtained directly/empirically
from the (transformed) time series of wave parameters. Figure 2.10 presents the annual wave
significant wave height roses for a selection of mid-harbour output locations for both Layout 0
and Layout 1, whereas the annual peak wave period roses for both layouts are given in
Figure 2.11. The roses were computed using the Hs, Tp and MWD time series. The numerical
values within the central circles of the roses represent the percentages of time per year for
which the specific parameter has a value lower than the lowest class value (0.30 m and 5.0 s
for respectively Hs and Tp).
We note that also for the normal conditions the wave climate inside the port is characterised by
multiple wave systems from different directions: a combination of local wind sea, swell waves
propagating into the harbour and multiple reflections from the obstacles. The mean wave

direction is, therefore, in most of the cases a direction that is not corresponding to any of these
specific individual wave systems. Care must, therefore, be taken when interpreting the
presented mean wave directions.
When comparing the panels of Figure 2.10, the sheltering effect of the land reclamation is made
visible. The percentages of time for which the wave heights are lower than 0.30 m (often used
as criterion for operational use of marinas) increases or remains equal for all locations and the
wave energy becomes more restricted to a smaller directional range. At locations P23 and P24,
these percentages are changed to (almost) 100% of the time below the criterion after
construction of the land reclamation.
From Figure 2.11 it is observed that in general terms the waves within the port become
somewhat shorter (peak wave periods decrease). Only on the locations in front of the land
reclamation (P3, P4, P20-P21 and P29), the waves become somewhat longer (peak wave
periods increase).
The resulting local wave climates have also been presented as joint-occurrence tables of Hs
(significant wave height) versus MWD (mean wave direction), Hs versus Tp (peak wave period)
and Hs versus Tp for all output locations are presented in Figure A.48 to Figure A.80 in
Appendix A.5.2.1 for Layout 0 and in Figure A.81 to Figure A.113 in Appendix A.5.2.2 for
Layout 1. It is noted that the same bin classes have been used as for the wave roses to classify
the wave data. The observations made for the wave roses are consistent with the information
presented in the joint occurrence tables.

Figure 2.10 Normal significant wave height (Hs) roses at a selection of mid-harbour output locations. Lower panel:
Layout 0, upper panel: Layout 1.

Figure 2.11 Normal peak wave period (Tp) roses at a selection of mid-harbour output locations. Lower panel:
Layout 0, upper panel: Layout 1.

2.6 Concluding remarks

Two port layouts have been considered with SWAN for four 200-year storm conditions and
mean annual wave conditions, being: (1) the present-day situation and (2) the situation
including the Coaling Island reclamation. The model results show that:
• the easterly extreme conditions, EAST95 or EAST90, generally result in the severest
wave conditions in the Gibraltar Port waters;
• in presence of the new land reclamation, wave energy entering through the southern port
entrance is partially dissipated on the rubble mound slopes of the new reclamation,
leading to an overall decrease of wave energy within the port basins;
• the Coaling Island reclamation provides a significant amount of sheltering of wave energy
for the entrances of the Cormorant Camber Marina and the Queensway Quay Marina;
• a small increase of wave energy (up to +0.15 m for extreme conditions) is expected only
along a section of the inner side of the detached port breakwater due to reflection of wave
energy off the vertical caisson head wall of the Coaling Island reclamation.

3 Hydrodynamic and flushing conditions
3.1 Introduction
A Delft3D – FLOW model has been setup for the project area to meet the study objectives of
Task 3, 4 and 5 as described in the Introduction (Chapter 1). Deltares has carried out studies
in this area before (WL | Delft Hydraulics, 2006a) which means that we can make efficient use
of available models and only need to update the schematizations and implement a number of
project-specific adjustments. The model will be setup in 3D mode, to allow density driven
currents relevant for the water exchange of the local basins. It will be forced with tide and wind
conditions. Since project-specific hydrodynamic measurements were not carried out, the model
is not further validated as part of the present study. However, we have good confidence in the
performance of the model since it has been successfully validated against ADCP
measurements in previous studies (WL | Delft Hydraulics, 2006a).
This Chapter reports the setup of the Delft3D - FLOW model and presents the local
hydrodynamics and flushing conditions of the present-day situation (without reclamation) and
the effect of the future situation (with reclamation). The Delft3D – FLOW model as setup in this
task will also be used as a basis for the sediment plumes dispersion assessment (Chapter 4),
after being extended with the sediment module of Delft3D to be able to implement sediment
spills in the model and compute the dispersion of sediment plumes due to dredging activities.
3.2 Hydrodynamic model setup
3.2.1 Computational grid

Figure 3.1 presents the series of Delft3D - FLOW curvilinear model grids used for flow modelling
in this study. The ‘overall’ model grid (shown in red in the left frame and taken from a previous
study carried out by Deltares) stretches about 50km in the alongshore direction, covers the
entire Bay of Gibraltar and the adjacent coastline. It has a resolution in the order of about 300m
in the offshore areas down to about 40m in the project area. This model is mainly used to
accurately implement the Mediterranean tide.
Two smaller scale grids have been specifically developed for this project. The right frame of
Figure 3.1 presents a zoomed-in overview, focussing on the ‘intermediate’ model grid (shown
in blue). This model grid is used to propagate the tide within the Bay of Gibraltar to the project
area. The boundary conditions of the intermediate model grid are derived from the overall
model grid. The resolution of this intermediate grid ranges from 100m in the offshore areas to
about 10-20m around the project area. The ‘detailed’ model grid (shown in turquoise in the right
panel of Figure 3.1 and in Figure 3.2) is online coupled to the intermediate grid using the so-
called domain-decomposition technique. This means that these model domains run in parallel
and no separate boundary conditions need to be prescribed for the detailed domain. The
detailed grid covers the area of the envisaged reclamation area, including the adjacent harbour
and marina basins north, east and south of the reclamation. The detailed grid has a resolution
ranging between 5 and 10m. To take into account possible variations of the current velocity
over the vertical, which is relevant for both the flushing and sediment dispersion, the
intermediate and detailed models were run in three-dimensional (3D) mode. In the 3D
hydrodynamic model, the σ-layer approach is used for the vertical schematization. This means
that the water column is divided into the same number of layers independent of the water depth,
each covering with a certain percentage of the total water depth. In that way the absolute
vertical resolution increases automatically in shallow areas and scales along with water level

variations in both time and space. The model includes 10 vertical layers, which is considered
appropriate for this study based on extensive experience with this type of modelling. The layers
are divided uniformly over the vertical; hence, each layer has a thickness of 10% of the local
water depth. A uniform layer division is justified here as the vertical gradients of the modelled
processes are not expected to be high.
Figure 3.1 Overview of the Delft3D - FLOW curvilinear model grids used in this study. Left frame: overall domain
(in red), intermediate domain (in blue), detailed domain (in turquoise). Right frame: zoom-in on the
intermediate grid (in blue) and the detailed domain (in turquoise).
Figure 3.2 Overview of the Delft3D - FLOW curvilinear model grids used in this study, zoomed in on the detailed
domain (in turquoise).

3.2.2 Bathymetry and other hydraulic-modelling features

The bathymetry schematizations of the three Delft3D FLOW hydrodynamics model domains
discussed in the previous subsection are based on the following data sources (with increasing
level of detail):
1. the GEBCO 2019 dataset (General Bathymetric Chart of the Oceans9) with a
resolution of 15 arcseconds (~460 m near Gibraltar, Figure A.1 in Appendix A.1),
2. digitized navigational charts with varying resolution (10-250 m, Figure A.2 in
Appendix A.1),
3. in-house depth schematization of inner port waters of the most recent wave model
(Deltares, 2015), and
4. isolines of a recent local depth survey provided by Lievense (Figure A.3 in
Appendix A.1).
The adopted schematization procedure has been the same as the bathymetry schematization
for the setup of the wave models, reported in Section 2.3.3. Effectively this means that the
bathymetries of both the hydrodynamics and wave models are similar, although locally the grid
setup and resolution of the wave and hydrodynamics model can be slightly different. The model
bathymetry of the present-day situation is presented in the left frame of Figure 3.3, the future
situation (including the reclamation) in the right frame. The reclamation has been schematized
using design drawings provided by the client and has been modelled as dry cells within the
computational grid. Also note the presence of the spur at the entrance of the Cormorant Camber
Marina and Queensway Quay Marina in the present-day situation. In agreement with the client,
the spur is removed from the scenario with the full reclamation in place, see the right frame of
Figure 3.3. The present-day situation and the future situation respectively presented left and
right in Figure 3.3 are considered in the modelling of hydrodynamics and flushing.
Apart from the schematization of the seabed bathymetry, some other relevant hydraulic
features are implemented in the model. The outer parameter of the Small Boats Marina (SBM),
as well as some other dry features, are schematized by dry grid points in the model, see dark
red coloured grid cells in Figure 3.4. Dry cells points remain permanently dry during a
computation and therewith are not part of the hydrodynamic calculations. In order to stimulate
the flushing, the SBM is connected to the open sea by 20 culverts: 16 on the west side and 4
on the south side of the SBM perimeter. The culverts are schematized in the numerical model
as open gaps in between an array of dry grid points representing the overall perimeter structure.
Accurate implementation of the culverts has been optimized by applying the so-called ‘porous
plates’ formulations. This is a technical model feature that is used to control the modelled
discharges through the culvert by means of a calibration factor, based on the locally expected
discharges computed by a detailed culvert model from an earlier project by Deltares.
In the future situation with the completed reclamation, the culverts at the southern side are no
longer in function and they are therefore not implemented in the model schematisation for that
situation. Only the culverts along the western section of the perimeter of the SBM remain.
The offshore breakwater (partly visible on the left side of Figure 3.4) is schematized as a ‘thin
dam’ in the intermediate domain of the model. Thin dams are infinitely thin objects along a
calculation grid line prohibiting the flow exchange between the computational cells at either
side of the dam. This approach is often applied if the size of the object is smaller than the local
resolution of the grid. Please note that the SBM perimeter, as described above, has been
9
https://www.gebco.net

represented by dry cells, since the higher resolution of the model in that area allows individual
cells to represent the width/extent of the structures here.
Figure 3.3 Model bathymetry schematization of the project area in the present-day situation (left frame) and the
future situation, including the reclamation. Bed level values are given with respect to Gibraltar Ordnance
Datum (GOD).
Figure 3.4 Overview of the model grid and bathymetry, including some technical model schematizations of the
SBM protection with culvert and the offshore detached breakwater. The 4 culverts at the southern side have
been schematized as 1 culvert, since they are located close together, but represent the discharge of all 4.
The floating mooring infrastructure (jetties) within the SBM is implemented as a floating
structure in the model. Effectively this means that the wind has no contact to the water surface

in this area representing the high density of jetties and moored boats in the SBM. Sensitivity
runs have not been executed to test what the effect is of the wind on the hydrodynamics within
the SBM, since it is expected to be minor and because it is not the focus of the present study.
3.2.3 Boundary conditions and forcing

The boundary conditions of the intermediate model have been derived from the overall model.
The boundary conditions of the overall model are a combination of astronomic water level and
current velocity boundaries, and therefore flexible with respect to timing. The boundary
conditions at the southern side of the intermediate model are defined as timeseries of water
levels, the conditions at the western boundary are defined as current velocities. It is common
practise and accepted within the modelling community to apply two different types of boundary
conditions on one model to achieve the highest computational stability and reliable results.
Applying different types of boundary conditions (either water level or local current) leaves the
model free to find its own solution for the other parameter locally, which you do not allow if you
force the model with a single parameter on all boundaries, often leading to instable results.
In addition to the tidal forcing, the model is also forced with wind conditions representative for
the area. Six representative wind conditions have been selected from a previously executed
modelling study by Deltares (WL | Delft Hydraulics, 2006a) and are imposed uniformly over
the model domains and constant in time (the wind condition has no influence in some parts of
the SBM, as described in Section in 3.2.2). An overview of the wind conditions is presented in
Table 3.1.
Table 3.1 Overview of modelled wind conditions

Wind condition # Wind direction Wind speed (m/s)
1 West 5.7
2 West South-West 9.3
3 East 4.5
4 East 9.5
5 North 3.6
6 No wind No wind
3.2.4 Model sensitivity analysis

Some sensitivity analyses on the model grid resolution, the computational timestep, the
diffusivity/viscosity and the loss coefficient of the porous plates (representing the culverts) have
been executed to arrive at the most stable, efficient and reliable model setup. The results of
these sensitivity analyses are not further described in the report since these are standard model
definition procedures and only the final, optimal modelling approach is presented here.
3.3 Model scenarios

The hydrodynamic simulations, also used for the flushing assessment with inclusion of an initial
tracer, have been carried out for the tidal period 1 June 2019 to 1 July 2019, covering
approximately 2 spring-neap tidal cycles and providing enough time to compute the flushing
(residence) times. Figure 3.5 presents the timeseries of the local water levels for the entire
duration of the simulations. The simulations included a spin-up time of 6 days in advance to 1
June, to make sure that the hydrodynamics are stabilized prior to the start of the flushing
computations.
In total 12 simulations have been carried out: 6 wind conditions for 2 situations: the present-
day situation (without reclamation, see the left frame of Figure 3.3) and the future situation (with

the full reclamation, see the right frame of Figure 3.3). A complete overview of the model
scenarios is presented in Table 3.2.
Table 3.2 Overview of the model scenarios for the hydrodynamics and flushing
Simulation # Situation Wind condition
1 West 5.7 m/s
2 West South-West 9.3 m/s
3 Present-day (no reclamation) East 4.5 m/s
4 East 9.5 m/s
5 North 3.6 m/s
6 No wind No wind
7 West 5.7 m/s
8 West South-West 9.3 m/s
9 Future situation East 4.5 m/s
10 (reclamation in place) East 9.5 m/s
11 North 3.6 m/s
12 No wind No wind
3.4 Impact of reclamation on hydrodynamics

Figure 3.5 presents the modelled water levels of the simulation period (1 June – 1 July 2019)
for wind condition 1 (see Table 3.1) for a location close to the envisaged reclamation, see red
marker in the right frame map. The blue line presents the water levels for the present-day
situation and the red line the water levels for the future situation (with reclamation). The figure
shows that the local vertical tidal range during spring tide is up to 1m and decreases to
approximately 0.4m during neap tide. There is no discernible impact of the reclamation on the
local water levels during normal wind conditions.
Figure 3.5 Timeseries plot of the water levels on a location just west of the envisaged reclamation for the present-
day situation (in blue) and the future situation (with reclamation, in red). The red marker on the right map
plot indicates the location.
Figure 3.6 shows the depth averaged velocities during spring-tide ebb (5 June 2019, 22:00h)
for wind condition 6 (no wind, see also Table 3.1) in the present-day situation (top left frame)
and the future situation (i.e. including reclamation, top right frame). The bottom frame shows
the difference of the two situations. It is observed that the present-day hydrodynamics are

generally low, with a slight increase just west of the SBM due to flow contraction between the
SBM and the further offshore located breakwater (located outside of the figure extents).
Figure 3.7 shows typical absolute depth-averaged flow velocities for the same time instance,
but including wind condition 1 (Table 3.1). The westerly wind of wind condition 1 has a slightly
decreasing effect on the southward directed flow velocity, compared to the situation without
wind (Figure 3.6). The model results show flow velocities up to about 3cm/s. The effect of the
completed reclamation (right frame) is present and the extent of the influenced surface area is
limited, the absolute increase is smaller than 1cm/s for the largest part. This is an insignificant
difference in relation to the accuracy level of models that can typically be expected in practical
situations like this. A slight decrease of flow velocities is observed north of the reclamation area.
The decrease of flow velocities can be explained by the sheltering effect of the reclamation.
Figure 3.8 present results in the same format and for the same tidal phase but for the situation
with wind condition 3 (see Table 3.1). For this condition flow velocities in the same order of
magnitude are modelled as compared to the situation without wind. This means that the effect
of this wind condition on the current velocities is limited. Also the areas of local increase and
decrease of flow velocities due to the reclamation are very similar of magnitude and absolute
value (shown by the difference plot in the bottom frame).
For each wind condition (Figure 3.6 to Figure 3.8) it can be observed that there is a local
increase of flow velocities in the entrance of the Cormorant Camber Marina and Queensway
Quay Marina. This is a combined effect of the removal of the spur, allowing more free flow in
the new situation, which means an increase of flow velocities. Also the reclamation causes a
local narrowing of the entrance, which also results in a local increase of velocities. For this time
instance there are no effects visible at the location of the SBM entrance.
It is noted that for this analysis one specific time instance during spring-tide has been selected,
(with relatively high flow velocities) to give an overall indication of the local tide and wind-driven
hydrodynamics that drive the flushing and sediment plume dispersion. This means that the
analysis described applies to this time instance only and the flow velocity patterns could look
different for other time instances. Nevertheless, the hydrodynamics generally will be very mild,
with or without reclamation, and will decrease even further for neap-tide conditions relative to
the spring-tide conditions considered here in detail.
This section presents and analyses results for wind conditions 1, 3 and 6 only. The results for
all the wind conditions are available in the same format in Appendix C, in respective order of
the wind conditions listed in Table 3.1.

Figure 3.6 Typical flow velocities during a spring tide for wind condition 6 (no wind). Top left: absolute flow
velocities in the present-day situation (no reclamation). Top right: flow velocities in the future situation (with
reclamation). Bottom: flow velocity difference plot (warm colours: increase of velocities with respect to
present-day; cold colours: decrease of velocities with respect to present-day). Blue vectors: present-day
situation, red vectors: future situation (with reclamation).

Figure 3.7 As Figure 3.6, now for wind condition 1.

Figure 3.8 As Figure 3.6, now for wind condition 3.
3.5 Impact of reclamation on flushing times

This section reports the results of the flushing simulations that have been carried out to
determine the effect of the reclamation on the flushing time in the area. To determine the
relative impact of the reclamation on the flushing times (sometimes also referred to as

‘residence times’), simulations have been carried out for the present-day situation and the
future situation, including the completed reclamation.
The flushing times are computed by means of the dilution rate of a numerical tracer in the
hydrodynamic model simulations, of which the setup is described in detail in Section 3.2. This
tracer is added in the project area (see Figure 3.9) and is dispersed by the currents (advection
and diffusion). The figure shows that the initial tracer field is applied in the entire project area
including the SBM and the two existing harbour basins south of the reclamation area. This
allows for a relative comparison of flushing times of different sub-sections of the area. The time
it takes for the tracer to drop below a certain threshold concentration is defined as the flushing
time. The threshold level is set on 37% (T37%) of the initial concentration, which is a typically
used threshold for this type of studies. Appendix B explains the theory of the flushing calculation
in more detail.
Figure 3.9 Schematization of the initial tracer field (red patch) used for the flushing simulations.
The computed residence times form a suitable basis for a relative comparison of the flushing
capacity for the present-day situation and future situation with reclamation, and to indicate
areas of poor flushing11.
The results of the flushing simulations are presented here for three different ambient conditions:
11
It is noted that the flushing rate of a water body is not conclusive for the expected water quality of this water body. The
water quality depends on the combination of flushing and the loadings (e.g. untreated sewer water discharges) on a
water body and can only be quantitatively assessed in an integral water quality assessment taking all loads into
account.

• Figure 3.10: No wind (Condition 6 in Table 3.1)

• Figure 3.11: Westerly wind (Condition 1 in Table 3.1)
• Figure 3.12: Easterly wind (Condition 3 in Table 3.1)
The presentation format is the same as for the flow velocity results presented in the previous
section. The top left frames present the flushing times for the present-day situation (without
reclamation), the top right frames present the flushing times for the future situation (with the
completed reclamation). The bottom frames present the differences between the flushing times
between the future situation and the present-day situation. The results of all conditions,
including the remaining wind conditions 2, 4 and 5 are presented in Appendix C.
Generally, it is observed that there are no large spatial differences in the computed flushing
times. The SBM shows to have the longest flushing time, in the order of 6-7 days. Even though
the SBM has several culverts in addition to the entrance at the north-east corner, the connection
of the SBM basin to the open sea is still limited, partly because of the sheltered entrance, hence
it takes longer to flush. The Cormorant Camber Marina and Queensway Quay Marina have
shorter flushing times, in the order of 5 days, because the entrance is more exposed to the
open sea.
The impact of the reclamation on the flushing times relative to the present-day situation is
limited, as indicated by the top right frames and bottom frames. Within the SBM basin, the
impact of the reclamation on the flushing times is shown to be neglectable. This means that
according to the modelling the closure on the culverts at the southern side of the SMB hardly
has an effect. This indicates that also in the present-day situation they do not contribute much
to the flushing of the SBM, because of limited head differences over the culvert structures. It is
observed that the flushing times in the Cormorant Camber Marina and Queensway Quay
Marina increases by approximately 1 day. This can be explained by the sheltering effect of the
reclamation on the flow velocities in the entrances to those basins. This is in line with the results
of the hydrodynamics presented in the previous sub-section, showing a decrease of the flow
velocities through the connection just south of the reclamation, although the absolute
differences are small. West of the reclamation, the flushing times slightly decrease. This is also
in line with the hydrodynamics, showing an increase of flow velocities due to flow contraction,
which has a reducing effect on the flushing times.
The impact of different wind conditions on the flushing times turns out to be low. The overall
patterns are the same and the observations described above apply practically speaking to all
the wind conditions.
It is concluded that the present-day flushing times in the area are in the order of a couple of
days and that the impact of the reclamation is very local and small. As stated, the flushing times
are only an indicator for the water quality and that also other influences should be considered
for a full water quality assessment.

Figure 3.10 Flushing time plots for wind condition 6 (no wind). Top left: flushing time in present-day situation
(without reclamation). Top right: flushing time in future situation (with reclamation). Bottom: flushing time
difference plot (warm colours indicate increased flushing time; cold colours indicate decreased flushing
time).

Figure 3.11 As Figure 3.10, now for wind condition 1 (westerly wind).

Figure 3.12 As Figure 3.10, now for wind condition 3 (easterly wind).


In this chapter, the hydrodynamic and flushing conditions in the project area have been
assessed, for both the present-day situation and the future situation with the reclamation in
place, under influence of the tide and several representative wind conditions. The following
conclusions are drawn from the results presented in this chapter:
• The flow velocities in the project area are mainly determined by the combination of the
tide and the wind.
• Under normal wind and spring tide conditions, the present-day flow velocities in the
project area are weak, in the order of 5cm/s. During neap tide the velocities are close
to zero.
• The reclamation has no discernible impact on the local water levels during normal wind
conditions.
• The reclamation has a slight influence relative to the present-day flow velocities,
showing an increase at the westernmost tip due flow contraction and a decrease at the
northern and southern side of the reclamation due to the sheltering effect of the
reclamation.
• Flushing times have been considered in this study as a proxy for water quality. The
computed flushing times in the present-day situation show that there are no large spatial
differences. Of the basins considered, the SBM shows to have the highest flushing
times in the order of 6-7 days. The Cormorant Camber Marina and Queensway Quay
Marina have slightly shorter flushing times in the order of 5 days, because they are
more exposed to the open sea.
• The computations show that the impact of the reclamation on the flushing times is
limited. There is no impact observed of the flushing times within the SBM, even though
the culverts at the southern side are closed off by the reclamation. This indicates the
limited effectiveness of those culverts already in the present-day situation. There is a
noticeable increase of flushing times of the Cormorant Camber Marina and Queensway
Quay Marina, due to the sheltering effect of the reclamation, but the absolute
differences are small (in the order of 1 day).
• The computational results indicate that the influence of different wind conditions on the
flushing times is limited. For each of the analysed normal wind conditions (no wind,
westerly wind, easterly wind) the patterns and order of magnitude flushing times are
rather similar.

4 Sediment plume dispersion
4.1 Introduction
In order to construct the Gibraltar Harbour coaling island, first the outer perimeter (also called
the bunds) of the reclamation will be constructed, after which the contained area will be filled.
Spillage and dispersion of sediment into the environment will only happen during the
construction of the outer perimeter. The potential effect of the dredging related sediment spill
on the environment, will be assessed in this chapter.
For the modelling of the far-field dispersion of the fine sediment dredging spills associated with
the construction of the bunds, the same detailed three-dimensional (3D) Delft3D-FLOW model
was applied as for the flushing task (Chapter 3). With this model, the flow conditions around
the reclamation were simulated including the dispersion of the spilled fine sediments due to the
construction dredging operations for different ambient conditions.
Three different dredge scenarios are considered in the dredge plume modelling. The locations
of the operations and associated timings for each of the three scenarios have been provided
by the Client. Deltares proposed the implementation of the sediment spill terms in the model,
partly based on the analysis of local bed sediment samples that have been collected by
Sergeyco Andalucia S.L. in 200912 and provided by the Client. The schematization of the spill
terms has been discussed with and agreed by the Client in earlier stages of the study.
4.2 Model approach
4.2.1 Model starting points

The hydrodynamic model as described in Chapter 3 is also used in the present sediment plume
dispersion task. This model includes the forcing of the tide and 6 representative wind forcing
conditions as described in the previous chapter.
The sediment dispersion modelling is carried out as first-order, excess sediment dispersion
modelling. This means that for the sediment dispersion modelling no background
concentrations are included in the model, but only the dispersion of the sediments spilled by
the dredging operations. This allows for analysis of the influence of the activities on the local
sediment concentrations separately, without an a-priory unknown and variable background
concentration value. The resulting suspended sediment concentrations will need to be added
to the (expected) background concentrations when assessing the total sediment concentrations
or turbidity in the project area. The numerical modelling computes suspended sediment
concentrations (i.e. in mg/L).
Resuspension of deposited material from the spills by waves is not considered in this study.
The locations of the dredge operations are sheltered from wave action by the offshore
(detached) breakwaters so very little wave energy is expected to reach the locations of the
dredging operations.
The simulations are carried out for different durations, covering the duration of the considered
dredging operations (considered further in the subsequent sections), including some extra time
12
Grading curves Sergeyco 2009.pdf

for the plume evolution after the operation has finished (e.g. the settling of the suspended
sediment).
4.2.2 Schematization of sediment spill terms

The dredge spills are modelled with 3 fine sediment fractions (<63μm), i.e. 10 μm, 20 μm and
50 μm with associated fall velocities, computed with Van Rijn (1993). These fractions are
assumed to represent 52%, 21% and 27% of the spill composition, derived from the sediment
grading analysis performed by Sergeyco, see also Figure 4.1. The sieving curve shows that
48% of the sediment sample passed the sieve and thus has a particle diameter of 80 μm13 or
less. The normalized distribution of fine sediment fractions in the dredge spill has been
determined by linearly extrapolating the grading analysis curve into the fine sediment segment,
indicated by the blue dashed line in Figure 4.1. It is noted that this is the most conservative
approach in terms of fine sediment distribution (i.e. the least favourable distribution, as a safe
modelling assumption). In order to cover the full range of fine sediment fractions it is assumed
that each of the three considered fine sediment fractions (10, 20, 50μm) represents a range of
fractions to make sure that the modelled spill fully represents the distribution of fine sediments
as derived from the sediment grading analysis. The 10 μm fraction represents the range
between 0 – 15 μm, the 20 μm fraction the range between 15 and 35 μm and the 50 μm fraction
the range between 35 and 80 μm. The percentual distribution of each of these ranges can be
derived from the grading analysis curve and is shown by the grey boxes in Figure 4.1. The
vertical dimension of these boxes represents the percentual distribution of each sediment class,
summing up to the total 48% of fine sediment fractions present in the analysed sediment
sample. Subsequently these percentages have been normalized to arrive at distributions to be
used for the schematization of the sediment spill in the numerical model. The spill rates as a
result of the dredging operation with the backhoe were agreed with the client to be 4 kg/s for
this study based on a separate assessment carried out by Deltares as part of this study (see
Appendix D), including three fine sediment fractions. This (conservatively) assumes that the
dredging for the outer perimeter structure will be carried out using a backhoe equipped with an
open bucket. An overview of the characteristics and the model schematization of the three
sediment fractions is presented in Table 4.1.
Figure 4.1 Analysis of the proportional presence of several fine sediment fractions (10, 20 and 50μm) in the model
schematization of the dredge spill term, based on the sediment grading analysis performed by Sergeyco.
13
We are aware that this is a crude approximation, but no other data were available. This assumption has been agreed
upon with the Client.

Table 4.1 Implementation of sediment fractions and associated sediment source terms applied in the numerical
model
Fraction Particle size, dp Fall velocity Fraction of total Spill rate [kg/s]
[μm] [mm/s] fines [%]
1 10 0.08 25 / 48 = 52% 0.52 * 4 = 2.08
2 20 0.33 10 / 48 = 21% 0.21 * 4 = 0.48
3 50 2.10 13 / 48 = 27% 0.27 * 4 = 1.04
4.2.3 Dredge scenarios

In total three dredging scenarios, representing different dredging areas and timings are
considered in the modelling, in agreement with the client. The characteristics of the three
Scenarios 1, 2 and 4 (following the numbering of the Client) are explained in this sub-section.
It is noted that each dredging scenario is simulated for the 6 wind conditions as described in
Section 3.2.3.
Scenario 1
Figure 4.2 shows the Scenario 1 dredge area (see the map figure in the top frame, black dashed
line), located south of the envisaged reclamation area and just north of the entrance of two
existing harbour basins. The dredge area has a total dimension of approximately 100m by
almost 50m and is subdivided in 6 sub-areas (boxes). In this scenario the reclamation is not
yet finalized, but in an intermediate state containing 100.000m3 of sediment (compare the
intermediate state of the reclamation indicated by the bright yellow area in Figure 4.2 with the
final state in the right frame of Figure 3.3). During this construction phase, a silt screen is placed
around the reclamation to block the sediment getting in suspension as a result of the
reclamation activities (i.e. the destination of the dredged sediment). In the modelling we assume
a 100% blockage of sediment dispersion from the reclamation area itself, so the only sediment
spill is as a result of the dredging activities carried out by the backhoe. The bottom frame of
Figure 4.2 shows the time sequence of the backhoe dredging operations in the different boxes
within the dredge area and the associated sediment spill rates (vertical axis). The bottom panel
shows that in this scenario the total operation takes about 2 days (48 hours) and moves in
sequential order from box 1 to box 6. The operations in respectively box 3 and 6 are suspended
for 12 hours (reflecting the working hours provided by the Client), reflected by the gaps in the
bar diagram. The underlying information provided by the Client is included in Appendix E.1.

Figure 4.2 Top frame: Scenario 1 dredge area (black dashed line), subdivided in 6 boxes. The bright yellow area
south of the SBM shows the assumed reclamation state during this dredge Scenario. Bottom frame: bar
diagram showing time sequence of the dredge operation, colour coded for the different boxes.
Scenario 2
Figure 4.3 presents the characteristics for dredge Scenario 2 in a similar manner as for
Scenario 1. The top frame shows the location of the dredge area (black dashed line), where
the northern end of the reclamation connects to the southern end of the SBM. This area has a
total dimension of approximately 170 x 30 m and is subdivided in 10 sub-areas (boxes). The
reclamation is in the same intermediate state as for Scenario 1, containing 100.000 m3. Also in
this case the reclamation is surrounded by a silt screen (implemented in the model by applying
a thin dam, but not visible in the figure), which means that there is only the sediment spill as a
result of the dredging activities with the backhoe. The bottom frame shows the sequence of
dredging activity in time, taking 3 days and 6 hours (78 hours). Also, in this scenario the
operations are suspended three times, when the operation is in respectively box 3, 6 and 8.
The underlying information provided by the Client is included in Appendix E.1.

Figure 4.3 Top frame: Scenario 2 dredge area (black dashed line), subdivided in 10 boxes. The bright yellow area
south of the SBM shows the assumed reclamation state during this dredge Scenario. Bottom frame: bar
diagram showing time sequence of the dredge operation, colour coded for the different boxes.
Scenario 4
Scenario 4 is the scenario that takes by far the most time and is executed over the largest area.
The top frame of Figure 4.4 shows the dredge area associated with Scenario 4, running from
the northern end of the reclamation in a curve towards the southern end. The area has a
dimension of 540 x 60 m and is subdivided in 54 different boxes. The reclamation is in an
advanced state compared to Scenarios 1 and 2, now also including the bunds at the northern
and southern end (following completion of the work described in Scenario 1 and 2). Also in this
scenario it is assumed that there is no suspension of sediment due to reclamation works as a
result of applying a silt screen. Therefore, only the spill term of the dredging activity remains.
The bottom frame of Figure 4.4 shows the time sequence of dredging operation, taking more
than 20 days (490 hours) in total. Due to the relatively large number of boxes in comparison to
Scenario 1 and 2, there is no legend included but they are colour coded in sequential order
from blue to red. The figure shows that there are 12 hour breaks each day. The underlying
information provided by the Client is included in Appendix E.1.

Figure 4.4 Top frame: Scenario 4 dredge area (black dashed line), subdivided in 54 boxes. Bottom frame: bar
diagram showing time sequence of the dredge operation, colour coded for the different boxes (in sequential
order from blue to red, legend is lacking due to the large number of boxes).
4.2.4 Evaluation criteria

There are no general criteria for evaluating the impact of dredge plumes. However, some areas
in the vicinity of the dredge locations are identified as ‘sensitive’ and will be looked at in more
detail in the following section. Timeseries plots of suspended sediment concentrations will be
presented near these locations to give an indication of the order of magnitude of suspended
sediment concentrations that can be expected there.

4.3 Results of the sediment plume dispersion
4.3.1 General behaviour of sediment plume dispersion

Due to the very weak hydrodynamics in the project area, the plume dispersion is mainly
governed by diffusive processes, spreading out in radial direction from the locations of the
dredge operations. Local flow velocities are in the order of a few cm/s and do not contribute
significantly to the dispersion of spilled sediment.
Since the sediment particles have a certain fall velocity, the suspended sediment concentration
does not only decrease because of dispersion processes, but also due to settling of the
sediments. Although the results have been presented as total (summed) suspended sediment
concentrations, 3 different sediment fractions were considered as part of the model
computations in each situation, each fraction with its own corresponding fall velocity. The
coarse fractions (e.g. 50μm) will settle closer to the dredging source than the smaller fractions
(e.g. 10μm). This means that further away from the source the composition of the sediment
plume typically consists more of fractions of smaller sediment than closer to the spill source.
The model allows for resuspension of deposited sediment under influence of flow velocities,
but flow velocities are expected to be too weak to resuspend deposited sediment in the situation
as considered.
4.3.2 Computed sediment plume dispersion

The computed sediment plume dispersion is presented as maps of time-averaged and
maximum depth-averaged SSC maps. For the maximum SSC maps, the 99 th concentration
percentiles were presented instead of the actual computed maximum concentrations, to
remove any possible numerical noise from the model computations. The presented 99th
percentile concentrations mean that 99% of the time, concentrations are lower than the
presented concentrations.
For readability, only a selection of wind condition scenarios is presented here, i.e. a normal
westerly and easterly wind condition (respectively wind condition 1 and 3 in Table 3.1). These
particular conditions were selected because the wind direction is expected to give the largest
difference in plume dispersion (if any). The maps are presented separately for dredge
Scenarios 1, 2 and 4. The results for the relatively short Scenario 1 and 2 are presented for a
spring tide situation, which is considered to be governing and resulting in the largest plume
extent. Scenario 4 is long enough to cover a full spring-neap cycle spring and therefore the
spring and neap tide is not separately considered. All remaining results are presented in
Appendix E.2.
The sediment plume dispersion for dredging activities associated to Scenario 1 are presented
in Figure 4.5 (mean sediment concentration) and Figure 4.6 (99th percentile). The left frame
presents the sediment concentration for the westerly wind conditions, the right frame for the
easterly wind condition. The figures show that the plume remains fairly close to the location of
the dredging activities and that the wind direction has a limited influence on the dispersion. The
spreading of the sediment is uniform in radial direction from the location of the sediment source,
mainly driven by diffusive processes. This is in line with the weak hydrodynamics in this area
(Chapter 3). In more dynamic environments with higher flow velocities, dredge sediment
plumes tend to have a more elongated shape in the orientation of the main flow direction, which
is not the case here. It is also observed that the sediment plume penetrates somewhat into the
Cormorant Camber Marina and Queensway Quay Marina, although in low concentrations with
maxima not exceeding 50 mg/L.

Figure 4.5 Sediment plume dispersion for dredge Scenario 1, mean depth-averaged concentration. Left frame:
westerly wind, right frame: easterly wind.
Figure 4.6 Sediment plume dispersion for dredge Scenario 1, maximum (99 th percentile) depth-averaged
concentration. Left frame: westerly wind, right frame: easterly wind.

easterly wind condition. Also for this dredge scenario with slightly longer duration than the
previous Scenario 1, the plume dispersion is limited and the direction of the wind does not lead
to significant differences in the shape of the plume. The plots with mean depth-averaged
concentrations show a slightly higher influence of the westerly wind compared to the easterly
wind. This is in line with hydrodynamics (Chapter 3), showing slightly higher flow velocities for
conditions with a westerly wind, resulting in a higher dispersion rate. In contrast to the dredging
activities of Scenario 1, the plots do not show any suspended sediment penetrating into the
Cormorant Camber Marina and Queensway Quay Marina. The plots with maximum
concentrations clearly show the signature of the boxed subdivision of the dredge area, which
is in line with the way how the dredge activities have been implemented in the model: moving
from box-to-box, but on fixed locations within the boxes. This results in local, relatively high
concentrations. In practice the generated plumes may be smoother, because the dredge
activities will not be as static as implemented in the modelling.
easterly wind condition. The main difference of Scenario 4 in comparison to Scenario’s 1 and
2 is the substantially longer duration and larger area of the dredge activities, which is reflected
in the results of the sediment dispersion. The extent of the sediment plume is larger, but also
in this case stays relatively close the dredge area. The mean concentration reaches values of
up to 50 mg/L and the maxima reach up to about 200 mg/L. Also for Scenario 4, the dredge
locations are well reflected in the maxima plots, showing concentration peaks on the locations
within each of the 54 boxes. In comparison to the results of Scenario 1, also for Scenario 4
there is some influence of the dredge activities in the Cormorant Camber Marina and
Queensway Quay Marina, showing increased concentrations, but in fairly low values as you
would typically expect for low dynamic areas. The hydrodynamics are mild and therefore do not
have the capacity to disperse the sediment over large distances, so the concentrations decline
in a radial direction from the source.

Figure 4.8 Sediment plume dispersion for dredge Scenario 2, maximum (99th percentile) depth-averaged

Figure 4.10 Sediment plume dispersion for dredge Scenario 4, maximum (99 th percentile) depth-averaged

4.3.3 Suspended sediment concentrations near the sensitive receivers

This sub-section presented the timeseries of local suspended sediment concentrations (one for
each dredge scenario) to be able to quantify the impact that can be expected at different
locations in the project area. Timeseries are presented for the locations indicated in Figure
4.11.
Figure 4.11 Map showing the locations for which sediment concentration timeseries are presented in Figure 4.12,
Figure 4.13, Figure 4.14.
Figure 4.12, Figure 4.13, Figure 4.14 respectively show the timeseries at these locations for
dredge scenarios 1, 2 and 4. From these figures it is observed that highest concentrations (in
the order of 200mg/l) can be expected for dredge Scenario 1 at the Island during dredge
Scenario 1, especially the location that is closest to the entrance of the Cormorant Camber
Marina and Queensway Quay Marina and the dredge operation itself. For dredge Scenario 4,
concentrations of similar magnitude are observed at the end of the operation, when the
backhoe is working close to the entrance of the Cormorant Camber Marina and Queensway
Quay Marina. Within the Cormorant Camber Marina and Queensway Quay Marina,
concentrations do not exceed 50mg/l. Sediment concentrations as a result of all the dredging
operations at the SBM are close to zero. If the duration of the dredge operations would increase,
the maximum concentration values found near the peninsula will probably scale accordingly
(linearly) up until an equilibrium is reached at some point in time. This might apply especially
for Scenario 1, which includes a relatively small dredge area.

Figure 4.12 Timeseries of depth-averaged suspended sediment concentrations due to dredge Scenario 1. Dredging
stops after 2 days.
stops after 3 days and 6 hours.
stops after almost 20 days.
4.3.4 Computed sediment deposition

Figure 4.15, Figure 4.16 and Figure 4.17 show the computed deposition as a result of
respectively the Scenario 1, 2 and 4 dredge activities. Similar to the presentation of the other
results presented in this chapter, it is noted that these figures show the excess sediment
deposition as a result of the dredging activities. The background sediment transport processes
are not considered in the modelling and are therefore not reflected in these figures. Regardless,

due to the mild hydrodynamics and limited amount of wave energy entering the Gibraltar Port
area, the background sediment transport processes are expected to be limited.
The sediment deposition of all scenarios shows a very similar pattern as the dispersion of the
sediment plume itself: the centre of gravity of the deposition patch is located on the dredge
area and the impact remains fairly close. The computed deposition as a result of the dredging
activities of Scenarios 1 and 2 reaches values of up to a decimetre in height close to the
dredging activity and decreases to a height of less than a centimetre about 100m away from
the dredge location. Due to Scenarios 1 and 2, some sediment deposition is expected within
the Cormorant Camber Marina and Queensway Quay Marina, but in quantities not exceeding
1cm in height.
The sediment deposition as a result of the dredging activity of Scenario 4 is computed to be

higher than for Scenario 1 and 2. In general, the cumulative effect of sediment depositing at
the bed is more dominant than the dispersion of suspended sediment in the water column,
because the deposition of sediment in this area is expected to be almost irreversible due to the
low energetic conditions and hardly resuspends. This means that the thickness builds up in
time along with the execution of the dredge operation and shows higher values for Scenario 4
than Scenarios 1 and 2 because the duration of the operation is longer. Still the associated
patch heights are limited and in the order of 1-2% of the local water depth.
Figure 4.15 Excess sediment deposition on the bed as a result of the Scenario 1 dredge activities


This chapter describes an assessment of the sediment plume dispersion and deposition as a
result of the dredge activities to construct the reclamation. In consultation with the Client, three
dredge scenarios have been defined (varying the location and duration of the dredge activity),
with associated sediment spills during the operation. The sediment spills have been

implemented in the model to assess the excess dispersion and deposition of sediment due to
the dredging operations. Hence the background sediment transport processes are not
considered in the model. In case the sediment concentration in the area needs to be assessed
in an integral way, the background concentration would need to be added if the information is
available at all.
The following conclusions are drawn from the results presented in this chapter:
• The dispersion of the dredge sediment plumes is mainly driven by diffusive processes,
resulting in a circular plume shape (uniform dispersion in radial direction from the
source). The hydrodynamics in this area are weak, hence the advective transport is
limited. Maximum values of up to a few hundred mg/L are observed close the location
of the dredging activity, quickly reducing to values below 50 mg/L at about 100-200 m
away.
• The direction of the wind has limited influence on the dispersion of the sediment and
shape of the dredge sediment plume.
• The sediment deposition shows very similar patterns as the dispersion area of the
suspended sediment plume itself. The computed sediment deposition on the bed
reaches values of up to a decimetre in height close to the location of the dredging
activity for Scenario 1 and 2 but drop to values less than 1 cm in height at 100-200 m
away from the dredge sediment source.
• The sediment deposition as a result of Scenario 4 is higher than for Scenarios 1 and 2:
up to a few decimetres in height at the location of the dredge activity. The height of the
patch also quickly drops at distances of 100-200 m away from the dredge activity.
Resuspension of deposited sediment is not expected here due to the low
hydrodynamics in the Gibraltar Port Area.
• According to the computations made, there is hardly an impact of the dredge activities
in the area north of the SBM. The results imply that an impact is also not expected in
case the duration of the dredge activities would be longer. This is mainly explained by
the weak hydrodynamics lacking the capacity to transport the sediment over a distance
more than a few hundred meters.
• There is an impact of the dredge activities on The Island, mainly from Scenario 1 and 4,
because it is located relatively close. However, the associated sediment concentrations
remain low and decrease quickly after termination of the dredging.
4.5 Discussion
The results of the model simulations described in this chapter show that the impact of the
dredging activities is local, quickly decreasing in magnitude at a distance of 100-200 m away
from the dredging activity due to the very low hydrodynamic conditions that lack the capacity to
transport the sediment over larger distances. It is expected that the extent and concentrations
of the sediment plume scale with the duration of the dredge activity, should they be slightly
longer than assumed in this study. Note that for much longer dredging activities, an equilibrium
may be reached in both extent and concentration. This also counts for the extent of the
sediment deposition area, however the thickness of the sediment is expected to scale linearly
with the duration of the dredging, as long as the operations are executed within a relatively
small area.
The practical implication of sediment deposition locally around the dredge activity is probably
a somewhat inefficient dredge operation since the spilled and deposited sediment needs to be
dredged again. Nevertheless, the sediment infill rates are much lower than the dredging rate,
so that will likely only have a limited impact on the planning of the dredging operations.

Mitigating measures could be taken if any impact on the Cormorant Camber Marina and
Queensway Quay Marina or the peninsula are to be avoided. Adopting time windows with flow
velocities leading away from those areas is not expected to be very effective, since the
advective transport is very low. Implementation of physical measures will probably be more
effective, such as applying a silt screen around the dredging activities.
Unless dredge operations will also take place further northward of the dredging areas
considered so far, no impact of dredging activities is expected on the area north of the SBM
because the hydrodynamics as they are modelled in this study are probably too weak to
transport it there.

5 Key findings
In this study the (potential) impact of the planned Coaling Island reclamation on the local wave
climate, the hydrodynamics in the harbour, the flushing of the port basins and the ecology
during construction (sediment dispersion) has been considered. Two situations have been
considered, being: (1) the present-day situation and (2) the situation including the Coaling
Island reclamation. The model results show that:
Wave conditions
• the easterly extreme conditions, EAST95 or EAST90, generally result in the severest
wave conditions in the Gibraltar Port waters;
• with the new land reclamation in place, wave energy entering through the southern port
entrance is partially dissipated on the rubble mound slopes, leading to an overall decrease
of wave energy within the port basins;
• only in a limited area along the inner side of the detached port breakwater, a small
increase of wave energy is expected due to reflection of wave energy off the planned
vertical caisson head wall of the Coaling Island reclamation;
• the Coaling Island reclamation provides a significant amount of sheltering of wave energy
for the entrances of the Cormorant Camber Marina and the Queensway Quay Marina,
leading to milder wave conditions in those areas.
Flow conditions
• the current in the project area is mainly driven by the tide and wind and reaches velocities
in the order of 5cm/s under normal wind conditions. There is a minimal impact of the
reclamation on the current velocities, showing local increase and decreases in the order
of 1cm/s, although such a difference may fall within the practical accuracy of the numerical
model.
• The reclamation has no discernible impact on the local water levels during normal wind
conditions
Flushing
• the flushing times in the project area are relatively short. The Small Boats Marina shows
the highest flushing times in the order of 6-7 days, mainly due to the limited connection to
the open sea. The two southern basins have slightly lower flushing times in the order of 5
days;
• the reclamation has a limited impact on the flushing times. The flushing times in the SBM
are not impacted and increase in the order of 1 day in the southern basins;
• the direction of normal wind conditions has no effect on the flushing times of the project
area.
Sediment dispersion
• the impact of the backhoe dredging activities is expected to stay relatively close to the
location of the dredge activity. Maximum sediment concentrations reach up to a few
hundred mg/L, but quickly decrease to typical background values 100-200 m away;
• the sediment deposition as a result of the backhoe dredging is in the order of a few
centimetres for the short dredging operations (Scenario 1 and 2) and reaches up to a few
decimetres for the long dredging operation (Scenario 4). However, in line with the extent
of the suspended sediment plumes, the impact remains close to the location of the dredge
activity.

References
Battjes, J.A. and H.W. Groenendijk, 2000: Wave height distributions on shallow foreshores,
Coastal Engineering. 40, 161-182.
Booij, N., Ris, R. C., and L. H. Holthuijsen, 1999: A third generation wave model for coastal
regions. Part 1. Model description and validation, J. Geophys. Res., 104(C4), 76497666.
Deltares, 2008, Gibraltar-West wave conditions: memo on analysis of the 10 October 2008
storm, H5284, author: B. de Sonneville, prepared for Van Oord.
Deltares, 2009a: West Gibraltar; Extreme wave conditions, H5284, report+memo, author: B.
de Sonneville, prepared for Van Oord.
Deltares, 2009b: Royal Gibraltar Yacht Club; Extreme wave conditions. Prepared for Gifford
LLP.
Deltares, 2012: Wave modelling to support the design of a marina within Gibraltar Port waters:
Extreme wave conditions, author: Z. Peng, prepared for Government of Gibraltar.
Deltares, 2015a: Mean wave conditions reaching a floating dock moored at Coaling Island
Gibraltar, author: H. Nogueira, prepared for Gibraltar Joinery & Building Services.
Deltares, 2015b: Small boats Marina in Gibraltar Port waters, Modelling of the 200-yr wave
conditions, author: A. Camarena Calderon, prepared for Government of Gibraltar.
Deltares, 2019: Eastside Coastal Protection Works (North); DES.2: Numerical modelling of the
200 yr return period extreme wave conditions, author: B. Reijmerink, prepared for Lievense.
Hasselmann, K., T.P. Barnett, E. Bouws, H. Carlson, D.E. Cartwright, K. Enke, J.A. Ewing, H.
Gienapp, D.E. Hasselmann, P. Kruseman, A. Meerburg, P. Müller, D.J. Olbers, K. Richter,
W. Sell and H. Walden, 1973: Measurements of wind-wave growth and swell decay during
the Joint North Sea Wave Project (JONSWAP), Dtsch. Hydrogr. Z. Suppl., 12, A8
Rijn, L. C. van, 1993. Principles of Sediment Transport in Rivers, Estuaries and Coastal Seas.
Aqua Publications, The Netherlands.
Rogers, W.E., P.A. Hwang, and D.W. Wang. 2003: Investigation of wave growth and decay in
the SWAN model: three regional-scale applications, J. Phys. Oceanogr., 33, 366-389.
Van der Westhuysen, A. J., 2010: Modeling of depth-induced wave breaking under finite depth
wave growth conditions. J. Geophys. Res, 115, C01008, 2010.
WL | Delft Hydraulics, 2006a: East Side development Gibraltar, Volume 1 Flow conditions,
H4725.50, author: B. van Vossen, prepared for EBG.
WL | Delft Hydraulics, 2006b: East Side development Gibraltar, Volume 2: Normal wave
conditions, H4725, author: S. Caires, prepared for EBG.

WL | Delft Hydraulics, 2006c: East Side development Gibraltar, Volume 4 Extreme wave
conditions, H4725.50, author: S. Caires, prepared for EBG.
WL | Delft Hydraulics, 2006d: Gibraltar West, Memo of extreme wave conditions. MCI-
18452/H4693-02/JvO, authors: J. van Os and S. Caires. Prepared for DHV.
Wu, J., 1982: Wind-stress coefficients over sea surface from breeze to hurricane, J. Geophys.
Res., 87, C12, 9704-9706
Zijlema, M., 2010: Computation of wind-wave spectra in coastal waters with SWAN on
unstructured grids. Coastal Engineering, 57, 267-277.
Zijlema, M., G.Ph. van Vledder and L.H. Holthuijsen, 2012: Bottom friction and wind drag for
wave models. Coastal Engineering, 65, 19-26.

TH File Compressed Main File 2mb

Uploaded by

Copyright:

Available Formats

TH File Compressed Main File 2mb

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TH File Compressed Main File 2mb

Uploaded by

Copyright:

Available Formats

Gibraltar Harbour

Coaling Island Reclamation Study

Client Project Attribute Pages

Gibraltar Harbour Coaling Island Reclamation Study

Client Project Attribute Pages

Version Date Author Initials Review Initials Approval Initials

Gibraltar Harbour Coaling Island Reclamation Study

3 Hydrodynamic and flushing conditions 28

4 Sediment plume dispersion 44

Gibraltar Harbour Coaling Island Reclamation Study i

4.2.4 Evaluation criteria 49

A Figures/Tables of wave study tasks A-1

B Background on computation of flushing times B-1

C Results of hydrodynamics and flushing simulations C-1

D Turbidity assessment of backhoe dredging D-7

E Dredge plume simulations E-1

Gibraltar Harbour Coaling Island Reclamation Study ii

Gibraltar Harbour Coaling Island Reclamation Study 1 of 63

1.4 Project team

1.5 Outline of this report

Gibraltar Harbour Coaling Island Reclamation Study 2 of 63

2.2 Metocean data used for model input

2.2.1 200-yr Offshore wave and wind conditions

The considered data sources covered the following periods:

Gibraltar Harbour Coaling Island Reclamation Study 3 of 63

Table 2.1 200-yr offshore wave and wind conditions

2.2.2 Normal offshore wave and wind conditions

Gibraltar Harbour Coaling Island Reclamation Study 4 of 63

Gibraltar Harbour Coaling Island Reclamation Study 5 of 63

Gibraltar Harbour Coaling Island Reclamation Study 6 of 63

2.2.3 Water levels

2.3 Wave modelling

Gibraltar Harbour Coaling Island Reclamation Study 7 of 63

2.3.2 Model domains

• Grid 0: an overall domain (250 m x 250 m, green),

Gibraltar Harbour Coaling Island Reclamation Study 8 of 63

Gibraltar Harbour Coaling Island Reclamation Study 9 of 63

2.3.3 Layout, bathymetry and structure schematizations

• Layout 0 - The current situation including the Small Boats Marina.

Gibraltar Harbour Coaling Island Reclamation Study 10 of 63

Gibraltar Harbour Coaling Island Reclamation Study 11 of 63

2.3.4 Output locations

Gibraltar Harbour Coaling Island Reclamation Study 12 of 63

2.3.5 Numerical and physics parameter settings

The applied criteria for numerical accuracy were set as follows:

All runs have been verified to have converged within 80 iterations.

Gibraltar Harbour Coaling Island Reclamation Study 13 of 63

2.3.6 Boundary and input conditions

Gibraltar Harbour Coaling Island Reclamation Study 14 of 63

Water level input

2.3.7 Run definitions

Gibraltar Harbour Coaling Island Reclamation Study 15 of 63

2.4 Modelling results for 200-yr wave conditions

Gibraltar Harbour Coaling Island Reclamation Study 16 of 63

Gibraltar Harbour Coaling Island Reclamation Study 17 of 63

2.4.2 Extreme wave conditions at the output locations

Total wave spectrum

Incoming wave spectrum

Gibraltar Harbour Coaling Island Reclamation Study 18 of 63

Gibraltar Harbour Coaling Island Reclamation Study 19 of 63