The Hongkong Electric Company Ltd
(hereinafter referred to as HK Electric) is considering the development of a
large-scale offshore wind farm in
Following the ERM Site Selection Study and submission of a report to HK Electric in
April 2008, two potential locations for development have been identified,
located at
In July 2006, HK Electric submitted a
Project Profile ([1]) to the Environmental
Protection Department (EPD) in accordance with procedures under the Environmental Impact Assessment Ordinance
(EIAO) for the development of the
wind farm. In response to this, an
EIAO Study Brief was awarded by EPD to outline the requirements for an
Environmental Impact Assessment (EIA) ([2]). This included the requirement to
undertake modelling.
This Method
Statement presents information on the approach for the water quality
assessment and modelling works for the study. Note that at the
time of completion of this Method
Statement the engineering information for both construction and operation
activities is not complete and therefore a general approach to how modelling
will be carried out based on a number of assumptions is provided below.
1.1
Objectives of the Modelling Exercise
The main objective of the modelling work
is to provide quantitative predictions of impacts to hydrodynamics and water
quality that will inform the impact assessment. The specific objectives of the modelling
exercise are:
·
To
identify and determine potential changes to the hydrodynamic regime post
development of the wind farm;
·
to identify
and quantify water pollutant emission sources;
·
to
determine the significance of impacts on sensitive receivers and potential
affected areas;
·
to
identify, predict and evaluate the residual environmental impacts (i.e. after
practicable mitigation); and
·
to
assess any cumulative effects expected to arise during the construction and
operation of the Project in relation to the sensitive receivers and potential
affected uses.
The construction and operational effects
will be studied by means of mathematical modelling using existing models that
will be set up by Deltares (formerly named WL |Delft Hydraulics) on behalf of
the Environmental Protection Department (EPD) or approved by the EPD for use in
environmental assessments.
1.2.1
Hydrodynamic
Model and Water Quality Model
A model has been developed by Deltares. The
These models have been calibrated extensively. The model covers the entire Hong Kong
waters, the Pearl River Estuary and the Sea Area in front of
The model used calibrated bathymetry at the wind farm
sites as shown in Figure 1.1.
Figure 1.1 Bathymetry
|
A refined model has been developed in order to
accurately model potential water quality and hydrodynamic impacts. The
refinement of the grid at was achieved by introducing two additional domains
around the farm and the dredging route (Figure
1.2). The finest grid in Lamma (green) has a resolution below 75m (i.e.
72.5m). The second finest grid in Lamma (red) presents cell sizes of about 150
- 250m. This will cover the landing
points where sensitive receivers are present as well as the wind farm site
proper.
The refined models were verified for both the dry and
wet season at a location to the south west of
Figure 1.2 Refined
Model Grid
|
Figure 1.3 Location
of Sites Selected for Model Verification of Update model vs Refined model
|
Figure 1.4 Model
verification Update model (blue) vs Refined model (red) − Water levels, depth
average current velocities and directions, salinity (West Season)
|
Figure 1.5 Model
verification Update model (blue) vs Refined model (red) − Water levels, depth
average current velocities and directions, salinity (Dry Season)
|
1.3
Coastline & Bathymetry
Hydrodynamic data will be obtained using coastline
and bathymetry for a time horizon representative of the construction and
operation of the facility (i.e., 2010 onwards).
The current patterns under the baseline situation
will be generated as an output of the hydrodynamic modelling. They will be presented as vectors
showing the current velocities (i.e. both direction and magnitude). At the SW Lamma site, the average
surface current velocity is 0.35 m s-1 in the wet season with a maximum of up to
0.61 m s-1 (2007 data); and an average of 0.15 m s-1
in the dry season with
a maximum of up to 0.26 m s-1 (2007
data).
For the hydrodynamic assessment, three scenarios will
be modelled, i.e. one for the baseline situation and one for each
of the operational cases for the two proposed sites (each of them covering a
complete spring/neap cycle for both the dry and wet seasons).
The
modelling will consider the impacts of the development of the offshore wind
farm at both sites during construction and operation (see Sections 3 and 4,
respectively). During the construction phase, modelling
will consider the impact dredging, jetting and foundation construction
works. During operation, the impact
of the wind farm on the hydrodynamic regime will be assessed,
1.6
General Assessment Assumptions
In carrying out the assessment, the worst case
assumptions have been made in order to provide a conservative assessment of
environmental impacts. These
assumptions are as follows:
·
The
assessment is based on peak dredging and jetting rates. In reality these will only occur for a
short period of time;
·
The calculations
of loss rates of sediment to suspension are based on conservative estimates for
the types of plant and methods of working;
·
For
foundation construction, the largest potential for sediment disturbance is
associated with the construction of monopile foundations with scour; and
·
Construction
of a pile and scour protection can occur simultaneously in one working day.
The
modelling will not consider the following aspects. These omissions have been previously agreed with EPD
for modelling works for other projects([3]), such as the Hong Kong LNG EIA and are in
line with modelling assessments for wind farm developments elsewhere.
·
The movement of marine vessels to and from
site, which could have a very localised effect on processes.
·
Scouring of bottom sediment around the
turbine foundation during operation.
This is excluded as it is expected that the release of sediment will be
minimal. If engineering design
determines that significant scouring will occur, then it is likely that
protection to minimise scour will be provided, which means that there should be
no seabed disturbance as a result of scour during the operational phase.
·
The
impacts in terms of contaminants released (i.e. TIN and NH3-N) and
DO depletion will not be modelled explicitly. Instead, they will be quantified on the
basis of the modelled maximum suspended sediment concentrations. This method has been used in a recently
approved EIA ([1]).
·
The
jacking-up operation for turbine foundation emplacement is likely to cause
negligible disturbance (far less than foundation construction) to the seabed
and hence no adverse water quality impacts arising from these activities are
expected.
·
Impacts
on hydrodynamics in the construction phase are typically only likely to be
associated with the presence of engineering equipment, e.g. jack-up barges,
placed temporarily on site. As such
equipment is only likely to be positioned at one site at a time for a
relatively short period of time, the effects on the hydrodynamic regime is
deemed to be very small in magnitude and localised over both temporal and
spatial scales. Therefore, it is
not proposed that these effects be modelled.
The water sensitive receivers (WSRs) have been
identified in accordance with Annex 14
of the Technical Memorandum on EIA
Process (EIAO, Cap.499, S.16). These WSRs are listed in Table 1.1. In addition to WSRs, modelling points
have been added adjacent to the cable route to understand the extent of impacts
associated with jetting activities.
EPD routine marine water and sediment monitoring stations (shown in Table 1.2), in addition to the WSRs,
were also included as discrete model output points.
Table 1.1 Water
Quality Sensitive Receivers (SRs) in the Vicinity of Southwest Lamma Site
Sensitive
Receiver |
Name |
ID |
Fisheries and Marine Ecological Sensitive Receivers |
||
Fisheries
Resources |
|
|
Spawning /
Nursery Grounds |
Spawning /
Nursery Grounds to the West |
SR22 |
|
Spawning / Nursery
Grounds to the East |
SR21 SR14 |
|
Spawning /
Nursery Grounds to the North |
SM6 |
|
Spawning /
Nursery Grounds to the South |
SM18 |
Fish
Culture Zone |
Lo Tik Wan |
SR2 |
|
Sok Kwu Wan |
SR3 |
Marine
Ecological Resources |
|
|
Potential
Coral Communities |
Nam Tsui to
Tai Kok hard coral communities |
SR4 SR23 SR9 SR10 SR19 SR24 SR20 SR21 |
Coral
Communities |
Lamma Power
Station Extension Seawall |
SR15 |
Horseshoe
Crab Nursery Grounds |
Sok Kwu Wan |
SR3 |
Marine Mammal
Habitat |
Southwest
Lamma Waters |
SR1 SM5 |
Green
Turtle Habitat |
Sham Wan |
SR6 |
|
|
SR1 SR6 SR13 SR14 |
Water Quality Sensitive Receivers |
|
|
Gazetted
Beaches |
Cheung Chau
Tung Wan |
SR7 |
|
Kwun Yam |
SR8 |
|
Hung Shing
Yeh |
SR9 |
|
Lo So Shing |
SR10 |
Seawater
Intakes |
Cheung Chau
|
SR11 |
|
Lamma Power
Station |
SR12 |
|
Yuen Kok |
SR13 |
Jetting Mixing Zone |
||
Mixing zone |
|
SR16 SR17 SR18 |
Table 1.2 EPD
Routine Marine Water and Sediment Monitoring Stations in the Vicinity to the
Wind Farm Site and
EPD
Monitoring Stations |
Respective
WCZ |
Marine
Water |
|
SM5, SM6, SM7, SM18 |
|
Marine
Sediment |
|
SS3,
SS4 |
|
For the construction phase the WAQ model will be used
to directly simulate the following
parameters:
·
suspended
sediments; and
·
sediment
deposition.
It is assumed that
the worst-case construction impacts will be at the commencement of dredging,
when there is no depression formed to trap sediments disturbed during works.
The assumptions on
working time in the model are summarised in Table
2.1.
Table 2.1 A
Summary of Working Time Assumed in the Model for Various Construction
Activities
Construction Activity |
Location |
Assumption of Working Time |
Foundation
Construction |
Open sea water |
24 hours per
day over 7 working days |
Submarine Cable
Circuit |
Open sea water
- jetting |
12 hours per
day over 7 working days per week |
|
Landing point –
grab dredging |
12 hours per
day ([4])
|
|
|
|
Grab dredgers will be utilised in the nearshore cable
landing area for both sites to construct a short underwater trench. The trench will be trapezium shape with
bottom width of 5 m. The upper
width shall be 8 to 12 m and the trench depth of 1.5 – 3.5 m deep. It is assumed that the trench length
will be a maximum of 100 m. Grab
dredgers may release sediment into suspension by the following mechanisms:
·
Impact
of the grab on the seabed as it is lowered;
·
Washing
of sediment off the outside of the grab as it is raised through the water
column and when it is lowered again after being emptied;
·
Leakage
of water from the grab as it is hauled above the water surface;
·
Spillage
of sediment from over-full grabs;
·
Loss
from grabs which cannot be fully closed due to the presence of debris;
·
Release
by splashing when loading barges by careless, inaccurate methods; and
·
Disturbance
of the seabed as the closed grab is removed.
In the transport of dredging materials, sediment may
be lost through leakage from barges.
However, dredging permits in
Sediment is also lost to the water column when
discharging material at disposal sites.
The amount that is lost depends on a large number of factors including
material characteristics, the speed and manner in which it is discharged from
the vessel, and the characteristics of the disposal sites. It is not necessary to address water
quality issues at the intended disposal site as these areas have already been
permitted by EPD and CEDD and the environmental acceptability proven. Hence modelling of impacts at disposal
sites does not need to be addressed and reference will be provided with respect
to these studies in the EIA Report
for this Project.
Loss rates have been taken from previously
accepted EIAs in Hong Kong ([5]) ([6]) ([7])
and have been based on a
review of world wide data on loss rates from dredging operations undertaken as part
of assessing the impacts of dredging areas of Kellett Bank for mooring
buoys ([8]). The
assessment concluded that for 8 m3 (minimum) grab dredgers working
in areas with significant amounts of debris on the seabed (such as in the vicinity
of existing mooring buoys) that the loss rates would be 25 kg m-3
dredged, while the loss rate in areas where debris is less likely to hinder
operations would be 17 kg m-3 dredged. It is assumed at this site from previous
information collected for the area, as well as the results of the geophysical
surveys undertaken from the present study, that there should be minimal debris ([9]) ([10]). The
value of 17 kg m-3 will therefore be used for this study. This takes into account the occasional
failure of the grab.
Generally, a split-bottom barge could have
a capacity of 900 m³. A bulk factor of 1.3 would normally be applied, giving a
dredging rate of 700 m³ per barge. The hopper dry density for an 800 to 1,000 m3
capacity barge is around 0.75 to 1.24 ton m-3.
The average release rates will, in fact, be somewhat
less than those indicated above. The instantaneous dredging (and loss) rates
will also decrease as the depth increases. This is because the assumed dredging
production rates are instantaneous rates that will not be maintained due to
delays for breakdowns, maintenance, crew changes and time spent relocating the
dredgers. The release rates that are to be modelled area, therefore, considered
to represent conservative conditions that will not prevail for any great length
of time.
The hourly production rate for sediment
dredging using a grab dredger of 8m3 adopted in the Kellett Bank
Study was 208.3 m3 hour-1 ([11])
using medium sized grabs (8 m3)
that will be used for this Project.
Its expected at this stage that dredging will be undertaken 12 hours per
day during daylight hours.
Therefore in a single day of activity 2499.6 m3 of sediment
will be dredged, which equates a release of 42,493.2 kg of sediment. The sediment release rate would
therefore equate to 0.984 kg s-1 with the assumption that only one
dredger will be utilised. The
leakage from grab dredgers can be throughout the water column as the dredger
lifts sediments from the seabed to the barge.
The use of cage-type silt curtains would be expected
to reduce the rate of suspended sediment levels typically between 76% and 81%.
However, for the purpose of this assessment, dredging without silt curtains
will be taken forward to provide a worst case scenario approach. If levels are seen to be unacceptable
then calculations can be made associated with a reduction expected with the
adoption of this mitigation.
It is assumed that
cable installation, will be largely undertaken using jetting methods. This method provides the greatest potential
for sediment release and therefore presents the worst case scenario.
Jetting speeds
have been taken as 360 m hr-1 for cable circuit installation ([12]). This rate relates to typical practices
by contractors in Hong Kong that would be involved in these works.
The maximum burial
depth for each installation will be 5 m and have a cross-sectional area of 0.75
m² (0.5 x 5 m x 0.3
m).
It will require
one pass of the jetting machine to reach the required burial depth. This will be temporary and instantaneous
disturbance to the seabed since the disturbed sediment is expected to settle on
the seabed in a short period after the jetting machine has passed. The rate of disturbance for the cable
installation will be 0.075 m3 s-1 ([13]).
It is conservatively assumed that 20% of the
disturbed sediment enters suspension and this would give a loss rate.
The loss rate used here has been used in previous projects for submarine
utility installations under the EIAO that
have been installed using jetting methods and have obtained Environmental
Permits:
·
Liquefied
Natural Gas (LNG) Receiving Terminal and Associated Facilities (AEIAR-106/2007). EP granted on 3 April 2007
(EP-257/2007).
·
The
Proposed Submarine Gas Pipelines from Cheng Tou Jiao Liquefied Natural Gas
Receiving Terminal, Shenzhen to Tai Po Gas Production Plant, Hong Kong – EIA
Study (AEIAR-071/2003). EP granted on 23 April 2003
(EP-167/2003).
·
132kV
Submarine Cable Installation for Wong Chuk Hang - Chung Hom Kok 132kV Circuits
(AEP-126/2002). EP granted on 2
April 2002 (EP-126/2002).
·
FLAG
North Asian Loop (AEP-099/2001). EP granted on 18 June 2001
(EP-099/2001).
·
East
Asian Crossing (EAC) Cable System (TKO), Asia Global Crossing (AEP-081/2000). EP granted on 4 October 2000
(EP-081/2000).
·
East
Asian Crossing (EAC) Cable System, Asia Global Crossing
(AEP-079/2000). EP granted on 6
September 2000 (EP-079/2000).
·
Submarine
Cable Landing Installation in Tong Fuk Lantau for Asia Pacific Cable Network 2
(APCN 2) Fibre Optic Submarine Cable System, EGS. EP granted on 26 July 2000 (EP-069/2000).
·
Telecommunication
Installation at Lot 591SA in DD 328, Tong Fuk, South Lantau Coast and the
Associated Cable Landing Work in Tong Fuk, South Lantau for the North Asia
Cable (NAC) Fibre Optic Submarine Cable System
(AEP-064/2000). EP granted in June
2000 (EP-064/2000).
To calculate the mass entrainment rate it is necessary to apply a dry
density for the material, which is conservatively assumed to be 600 kg m-3. As this dry density has been assumed in
a number of the assessments listed above and approved under the EIAO it is considered appropriate for
use in the present study.
Using the above assumptions, it is therefore determined that the maximum
sediment loss rate would be:
0.075 m3
s-1 x 20% x 600 kg/m³ = 9 kg s-1
The sediment will be entered into the model in the model layer closest
to the seabed because this will represent the entrainment of sediment to
suspension from the layer of fluid mud flowing over the existing seabed. This approach is considered valid as the
jetting machine is fluidising the seabed sediments and not excavating the
sediments, consequently there will be little vertical entrainment of sediment
into the water column.
The sediment release rate can be lowered by reducing
the jetting speed. However, the
above figure will be assumed for the modelling as this present the maximum
jetting speed and therefore the maximum release rate.
The sediment will be entered into the model within a
series of grid cells to represent the jetting machine moving along the cable
route. Thus each grid cell will
represent a section of the cable route and sediment will be entered into that
grid cell for the length of time it takes the jetting machine to pass the
length of that cell, based on the jetting machine speeds given above. Once the jetting machine has passed that
grid cell, sediment will then be entered in the next grid cell on the
route. The sediment release in the
bed layer (constitute 10 %) of the water column will be assumed in the model.
It should be noted that theses assumptions have been
adopted in the previous approved EIAs and a number of jetting contractors have
confirmed that these assumptions are reasonable and practical.
2.4
Turbine Foundation Construction
The foundation options for the wind farm are as
follows:
·
multi-pile
(tripod); and
·
monopile
The monopile structure is anticipated to
have a diameter of 5 to 7 m which will lead to a physical footprint of
approximately 38.5 m2 with a pile wall thickness of approximately 80
mm. For tripods, the diameter of
each tripod pile is estimated to be 1.3 m with a 7 m separation distance
between each tripod pile. The
subsequent physical footprint will be in the order of 22 m2 for each
tripod group (worse case triangular area).
The below seabed pile wall thickness of tripod foundations will be
approximately 50 mm.
For modelling purposes it is assumed that
foundation scour protection will be constructed as this would provide
opportunity for greatest sediment disturbance. It is assumed that this scour protection
will have an overall width of 30m and length of 30m and overall area of 900 m2.
The area of scour revetment would therefore encompass
the area of turbine foundation disturbance.
Foundations can be constructed using percussive pile
or boring techniques. The piles
would be comprised of tubular steel.
If percussive driving is taken forward the large majority of sediment will
be collected within the pile and not released into the water column. For boring works, it is normal practice
for casing to be driven into the seabed before drilling. The purpose of this drill sleeve is to
contain fine material and to prevent the excavated hole from collapsing. Therefore there should be a very small
release of fines associated with piling activity.
It is not common practice to model the release of
sediment associated with piling activity and construction of scour
revetment. However, given the
potential for sediment release the following assumptions are made:
·
Only
surface sediment (taken as material up to 1m below the surface of the seabed)
agitated by piling and/or construction of scour protection could be released
into the water column;
·
Modelling
will only consider the area of disturbance associated with the construction of
scour protection as the area of turbine foundation is enclosed within this
area. Calculations can be made on
the potential impact of constructing turbine foundations, which have a much
lower footprint of impact;
·
The
whole area in the footprint of the pile will be disturbed. In reality this is unlikely as pile
walls are relatively thin and these would provide the area that causes any
agitation;
·
Only
20% of the sediment agitated will be released into suspension as per standard
calculations agreed for jetting techniques; and
·
The
release of sediment will be instant and will only occur as a single event in
any day of working. In reality,
this is unlikely. However, it presents the worst case event for sediment
release.
For scour protection works in a 24 hour working day
the maximum volume of sediment released would be 180 m3 (20% of the
area of disturbance for one pile foundation). Using the dry density value of 600 kg m-3,
this equates to an overall release of 108,000kg of sediment per turbine in a
single working day. If this
sediment were to be released as a constant through the working day, this would
relate to a loss rate of 1.25 kg s-1.
The sediment will be entered into the model in the
model layer closest to the seabed.
2.5
Indicative Construction Programme and Sequence
The provisional timeframe
for the completion of construction activities is provided below. More detailed information on the activity
schedule will be obtained prior to commencement of modelling, including a daily
programme.
·
Foundation
construction - June
2011 to December 2012
·
Cabling
and offshore substation - June 2012 to
June 2013
·
Wind
turbine installation - January
2013 to September 2013
For simulating sediment impacts the following general
parameters has been used for suspended sediments once disturbed:
·
Settling
velocity – 0.5 mm s-1
·
Critical
shear stress for deposition – 0.2 N m-2
·
Critical
shear stress for erosion – 0.3 N m-2
·
Minimum
depth where deposition allowed – 1 m
·
Resuspension
rate – 30 g m-2 d-1
·
Wave
calculation method – Tamminga
·
Chezy
calculation method – White/Colebrook
·
Bottom
roughness – 0.001 m ([14])
·
Fetch
for wave driven erosion – 35 km
·
Depth
gradient effect on waves – absent
The above parameters have been used to simulate the
impacts from sediment plumes in
Delft3D-FLOW will be carried out to
simulate the operational hydrodynamic conditions for both proposed sites. Three scenarios will be modelled, i.e. one for the baseline situation
and one for each of the operational cases for the two proposed sites (each of
them covering a complete spring/neap cycle for both the dry and wet seasons).
It is presently assumed that the turbines
will be rated 2.3 – 3.6MW and the total number installed will be no greater
than 35 units. At the SW Lamma site
is it preliminarily proposed that there will be a separation between turbines
of 360 m on the north / south axis and 650 m on the east / west axis. Should 3.6MW class wind turbine be
selected, the number of wind turbines would be reduced to around 28 to 30 in
order to maintain the wind farm capacity of around 100MW
As discussed, monopile foundations have the greatest
potential physical blockage to the physical environment and therefore have the
most potential to cause changes to processes. The dimensions of these foundations will
therefore be used for modelling as part of a worst case scenario approach.
Since the 50 – 75m resolution in the most
detailed model domains is insufficient to resolve the individual piles, the
hydraulic structure option was used to include the effects of the sub-grid
piles on the flow in the model.
This option is commonly used to include piles, such as bridge piles, in
hydrodynamic modelling and is based on an additional quadratic friction term to
the momentum equations. The energy loss term for this friction is derived by
the following equation, which includes pile diameter and number relative to the
grid dimensions:
Since the wet and dry seasons result in
entirely different flow conditions, the model simulations are (commonly)
carried out for both seasons separately. This results in the following
scenarios:
1. West site, base case, dry season
2. West site, base case, wet season
3. West site, including wind farm, dry
season
4. West site, including wind farm, wet
season
At present the identified potential concurrent
projects are the marine dumping activities near Ninepins Islands, East of Tung Lung
Chau and South Cheung Chau, proposed Hong Kong Offshore Wind Farm in
South-eastern Waters (EIAO Ref: EIA Study Brief No. ESB-146/2006),
potential sand reserves in the eastern waters. However, there are not sufficient details
of these projects at this preliminary stage to determine the possible
concurrent construction/operational activities. Therefore, modelling works for
cumulative water quality impacts during the construction phase and/or
operational phase arising from the concurrent projects with the proposed
Project will be decided, if necessary, upon collection of adequate information
on the potential concurrent projects.
The modelling work will provide quantified
information on the potential impact of construction works on water
quality. This information will
inform the assessment of potential effects of construction on water sensitive
receivers with reference to Hong Kong Water Quality Objectives. In addition, the modelling will identify
the impact of the operation of the wind farm on hydrodynamic processes.
The modelling results will be provided as a Technical
Appendix to the EIA Report and information summarised as appropriate in the
water quality and hydrodynamic processes assessment sections of the EIA Report.