ArticlePDF Available

Hydrodynamic impact and power production of tidal turbines in a storm surge barrier

Authors:

Abstract and Figures

To estimate the impact on energy production and environment of tidal turbines placed in the Eastern Scheldt Storm Surge Barrier a Computational Fluid Dynamics (CFD) study has been carried out on the additional head differences induced by the turbines. The CFD model focusses on a single gate opening of the Storm Surge Barrier and includes half of the adjoining gates on either side. In this 40 m wide Gate a 1.2 MW array existing of five Tocardo T2 tidal turbines has been installed as part of a demonstration project in 2015. Transient computations of the barrier with and without the turbine array were carried out for a range of quasi stationary tidal phases. The turbines are resolved in detail as rotating equipment: real-time rotation of the turbine blades (involving the displacement of the mesh nodes in an unsteady setting) is implemented, and torque and thrust for the prescribed speed of rotation is provided as output. The results for velocity, power and thrust are compared with field experiments to validate the model. Based on these computations an estimate of the effect of turbines on the discharge capacity of the storm surge barrier is given. This information will be used to parameterize the tidal turbines in the far-field hydrodynamic model of Eastern Scheldt estuary for the ultimate assessment of the effect of tidal turbines on energy production and on the environment.
Content may be subject to copyright.
Abstract—To estimate the impact on energy production and
environment of tidal turbines placed in the Eastern Scheldt
Storm Surge Barrier a Computational Fluid Dynamics
(CFD) study has been carried out on the additional head
differences induced by the turbines.
The CFD model focusses on a single gate opening of the
Storm Surge Barrier and includes half of the adjoining gates
on either side. In this 40 m wide Gate a 1.2 MW array
existing of five Tocardo T2 tidal turbines has been installed
as part of a demonstration project in 2015.
Transient computations of the barrier with and without the
turbine array were carried out for a range of quasi stationary
tidal phases. The turbines are resolved in detail as rotating
equipment: real-time rotation of the turbine blades
(involving the displacement of the mesh nodes in an
unsteady setting) is implemented, and torque and thrust for
the prescribed speed of rotation is provided as output. The
results for velocity, power and thrust are compared with
field experiments to validate the model.
Based on these computations an estimate of the effect of
turbines on the discharge capacity of the storm surge barrier
is given. This information will be used to parameterize the
tidal turbines in the far-field hydrodynamic model of
Eastern Scheldt estuary for the ultimate assessment of the
effect of tidal turbines on energy production and on the
environment.
Keywords—CFD validation, Tidal Turbines, Hydraulic
Structure, Free Surface Flow
I. INTRODUCTION
HE Eastern Scheldt Storm Surge Barrier consists of
62 individual gates, and is constructed of concrete
pillars, top beams and sill beams connecting to a rockfill
sill construction and about 600 m of bed protection on both
sides, see [1]. During ebb and flood the maximum head
difference over the barrier is about 1 m, with maximum
velocities of 4 m/s and higher, making it an ideal site for
the generation of tidal stream energy.
In 2015 an array of five tidal turbines was deployed in
Gate #08 of the Roompot Section of the barrier in the
Submission 1428. Tidal hydrodynamic modelling. The work is part
of the Dutch Marine Energy Centre (DMEC) project funded by the
European Regional Development Fund (ERDF) and co-financed by
the Province of Noord-Holland.
T. S. D. O’Mahoney, A. de Fockert, A. C. Bijlsma, Deltares,
Boussinesqweg 1, 2629 HV Delft, The Netherlands (e-mail:
framework of a tidal power pilot project. This project
initiated the development of numerical modelling tools to
assess t he effects of the tidal turbines in thi s type of barr ier,
which need to be determined as part of the permit process.
Previously, the CFD modelling of a tidal turbine in the
proximity of the free surface was addressed, [2]. As a
second step, the CFD modelling of the flow through the
gates of the barrier (without turbines) was studied and
compared with field measurements of ADCP horizontal
and vertical profiles in the gate opening [3].
Analysis of the performance of tidal turbines has been
conducted with BEM methods ([4, 5]), CFD with frozen
rotor [6] or actuator disc [7, 8]. More recently CFD methods
with rotating turbine blades in a sliding mesh
configuration have been used to give a more detailed view
of the flow f ield and turbulence at the i nlet and wake of the
turbines. Most experiments [9-12] consider isolated
turbines at lab scale for a marine current application where
tom.omahoney@deltares.nl,anton.defockert@deltares.nl,
arnout.bijlsma@deltares.nl ).
P. de Haas, formerly Tocardo Tidal Power BV, Sluiskolkkade 2,
1779 GP, Den Oever, The Netherlands (e-mail:pdhaas@haske.net)
Hydrodynamic impact and power production
of tidal turbines in a storm surge barrier
Tom S. D. O’Mahoney, Anton de Fockert, Arnout C. Bijlsma and Pieter de Haas
T
Fig. 1. Location of Gate+08 of the Roompot section of the Eastern
Scheldt Barrier, shown on a satellite image (source: Google Earth)
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
127
Manuscript received 16 March; accepted 30 March; published 30
November, 2020. This is an open access article distributed under the terms of
the Creative Commons Attribution 4.0 licence (CC BY
https://creativecommons.org/licenses/by/4.0/). This article has been subject to
single-blind peer review by a minimum of two reviewers. The work is part
of the Dutch Marine Energy Centre project funded by the European Regional
Development Fund and co-financed by the Province of Noord-Holland.
T. S. D. O’Mahoney, A. de Fockert, A. C. Bijlsma, Deltares, Boussinesqweg 1,
2629 HV Delft, The Netherlands (e-mail: tom.omahoney@deltares.nl,
anton.defockert@deltares.nl, arnout.bijlsma@deltares.nl). P. de Haas,
formerly Tocardo Tidal Power BV, Sluiskolkkade 2, 1779 GP, Den Oever, The
Netherlands (e-mail:pdhaas@haske.net). Digital Object Identifier
https://doi.org/10.36688/imej.3.127-136
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
the flow is unidirectional. Subsequently, CFD validation
efforts have therefore mostly concentrated on such
situations. ([13-16]). Very few studies [17] have studied
flow directionality. As the turbines in all these studies are
isolated the blockages are relatively low (from 6 % in [8],
12 % in [10], to 17 % in [18, 19]), although in one study a
very high blockage was used [20]. Higher blockages have
also been studied for cross flow turbines [21]. However,
turbines located i n a Storm Surge Barrier will have a higher
blockage than those available in this literature.
Similarly, the location of turbines in a pre-existing
hydraulic structure leads to complex inflow conditions
and interactions with the structure. Some authors have
shown the importance of inflow velocity profile and inflow
turbulence on turbine performance, wake development
and the performance of CFD models [7, 22, 23]. The
interaction with other turbines in arrays, both in terms of
performance and wake recovery, has been studied but
only in the absence of other structures [24-27]. Similarly an
uneven bathymetry is often not included although the
probable importance is stressed [12]. Field measurements
in a real bathymetry have been made [28], on a site before
turbine installation, and consequently lab scale
measurements were scaled up for the CFD simulations
with a turbine [16]. Alternatively the CFD has been
performed at full-scale with measured inflow profiles [29].
The most recent work has measur ed in the field around the
turbine fence in the Eastern Scheldt Barrier [30].
The aim of the current paper is to combine the
modelling of bathymetry, surrounding structures and
rotating blades to simulate a field scale tidal turbine array
with high blockage owing to its location in a Storm Surge
Barrier. Comparison with field scale measurements at the
location of the turbines, with and without turbines is made
in order to validate the CFD model. The CFD results give
information about the interaction of the wake with that of
the barrier and bathymetry, in terms of head difference
and turbulence, and can help to inform modelling efforts
at a larger system scale, which would be used for
environmental impact assessments.
II. METHODOLOGY
The approach that has been taken in this study involves
the comparison of CFD simulations with field
measurements of velocity profiles, power and thrust in
order to validate a numerical modelling approach. The
simulations and measurements are made for situations in
with and without turbines. Subsequently, the validated
CFD model is used to calculate the effect of the presence of
the turbines on the discharge characteristics (discharge
coefficient) of the barrier, something which is not directly
measured. The numerical tool would then also be available
to study the effect on the flow field, velocities and
turbulence near the bed, and on the bed protection. This
section gives a short summary of the available
measurements and the conditions used for the CFD model.
A. Velocity measurements
ADCP measurements have been carried for both the
situation before turbine deployment (2011) and during
turbine deployment (2016, 2017). The locations of the
Fig. 2. Detail image of t
he Roompot section with the location of
Gate#08 (in red) and the approximate extent of the CFD domain (in
green)
a)
b)
Fig. 3. Overview of the ADCP measurement: a) in 2011 without
turbines and b) in 2015/2016 mounted from the turbines. Beams 1 and
3 are in the horizontal plane
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
measurements are shown in Figure 3. The vertical
locations are given relative to Amsterdam Ordinance
Datum (NAP) which is approximately mean sea level.
In 2015, 5 turbines were installed on the Eastern Scheldt
side of Gate #08 of the Eastern Scheldt Barrier Roompot
section. The middle and outer turbines were equipped
with two ADCP’s, one pointed forwards (towards the
North Sea) and one pointed backwards (towards the
Eastern Scheldt), see Figure 3. The ADCP devices measure
along 1 beam by default but have also been used to
perform 5-beam measurements. Different periods have
been measured. The performance of this type of turbine in
isolation has been studied previously [2].
For each measurement a median is made over a number
of tidal cycles of the measured variable (velocity, turbine
RPM, turbine power or thrust) at each head to give a
relation between the head difference and the median of the
measured variable. The measured results shown in this
paper are therefore the median values of the moments in
the tidal cycles for which the heads are those as given in
Table 1. For each head there are two moments in the tidal
cycle at which it occurs; at the start and end of the tidal
phase. Because of the inertia in the system the velocities at
the end of the tidal phase are higher than at the start. These
two moments are given separately in the results section of
this paper.
The four cases chosen for this study include two ebb
cases and two flood cases. The four cases are moments in
the tidal cycle at which the ADCP data has been analysed.
The range of the water levels and heads is determined by
the range of ADCP data for which a positive quality check
is available. They do not cover the entire range of
operation of the turbines.
B. Power measurements
The generated power of the turbines has been analysed
in a similar way to that of the rotation rates; giving a
median power over many tidal cycles as function of the
head difference. The resulting power for the head
differences in each of the simulation conditions is given in
Table 2. The power generated in the simulations is not
specified but is a result of the simulation. The moment and
force on the turbines is a direct output from the
simulations, calculated by Star-CCM+ as an integration of
pressure across the surface of the blades. The information
of the measured power is used for comparison with the
CFD in the results section. The values as specified in Table
2 are measured power values to the rotor after being
corrected for the system losses.
It was seen that the power that is generated at the start
of the tidal phase is significantly lower than at the end of
the tidal phase. This is most likely the result of the
approach flow towards the turbines, which for ebb is
straighter at the end of the tidal phase than during the start
of the tidal phase. For flood the oblique flow is not present.
A hydrodynamic model (Delft 3D) of the Eastern Scheldt
estimated that the flow angle can be as high as 31° to the
axis of the turbines at 200 m from the barrier for the low
heads at the start of the ebb phase. For the higher heads
and at the end of the ebb phase the angle is approximately
6°. An oblique approach flow leads to a reduced power
production, particularly in turbine 5.
C. Thrust measurements
Thrust measurements have been performed by Tocardo
on the 3th and 5th turbine in the barrier. Similar to the
power measurements, these measurements have been
analysed and used as validation by the CFD model. The
thrust is measured with thrust gauges, which are mounted
on the strut of the turbine. The values as presented in Table
3 are analysed based on the head differences over the
barrier, similar to the power and ADCP measurements
D. CFD simulations
The simulations have been performed using the Star-
CCM+® software, version 11.02.10, a proprietary and
TABLE I
SUMMARY OF THE CONDITIONS USED FOR THE ANALYSIS
Tide Case
Number
Water level
North Sea
w.r.t. NAP
Water level
Eastern Scheldt
w.r.t. NAP
Head
[m]
Ebb NT,1/WT.1 0.93 1.13 -0.20
NT.2/WT.2 0.68 1.00 -0.32
Flood
NT.3/WT.3 -0.57 -0.77 0.20
NT.4/WT.4 1.12 0.57 0.55
NT represents No Turbines, WT represents With Turbines
Fig. 4. Contour plot of the bed elevation. Dimension are w
ith
reference to NAP
TABLE 2
SUMMARY OF THE MEASURED POWER OF THE TURBINES
Tide Case T1
[kW] T2
[kW] T3
[kW] T4
[kW] T5
[kW]
Ebb WT.1 17.16 16.11 14.61 15.65 13.17
WT.2 36.38 34.86 33.45 34.80 33.03
Flood
WT.3 50.51 51.06 48.10 50.20 51.05
WT.4 219.99 223.28 214.06 217.36 222.02
TABLE 3
SUMMARY OF THE MEASURED THRUST OF THE TURBINES
Tide Case
T
1
[kN]
T2
[kN]
T3
[kN]
T4
[kN]
T5
[kN]
Ebb WT.1 n/a n/a -20.00 n/a -25.43
WT.2 n/a n/a -33.97 n/a -40.06
Flood
WT.3 n/a n/a 43.25 n/a 45.39
WT.4 n/a n/a 105.04 n/a 117.68
O’MAHONEY et al.: HYDRODYNAMIC IMPACT OF TIDAL TURBINES IN A STORM SURGE BARRIER
129
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
commercial CFD package. Star-CCM+ is able to solve the
Navier-Stokes equations for mass and momentum in a
three-dimensional domain, both steady-state or transient.
The package also includes its own meshing algorithm.
Different types of boundary conditions, numerical
schemes and turbulence models are available.
Star-CCM+ has been used successfully by Deltares in
recent years for the simulation of free surface flow in and
around hydraulic structures, e.g. shipping locks and
sluices, sewage systems and tidal energy applications.
Previous validation of flows around tidal turbines [3] and
the Eastern Scheldt Barrier [4] have also been performed
with this software.
The flow through the barrier is simulated including the
free surface. The highly detailed pillar and sill geometry
result in an extremely turbulent flow downstream of the
barrier and in this work an eddy-resolving turbulence
model is used.
For the CFD simulations the water levels are considered
to be constant, which has the effect of excluding tidal
inertia effects from the model. This is further addressed in
the discussion section.
III. NUMERICAL SETUP
E. General model settings
The equations are solved in transient form, with 1st
order temporal discretization, and a segregated pressure-
based solver. The equations are discretized on a structured
3D grid. The Finite Volume method is used with 3rd order
combination of MUSCL and Central Difference for the
discretisation of the advection terms. To model turbulence
near the turbines and near the sill of the barrier, Improved
Delayed Detached Eddy Simulation (IDDES) turbulence
model has been applied. This model uses a Large-Eddy
Simulation approach in the bulk flow and a Reynolds-
Averaged Navier Stokes approach in the near wall region.
The near wall model uses the SST k-Omega model with
wall functions. An LES model is used for the vast majority
of the domain. To model the free surface air and water are
included as immiscible phases in the Volume of Fluid
(VOF) model. A time step of 0.05 seconds is used for the
final simulations although some of the spin-up of the
simulations is performed with a larger time step 0.1-0.5 s.
In all cases the time step and mesh are sufficient to achieve
a sharp interface for the volume fraction of water between
air and water, this interface being no more than 1 or 2 cells
in the vertical at all locations.).
F. Geometry
The geometry of the CFD simulations encompasses a
region approximately 200 m upstream and 200 m
downstream of the barrier. The bathymetry (see Figure 4)
has been generated from multibeam data (resolution
0.5 m) of the Eastern Scheldt region of 2017 provided by
Rijkswaterstaat and converted to an stl file, with a
resolution of 2 m. It is integrated with the 3D CAD of the
barrier and the turbines.
The geometry encompasses one full port of the storm
surge barrier (Roompot 8) and half of the neighbouring
ports on each side. The geometry has been updated and
verified based on the as built drawings from
Rijkswaterstaat and photographs of the structure during
installation.
The turbine and array geometries have been provided
by Tocardo, the manufacturer and operator of the turbines.
Each turbine has two blades and is mounted on a support
from above. There are 5 turbines in the port of Roompot 8,
each with a diameter of 5.2 m. The distance between the
axes of each turbine is 6.7 m. The configuration can be seen
in Figure 5 with all turbine blades in the same orientation.
In the simulations the turbines do not rotate in sync but at
different (but constant) rotating speeds from each other.
After some spin-up time the turbines are at very different
orientations from each other at any given time. A rotating
zone made of a cylinder of 6 m diameter and 3.2 m length
is located around each turbine hub. Ideally a larger
rotating zone should be used for CFD calculations so that
a)
b)
Fig. 5. 3D CAD view of the geometry of a) the barrier as seen from
the North Sea, and, b) the five turbines with support structure
TABLE 4
SUMMARY OF THEBOUNDARY CONDIT IONS FORTHE CFD
Boundary Conditions
North Sea / Eas
tern
Scheldt Inflow profile and water levels
Bed
No-slip wall (smooth – no roughness)
Barrier
No-slip wall with roughness
Sides (in breadth)
Symmetry
Top
Atmospheric pressure
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
the interpolation between the rotating zone and the
stationary zone (occurring at the interaction surface
between the two zones at each time step) is performed at a
location where the gradients of the flow are small and the
numerical errors in interpolation have a small effect.
However, for these simulations the rotating zone around
the turbines is limited in size owing to the requirement that
they don’t overlap with each other or have a very small
gap between them where cell quality would be low. A zone
of 6 m diameter has been used because it allows for good
quality cells between the different rotating zones and is
therefore considered acceptable for this study.
G. Mesh
The mesh has been generated in Star-CCM+ using
multiple areas of mesh refinement. The largest cells in the
domain are in the far field near the boundaries and are
cube cells of size 1.2 m. Given that the water depth is
approximately 30 m in this region, this gives a minimum
of 20 cells in the vertical. The mesh is refined in the region
60 m upstream and downstream of the barrier such that
the cells in this region become slightly elongated with the
longitudinal and transverse dimension still 1.2 m, but with
vertical dimension 60 cm. The water depth in this region
can be as low as 18 m but still this mesh gives
approximately 30 cells in the vertical. Around the barrier
sill itself the cells are refined further and are again cubic
with dimension 16 cm. The height of the opening above the
sill is approximately 10.5 m but the water depth can be 9
m. In any case the mesh has at least 40 cells in the vertical.
The total number of cells for simulations with turbines is
approximately 55 million and the simulations took 4 weeks
using 40 cores on a cluster of an Intel Xeon E3-1276 v3
processors of 3.6 Ghz.
H. Boundary conditions
The boundary conditions are a combination of water
levels and inlet velocity along with no-slip boundaries for
the wall boundaries at the barrier and bed. The conditions
at the inlet and outlet boundaries are constant throughout
the simulation giving a flow field without inertial effects
associated with a varying tidal water level.
1) Inflow velocity profile
ADCP data of the velocity profile in the Eastern Scheldt
is available for a number of measurement points. MP0002
is in the Eastern Scheldt 750 m from the barrier and at a
location where the bathymetry is representative for the
inflow conditions to Roompot 8. A comparison of
measured velocity profiles in the Eastern Scheldt shows
that a logarithmic profile is representative.
Figure 6 shows comparison of the measured profile
during Ebb which would be representative of the inflow
condition for the CFD model, along with an example of the
logarithmic profile used for the CFD (scaled to the depth
at that location). The measurements show a variation in the
profile between the start of the tidal phase and the end of
the tidal phase due to inertia of the tide. The CFD profile
is between these two measured profiles and is therefore a
good model of the average inflow conditions at the barrier.
2) Water levels
Water levels are set at the downstream boundaries with
a profile of the volume fraction of water, where the volume
fraction is 1 below the water level and 0 above it. The
downstream boundary is the Eastern Scheldt for flood
cases and the North Sea for Ebb cases. An initial water
level is also set for the upstream boundary in the same way
along with a velocity profile. This initial water level may
not be consistent with the discharge through the barrier
and the downstream water level and therefore the
upstream water level varies during the simulation. In
order to have results for the desired set of conditions in
Table 2, the velocity profile is scaled to adjust the discharge
until the achieved upstream water level is within 1 cm of
the desired level. This water level is always averaged over
the breadth of the domain and then time-averaged.
3) Wall boundaries
The barrier itself includes the sill and the pillars which
are of concrete. They are modelled as no-slip walls with a
roughness of worn concrete (1 cm). The bed includes the
small-scale variations of the bed protection and seabed
which have been captured by the multibeam data of 0.5 m
resolution and has been modelled as a no-slip wall without
additional roughness. The sides of the domain located at
the halfway point between the two pillars in gate 7 and 9
so that a symmetry boundary condition can be used.
4) Turbine rotation rate
The rotation speed of the turbines is applied to the
model as an input parameter. The rotation speeds are
based on measured RPMs at the turbines. The rotation
speed in the simulations is therefore chosen to correspond
to the head difference of the simulation; a different speed
is therefore given to each turbine (for more information see
[31]). A summary of the rotation speeds used in the
simulations is given in Table 5.
Fig. 6 Measured profile at MP0002 versus logarithmic vertical
profile used in CFD
O’MAHONEY et al.: HYDRODYNAMIC IMPACT OF TIDAL TURBINES IN A STORM SURGE BARRIER
131
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
IV. RESULTS
I. Discharge coefficients
The discharge through the barrier depends on resistance
of the barrier and the head difference across the barrier.
The discharge coefficient gives the relation between the
head and the discharge for a given situation. It is
dependent on the cross section: in case of the Eastern
Scheldt barrier this is also dependent on the water level.
The discharge can be described by the following formula.
=

2
(1)
In which Q is discharge [m3/s], µ discharge coefficient
[ - ], A the wetted cross-sectional area in the port, [m2], g
gravitational constant [m/s2] and Δh water level difference
[m].
The discharge coefficients from the simulations were
calculated for each CFD case. A distinction is made for a
coefficient over the whole domain and between each port
of the barrier. For the simulations without turbines these
should be identical because of the symmetry in the
geometry and boundary conditions. The Star-CCM+
monitors give a constant discharge at the inlet boundary,
owing to the specification of the inlet profile there. The
discharge at the outlet boundary varies during the
simulation owing to the unsteadiness of the flow. The
value reported in Table 6 is based on a monitor for
discharge through a vertical cross-section at the barrier
which is averaged over 100 s of simulation time after
reaching a statistically steady-state.
For the ebb cases the discharge coefficient (µ) is almost
identical between the cases 1 and 2. The maximum
difference is 2 %. This can be attributed to the very similar
water levels upstream in each case (the difference is only
13 cm) and the small difference in head (only 12 cm). For
the flood cases the upstream water level varies between
the cases by 1.69 m. For a barrier of this type with a lintel
support which is submerged in water for the higher water
levels the discharge coefficient is expected to be dependent
on water levels as well as geometry. However, even in this
case the difference is not large between case 3 and 4,
approximately 3 %.
The effect of the presence of the tidal turbines can be
determined by comparing the simulations with and
without turbines for the coefficient through Gate Roompot
8. The discharge capacity hardly changes during flood and
decreases by about 5 % during ebb. The main reason for
this is the fact that the turbines are positioned upstream of
the barrier sill during ebb and sufficiently far downstream
of the barrier sill during flood. During ebb, the turbines are
blocking the flow towards the barrier, while during flood
the influence of the turbines on the flow becomes effective
after the flow has passed the narrowest cross section of the
gate (above the sill).
J. Velocity profiles
1) Barrier without turbines
The ADCP measuring devices for the vertical profiles
were mounted on the Eastern Scheldt face of the barrier
sill. In the Ebb case this is in a location upstream of the
point of flow separation. The vertical velocity profile
(Figure 7) shows therefore simply an approximately
logarithmic profile with a clear boundary layer starting
from the height of the top of the sill. The measurements are
split between a profile from the start of the tidal phase and
that from the end of the tidal phase. The difference
between the two profiles is due to inertia in the Eastern
Scheldt.
The CFD simulations are made with constant boundary
conditions (water levels) and the results are taken once the
flow field has reached an equilibrium (a statistically steady
TABLE 5
SUMMARY OF THE ROTATION RATES OF THE TURBINES
Tide Case T1
[rpm] T2
[rpm] T3
[rpm] T4
[rpm] T5
[rpm]
Ebb WT.1 -24.33 -24.03 -23.03 -23.66 -21.88
WT.2 -31.29 -31.17 -30.52 -30.97 -30.36
Flood
WT.3 34.95 35.42 34.50 35.07 35.14
WT.4
44.57
44.86
44.47
44.61
44.66
TABLE 6
RESULTS OF THE DISCHARGE CO EFFICIENTS
Tide Case Head
[m]
µ Gate
#08
µ Side
Ebb NT.1 -0.20 0.96 0.98
NT.2 -0.32 0.98 0.99
Flood
NT.3 0.20 0.86 0.87
NT.4 0.55 0.88 0.89
Ebb
WT.1 -0.20 0.93 0.97
WT.2 -0.32 0.92 0.98
Flood
WT.3 0.20 0.85 0.85
WT.4 0.55 0.88 0.88
NT represents No Turbines, WT represents With Turbines
Fig. 7 Vertical ADCP profile from the measurements without
turbines (in blue) and the CFD (in red) – Ebb Case NT.1
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
state), this means that the CFD results are free from inertia.
It is therefore expected that the CFD profile should fall
between the two measured profiles. For this vertical profile
of the ebb case NT.1 this is the case. The agreement with
the numerical model is therefore considered good.
In the flood case the ADCP is located downstream of the
sill and therefore is in the recirculation zone. The
comparison of the CFD results with the measurements is
shown in Figure 8. In this case the variation between the
start of the tidal phase and the end of the tidal phase in the
measurements is small and the velocity in the bulk of the
flow is predicted very well by the CFD model. The size of
the recirculation zone is slightly larger in the CFD than in
the measurements. The horizontal flow profile (not
shown) has similar characteristics to that in the ebb case
and the agreement with CFD is equally good.
2) Barrier with turbines
For the ebb case (Figure 9) the inflow profile is at the top
of the figure. This shows a region of rising speed (because
of the coordinate system used the velocity here is negative
and becomes more negative as it accelerates towards the
barrier). The turbines are located upstream of the barrier,
so the velocity magnitude drops in front of the barrier in
the wake of the turbines. Note that for the simulation
without the turbines the flow accelerates all the way to the
sill and beyond, owing to the contraction of the flow which
continues downstream of the sill in that case. Downstream
of the turbines the velocity recovers. The agreement with
the measurements is good although downstream of the
turbines the uncertainty in the measurements is large.
Upstream of the turbines the CFD simulations show a
small underestimation of the velocity. Also, the CFD
simulation with turbines shows lower velocity magnitudes
than the simulation without turbines. This is owing to the
difference in discharges in the two simulations owing to
the extra resistance of the turbines (the profile is insensitive
to the exact location in the vertical at which it is made). A
slight underestimation of the velocity magnitude in the
CFD when compared to the measurements could also
suggest that the discharge is underestimated by the CFD.
For the flood case (Figure 10) the inflow profile is at the
bottom of the graph showing the flow acceleration
towards the barrier. In the flood case the velocity
magnitude is lower compared to the measurements and
the difference is greater than in the ebb case. The difference
with the simulation without turbines is less in this case but
is still consistent with extra resistance from the turbines.
However, even the CFD simulation without turbines
shows lower inflow velocities than the measurements of
velocity with turbines for the same case. This is
inconsistent and given that the velocities without turbines
have been well validated against the 2011 measurements
Fig. 9 Longitud
inal ADCP profile from the measurements with
turbines (in blue) and the CFD (in red) – Ebb Case WT.1
Fig. 8 Vertical ADCP profile from the measurements without
turbines (in blue) and the CFD (in red) – Flood Case NT.4
Fig. 10 Longitudinal ADCP profil
e from the measurements with
turbines (in blue) and the CFD (in red) – Flood Case WT.4
O’MAHONEY et al.: HYDRODYNAMIC IMPACT OF TIDAL TURBINES IN A STORM SURGE BARRIER
133
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
this throws some doubt on the absolute values of the 2016
measurements. In the analysis of the measurements [31], it
was observed that the flow velocity during flood is about
5-10% larger for the situation with turbines than for the
situation without turbines, whilst the opposite would be
expected. The measurements with turbines are therefore
concluded to show values that are too high.
In the wake region the recovery of the velocity is faster
in the CFD model than in the measurements. For the flood
case the wake region is much larger than in the ebb case
because the flow is not subsequently squeezed through the
barrier (see Figure 11), although note that the ADCP in the
wake during ebb is not in line with the turbine axis (for
flood it is). The recovery is not complete after 50 m (the
CFD domain extends 200 m downstream of the barrier but
only 50 m are plotted in the figure).
The agreement of the CFD with the measurements is
acceptable. Some of the details of the flow patterns are not
predicted correctly, such as the distance of the wake
recovery and from the comparison of the velocity profiles
it can be concluded that the CFD slightly underestimates
the discharge. This could be attributed to the lack of
sufficient mesh resolution in the wake region to model the
turbulence in this region.
K. Turbine power
Tocardo controls the operation of the turbines in the
Eastern Scheldt Storm Surge Barrier based on the
rotational speed of the turbine. The rotational speed of the
turbine will be tuned in such a way that the maximum
amount of energy can be subtracted from the turbine. This
means that the ADCPs are not used to control the turbine
operation.
In the CFD model, the rotational speed of the turbines is
used as an input parameter. Together with the torque of
the turbine, the power is calculated via the formula: P = τω,
in which τ is the torque (Nm) and ω is the angular velocity
[rad/s]. The generated power is compared with the
measured power from Tocardo. The measured power is
the power that has been delivered to the rotor after
correction for the system losses. Tocardo provided a
conversion relation between the grid power and the power
on the rotor. Due to energy losses to the system, the power
on the rotor is approximately 10% higher.
In Table 7, the obtained power by the CFD model is
presented. The deviation on the mean velocity for the
higher power output cases is below 10%. It has to be noted
that the standard deviation for the modelling results is in
the order of 5-10% of the mean power. For the
measurements, the standard deviation of the power is in
the order of 10-20%, with lower standard deviations for the
cases with the larger head difference. Based on this
analysis, it is concluded that the CFD model represents the
measured power well.
L. Turbine thrust
The thrust, which is obtained from the CFD model is
presented in Table 8. The differences with the measured
values is approximately 10 kN. For the case with higher
forcing on the turbines WT.4, the relative difference is
maximum 5 % of the measured thrust. Similar to the
power, the standard deviation of the modelling results is
5 %, while the standard deviation of the measurements is
in the order of 10-30 %. Where the cases with the lower
head difference show the larger standard deviations. From
this analysis, it is concluded that the CFD model represents
the measured thrust sufficiently accurately.
V. DISCUSSION
M. Validation of the model
The CFD model has been validated against field
measurements taken from ADCPs mounted at the barrier
during operation. The agreement in velocity at the ADCP
profile locations is good for the cases and head differences
studied. The agreement is better for the ADCP
measurements taken in the situation without turbines. For
TABLE 7
SUMMARY OF THE SIMULATED POWER OF THE TURBINES (MEASURED IN
BRACKETS)
Tide Case T1
[kW] T2
[kW] T3
[kW] T4
[kW] T5
[kW]
Ebb
WT.1 9.1
(17.2) 9.8
(16.1) 10.6
(14.6) 10.7
(15.6) 12.0
(13.2)
WT.2 19.8
(36.4) 20.8
(34.9) 22.2
(33.5) 22.5
(34.8) 24.8
(33.0)
Flood
WT.3 49.0
(50.5) 48.4
(51.1) 48.6
(48.1) 48.4
(50.2) 48.5
(51.1)
WT.4 218.2
(220.0) 201.5
(223.3) 197.9
(214.1) 199.4
(217.4) 208.8
(222.0)
TABLE 8
SUMMARY OF THE SIMULATED THRUST OF THE TURBINES (MEASURED IN
BRACKETS)
Tide Case T1
[kN] T2
[kN] T3
[kN] T4
[kN] T5
[kN]
Ebb
WT.1 - 12.2 - 12.8 - 13.4
(- 20.0) - 13.5 - 14.4
(- 25.4)
WT.2 - 23.1 - 23.9 - 24.8
(- 34.0) - 25.1 - 26.5
(- 40.1)
Flood
WT.3 38.5 38.6 38.5
(43.2) 38.6 38.5
(45.4)
WT.4 114.6 111.4 110.6
(105.0) 110.9 113.5
(117.7)
Fig. 11 Mean velocity for a vertical cross section between the
turbines for flood case 4 without turbines (upper figure) and with
turbines (lower figure). Only a detail around the sill is shown.
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
the situation with turbines the inflow velocities are slightly
underpredicted. This could indicate that the discharges are
also slightly underpredicted in the case with turbines.
However, it has to be noted that the field measurements of
the situation with turbines show about 5-10 % higher
velocities than for the situation without turbines. This
estimate is made by comparing the velocities measured at
the sill without turbines, but at the height of the turbine
axis, with the subsequent measurements when the
turbines are installed. Similarly, the details of some of the
profiles, the size of the recirculation zone and the size of
the wake recovery zone, show some discrepancies but for
the bulk parameter of a discharge coefficient this is of less
importance. Also, in general the good agreement between
the measurements and the CFD give confidence that the
quality of the predicted flow field is high.
The predictions of the power and thrust from the
turbines are validated as well. For both parameters, the
standard deviation of the measurements is significant
(between 10 % and 30 %). The agreement of the CFD
model with the measurements at the high flow rate (WT.4)
is quite good (<10 %).
N. Inclusion of rotating zones around the turbines
The largest uncertainty in the results of the CFD
simulations is created by the rotating zones around the
turbines. The earlier validation [2] of the CFD model with
turbines was performed with a larger rotating zone around
the turbines than was possible to apply in these
simulations. A larger zone is generally preferable because
it allows the interpolation between rotating and stationary
zones to be made at a location where the gradients are
small and the numerical errors will similarly be smaller.
For the simulations of 5 turbines in a row in the opening of
the barrier it is not possible to create a larger zone.
Similarly, for the flood case, the location of the rotating
zone in a region of very high gradients downstream of the
flow separation at the sill, makes the numerical errors of
interpolation greater. The expected effect of this is to
increase numerical diffusion which is a form of extra
resistance in simulations which are not fully resolved.
The inclusion of the rotating turbines in the simulations
is state-of-the-art (see discussion of literature in Section
1.1). The validation performed here with the turbines
shows that this model is possible of generating validated
mean profiles whilst still resolving a large proportion of
the unsteady turbulence generated by the turbines and
thereby giving insights into the true effect of the turbines
on the downstream flow.
O. Biofouling of the Eastern Scheldt Barrier
During installation of the horizontal and vertical
ADCP’s in 2011 by Partrac, biofouling was observed on the
sill. A layer of approximately 10cm of mussels had to be
removed from the sill before installation of the ADCP
equipment. This layer of biofouling has not been included
in the CFD simulations, while it may have an effect on the
details of flow through the barrier.
Some simulations with different roughness were made
during the validation process of the model but adding a
roughness as much as 10 cm is not possible in the current
setup because in the model currently used the roughness
is limited to the size of the first cell near the wall. An
investigation of other methods to include roughness in the
model was not carried out.
P. Dynamic effects in the Eastern Scheldt
From the measurements it became clear that due to
inertia, the velocities are much higher at the end of the tidal
phase than at the beginning of the tidal phase. This effect
is not included in the CFD model because the water levels
were imposed as constant boundary conditions. The CFD
simulations also do not include the oblique approach flow
which is present at the start of the tidal phase for the ebb
cases. This oblique flow is the result of the turning of the
tide in the local geometry of the Eastern Scheldt. This
oblique approach flow angle could have played a role in
the validation of the power measurements.
VI. CONCLUSION
The CFD simulations performed of the Eastern Scheldt
Barrier were conducted for the purpose of the prediction
of discharge capacity and flow patterns downstream of the
barrier. With this objective in mind the following
conclusions can be drawn:
Q. Model is validated for discharge coefficients:
The agreement between the CFD and ADCP field
measurements of velocity profiles is good. The small
discrepancy in a number of details is not sufficient to result
in significant differences in the total flow through the
barrier. The agreement in the profiles is also not sensitive
to the exact location in the domain at which the profile is
drawn. This gives confidence that the comparison made is
valid for a large part of the flow field. For the cases with
turbines it appears that the inflow velocities (and
consequently discharges) are slightly underpredicted.
However, it is unclear if this is caused by the numerical
model or by the flow measurements, which have shown
that the upstream flow velocities are higher for the
situation with turbines than without turbines. It is
concluded that the developed model is suitable for an
assessment of the effect of the turbines on the discharge
coefficient.
R. Model validation for power and thrust
The discrepancy between the CFD calculated turbine
power and the measured turbine is small for high
discharges and within the ranges of uncertainty. For
simulations where the agreement in inflow velocity
profiles is better the agreement with the measured power
is also improved. This means the model should perform
well for the prediction of the optimal location of the
turbines in the flow field for power generation.
O’MAHONEY et al.: HYDRODYNAMIC IMPACT OF TIDAL TURBINES IN A STORM SURGE BARRIER
135
O’MAHONEY et al.: HYDRODYNAMIC IMPACT AND POWER PRODUCTION OF TIDAL TURBINES IN A STORM SURGE BARRIER
S. Effect of turbines on discharge coefficient
The turbines have only a small effect on the discharge
coefficients. For the ebb case the reduction in discharge
coefficient between simulations without and with turbines
is approximately 5%. For the flood cases the reduction was
negligible in both cases but this is likely within the range
of accuracy of the simulations. The small effect on the
discharge coefficient for flood is attributed to the fact that
the turbines are placed sufficiently far downstream of the
sill. During ebb, the turbines are upstream of the barrier,
which causes a reduced discharge coefficient.
The relative small effects on the discharge coefficients of
the barrier gate in question imply that the effects of the
tidal array on the environment (currents, tidal ranges, tidal
flats) of the Eastern Scheldt estuary will not be large. The
work on the environmental effects will be presented
separately.
ACKNOWLEDGEMENT
The work is part of the Dutch Marine Energy Centre
(DMEC) project funded by the European Regional
Development Fund (ERDF) and co-financed by the
Province of Noord-Holland. Tocardo Tidal Power BV is
gratefully acknowledged for the permission to publish the
results of the ADCP, thrust, power and RPM
measurements.
REFERENCES
[1] Rijkswaterstaat, "Design Plan Oosterschelde Storm Surge
Barrier: Overall design and design philosophy," Rijkswaterstaat,
Rotterdam, The Netherlands1994.
[2] A. Tralli, A. Bijlsma, W. te Velde, and P. de Haas, "CFD Study
on Free-Surface Influence on Tidal Turbines in Hydraulic
Structures," Proc. of the ASME 34th International Conference on
Ocean, Offshore and Artic Engineering,no. OMAE2015-41187,
2015.
[3] A. Bijlsma, A. Tralli, W. Verbruggen, and P. De Haas, "Detailed
dynamics of the Ea stern Scheldt storm surge barrier: validation of
a CFD approach," Proc. of the 4th International Symposium of
Shallow Flows, 2017.
[4] W. M. J. Batten, A. S. Bahaj, A. F. Molland, and J. R. Chaplin,
"The prediction of the hydrodynamics performance of marine
current turbines," Renewable Energy, vol. 33, pp. 1085-1096,
2008.
[5] S. R. Turnock, A. B. Phillips, J. Banks, and R. Nicholls-Lee,
"Modelling tidal current turbine wakes using a coupled RANS-
BEMT approach as a tool for analysing power capture of arrays
of turbines," Ocean Engineering, vol. 38, pp. 1300-1307, 2011.
[6] C. H. Jo, J. Y. Yim, K. H. Lee, and Y. H. Rho, "Performance of
horizontal axis todal current turbine by blade configuration,"
Renewable Energy, vol. 42, pp. 195-206, 2012.
[7] S. Gant and T. Stallard, "Modelling a tidal turbine in unsteady
flow," Proc. 18th Int. Offshore and Polar Engineering
Conference, pp. 473-480, 2008.
[8] W. M. J. Batten, M. E. Harrison, and A. S. Bahaj, "Accuracy of
the actuator disc-RANS approach for predicting the performance
and wake of tidal turbines," Phil Trans R Soc A, vol. 371, no.
20120293, 2013.
[9] L. Myers and A. S. Bahaj, "Power output performance
characteristics of a horizontal axis marine current turbine,"
Renewable Energy, vol. 31, pp. 197-208, 2005.
[10] L. Myers and A. S. Bahaj, "Wake studies of a 1/30th scale
horizontal axis marine current turbine," Ocean Engineering, vol.
34, pp. 758-762, 2007.
[11] X. Sun, J. P. Chick, and I. G. Bryden, "Laboratory-scale
simulation of energy extraction from tidal currents," Renewable
Energy, vol. 33, pp. 1267-1274, 2008.
[12] L. Myers and A. S. Bahaj, "Experimental analysis of the flow
field around horizontal axis tidal turbines by use of scale mesh
disk rotor simulators," Ocean Engineering,vol. 37, no. 218-227,
2010.
[13] M. E. Harrison, W. M. J. Batten, L. Myers, and A. S. Bahaj, "A
comparison between CFD simulations and experiments for
predicting the far wake of horizontal axis tidal turbines," Proc.
8th European Wave and Tidal Energy Conference (EWTEC),
2010.
[14] M. J . Laws on, Y. Li, and D. Sale, "Devel opment a nd Verifi catio n
of a Computational Fluid Dynamics Model of a Horizontal Axis
Tidal Current Turbine," Proc. 30th ASME Int. Conf. on Ocean,
Offshore and Arctic Engineering, no. OMAE2011-49863, 2011.
[15] W. Otto, D. Rijpkema, and G. Vaz, "Viscous Flow Calculations
on an Axial Marine Current Turbine," Proc. 31th ASME Int. Conf.
on Ocean, Offshore and Arctic Engineering, no. OMAE2012-
83452, 2012.
[16] A. Mason-Jones et al., "Non-dimensional scaling of tidal stream
turbines," Energy, pp. 820-829, 2012.
[17] C. Frost, C. E. Morris, A. Mason-Jones, D. M. O'Doherty, and T.
O'Doherty, "The effect of tidal flow directionality on tidal turbine
performance," Renewable Energy,vol. 78, pp. 609-620, 2015.
[18] A. S. Bahaj, W. M. J. Batten, and G. McCann, "Experimental
verification of numerical predictions for the hydrodynamic
performance of horizontal axis marine current turbines,"
Renewable Energy, vol. 32, pp. 2479-2490, 2007.
[19] A. S. Bahaj, A. F. Molland, J. R. Chaplin, and W. M. J. Batten,
"Power and Thrust measurements of marine current turbines
under various hydrodynamic flow conditions in a cavitation
tunnel and towing tank," Renewable Energy, vol. 32, no. 3, pp.
407-426, 2007.
[20] J. Whelan, J. M. R. Graham, and J. Peiro, "A free-surface and
blockage correction for tidel turbines," Journal of Fluid
Mechanics, vol. 624, pp. 281-291, 2009.
[21] C. A. Consul, R. H. J. Willden, and S. C. McIntosh, "Blockage
effects on the hydrodynamic performance of a marine cross-flow
turbine," Phil Trans R Soc A, vol. 371, no. 20120299, 2013.
[22] T. Blackmore, W. M. J. Batten, and A. S. Bahaj, "Influence of
turbulence on the wake of a marine current turbine simulator,"
Phil Trans R Soc A, vol. 470, no. 20140331, 2014.
[23] T. Blackmore, L. Myers, and A. S. Bahaj, "Effects of turbulence
on tidal turbines: Implications to performance, blade loads and
condition monitoring," International Journal of Marine Energy,
vol. 14, pp. 1-26, 2016.
[24] M. J. Churchfield, Y. Li, and P. J. Moriaty, "A large-eddy
simulati on study of wake propagation and power pr oduction in an
array," Phil Trans R Soc A, vol. 371, no. 20120421, 2013.
[25] T. Daly, L. Myers, and A. S. Bahaj, "Modelling of the flow field
surrounding tidal turbine arrays for varying positions in a
channel," Phil Trans R Soc A, vol. 371, no. 20120246, 2013.
[26] T. Stallard, R. Collings, T. Feng, and J. Whelan, "Interactions
between tidal turbine wakes: experimental study of a group of
three-bladed rotor s," Phil Trans R Soc A, vol. 371, no. 20120159,
2013.
[27] A. Olczak, T. Stallard, T. Feng, and P. K. Stansby, "Comparison
of a RANS blade element model for tidal turbine arrays with
laboratory scale measurements of wake velocity and rotor thrust,"
Journal of Fluids and Structures, vol. 64, pp. 87-106, 2016.
[28] T. O'Doherty, A. Mason-Jones, D. M. O'Doherty, P. S. Evans, C.
Wooldridge, and I. Fryett, "Considerations of a horizontal axis
tidal tu rbine," Proc. Inst of Civil Engineers. Energy, vol. 163, no.
EN3, pp. 119-130, 2010.
[29] U. Ahmed, I. Afgan, D. D. Apsley, T. Stallard, and P. K. Stansby,
"CFD simulations of a full-scale tidal turbine: comparison of LES
and RANS with field data," Proc. 11th European Wave and Tidal
Energy Conference (EWTEC), 2015.
[30] M. Verbeek, R. J. Labeur, and W. Uijttewaal, "Field observations
of rotor thrust and bypass-acceleration at a tidal fence," presented
at the 6th Oxford Tidal Energy Workshop, Oxford, UK, 2018.
[31] W. Verbruggen, "Analysis ADCP data Eastern Scheldt Barrier
with and without turbine deployment," Deltares2018.
INTERNATIONAL MARINE ENERGY JOURNAL, VOL. 3, NO. 3, NOVEMBER 2020
... The large-scale assessments focus on the main intertidal flats in the estuary. This article presents a brief description of the estuary and the pilot plant, after which the impacts are addressed based on the measurements from the monitoring campaign and the numerical models which were validated against these measurements [18]. At the end of this article the potential for upscaling tidal power from the storm surge barrier is discussed based on the conclusions of the impact of this pilot plant compared to other impacts such as sea level rise. ...
... To assess the impact of the tidal turbines on the environment, a CFD model (as described in Ref. [18]) was made extending 200 m at both sides of the barrier and consisting of 2 gates (Gate 8 and half the adjacent Gate 7 and Gate 9). The geometry of the storm surge barrier and the turbines including turbine blades were reproduced in detail. ...
... The bathymetry was based on a multibeam survey (0.5 m resolution) and converted to a resolution of 2 m in the model. Simulations were carried out for situations with and without turbines and the results were validated against the velocities measured by the ADCP measurement campaigns of both 2011 without turbines and 2016 with turbines installed (see Ref. [18]). In the simulations with turbines, the turbines were represented by means of an overset mesh, in which the actual rotation of the turbine blades was modelled. ...
Article
Hydraulic structures can be a promising place for tidal energy extraction due to the high flow velocities, easy access to the power grid and easy access for maintenance. However, quantification of the impacts of a tidal power plant in a hydraulic structure is not straight forward. In 2015 a pilot plant consisting of an array of five Tocardo tidal turbines was installed in the Eastern Scheldt Storm Surge Barrier in the Netherlands. This pilot was accompanied by monitoring studies to verify that the operation of the plant had no adverse impact on the barrier and its surroundings. This paper presents the assessment of the hydraulic impact of the tidal power plant in the storm surge barrier based on an analysis of water level and current measurements, combined with numerical modeling and followed by an assessment of the environmental impact with emphasis on the effects on the intertidal areas in the estuary. This validation approach by a pilot plant is imperative to understand the interaction between tidal turbines and the hydraulic structure on the local scale. This understanding gives extra credibility to the predictions of the extrapolated large-scale and large array assessments which will always be fully numerical.
... 1. Introduction. Barrier simulations are difficult due to the barriers' inherently thin structure [23,19]. This means that more resolution is required at the barriers. ...
Preprint
We apply the cell merging method to a model shallow water problem with a permeable boundary. We use a cut cell approach which is more easily and systematically scalable with different shapes of boundaries. The novel cell merging method presented in this paper uses both wave propagation algorithm and gradient reconstruction for second order corrections, along with minmod and Barth-Jespersen limiters. We observe second order convergence in two model problems, one with a linear boundary and other with a composite, V-shaped boundary. We assess the effectiveness of these two boundaries by doing a realistic scenario including an island and observing inundation at the peak of the island.
Article
Full-text available
The tidal flow between bridge pillars and through open barriers is a promising source of ocean energy which can be exploited using tidal stream turbines, as proven recently by operational demonstration plants. The aim of this study is to clarify the consequences for the power output of tidal turbines when placing them in a hydraulic structure. To this end, experimental measurements of turbine power and wakes are performed, using a down-scaled turbine mounted at a submerged weir. The results are compared to an analytical model, validating its range of application for optimising turbine-weir geometries. The experimental data show that the power coefficient of the turbine can be increased by optimising the blockage of the channel and the distance between the turbine and the structure, which is related to the wake configuration. In this way, the power coefficient increased by 40 percent when the turbine was re-positioned from the upstream to the downstream end of the structure. The theoretical model could reproduce the measured power within 10 percent accuracy, proving its value as a rapid assessment tool. As such, this work advances the knowledge needed to meet targets on the transition towards renewable energy.
Conference Paper
Full-text available
A Computational Fluid Dynamics (CFD) model of the detailed hydrodynamics of a gate of the Eastern Scheldt storm surge barrier was build and validated against literature and ADCP data. Using the CFD code StarCCM+, a two-phase air-water system was simulated via the Volume of Fluid (VOF) method, on a highly detailed geometry featuring a realistic bathymetry profile and gate structure. Unique horizontal and vertical up-looking ADCP measurements were available for the existing condition. For quasi stationary instants in a tidal cycle comparisons were performed against transient Reynolds-Averaged Navier-Stokes (RANS) computations with steady boundary conditions. Compared data included the head loss across the barrier, and the detailed velocity pattern in a horizontal and a vertical plane at specific locations close to the structure. The validated model is to provide quantitative data (discharge relation, velocity patterns, turbulence intensity) on the flow through the gates. Ultimately, it will be coupled to a far-field model including the entire Eastern Scheldt estuary and adjacent coastal waters to assess the effect of the tidal turbines in the Eastern Scheldt storm surge barrier.
Conference Paper
Full-text available
CFD simulations have been performed for a geometry-resolved full-scale tidal-stream turbine and compared with experimental data from the EMEC test site in the Orkney Isles. The mesh comprises two regions: a rotating part, containing the turbine, and a stationary outer part, including the support tower. A sliding-mesh interface couples the two parts. Initially, Reynolds-averaged Navier-Stokes and large-eddy simulations were performed using an inflow velocity profile representative of the test site but low inflow turbulence, yielding satisfactory mean power coefficients. LES with synthetic turbulence prescribed at inlet was then employed to try to predict realistic load fluctuations. Load fluctuations (power, thrust and blade bending moments) may arise from onset mean velocity shear, influence of the support tower, blade-generated turbulence, approach-flow turbulence and waves. Inflow statistics were prescribed to match the vertical distribution of mean velocity, Reynolds stresses and length scales determined from a channel-flow simulation, with additional factoring of stresses and length scales to match as far as possible those measured on-site. LES simulations with synthetic turbulence at inflow satisfactorily reproduces the spectral distribution of blade bending moments provided that spectra are normalised by variance to reflect the relatively small number of rotations computed.
Article
Full-text available
A Reynolds averaged Navier Stokes (RANS) model has been compared with laboratory scale measurements of the loading and wake of arrays of horizontal axis three-bladed rotors. The diameter D of each rotor was 0.27 m and axis was at mid depth in a channel of depth and width . Mean flow speed was 0.46 m/s, giving a Reynolds number of , with turbulence intensity of 12%. The commercial software StarCCM+ is employed. Steady flow is modelled and each rotor represented by imposing radial variation of streamwise force on the rotor defined by a blade element method. When experimental values of turbulent kinetic energy and dissipation rate are defined at the model inflow, the transverse profile of depth-averaged velocity of a single rotor wake is predicted reasonably for downstream distances greater than . Within to downstream of the rotor plane the wake is approximately axisymmetric and both streamwise velocity and wake rotation are reasonably well predicted for radii greater than half the rotor radius. Inclusion of tip generated turbulent kinetic energy as predicted by blade resolved RANS CFD improves predictions of streamwise velocity and wake rotation within downstream. For a single row of rotors partially spanning the channel the increase of thrust coefficient relative to an isolated rotor due to global and local blockage is predicted, but the overall magnitude of thrust is overpredicted. Downstream of a single row the velocity deficit along the axis of each rotor is closely predicted while the deficit between adjacent wakes is underpredicted until approximately downstream. For a second row of rotors located at downstream the square of the rotor averaged velocity, indicative of thrust, is predicted to within 3% and 12% respectively for rotor axes aligned and staggered relative to those on the front row. For these downstream rotor positions similar accuracy may be obtained by superposition of the self-similar velocity deficit profile of isolated rotor wakes only. For multiple rows, thrust on the front row of rotors is increased relative to the same row in isolation. Rotors on downstream rows partly exposed to the higher velocity flow bypassing the array also experience high values of thrust. For arrays of up to twelve rotors the average thrust of an individual turbine in a row is typically predicted to within 10% for the first row and 20% for the second row and within 38% for the inner rotors of the third row. The accuracy of a RANS blade element method using commercially available software has thus been assessed for the complex wake flows of arrays of up to twelve three-bladed rotors in a shallow turbulent flow.
Article
Full-text available
With many Tidal Energy Conversion (TEC) devices at full scale prototype stage there are two distinct design groups for Horizontal Axis Tidal Turbines (HATTs). Devices with a yaw mechanism allowing the turbine to always face into the flow, and devices with blades that can rotate through 180° to harness a strongly bi-directional flow. As marine turbine technology verges on the realm of economic viability this paper reveals the performance of Cardiff University's concept tidal turbine with its support structure either upstream or downstream and with various proximities between the rotating plane of the turbine and its support stanchion. Through the use of validated Computational Fluid Dynamics (CFD) modelling this work shows the optimal proximity between rotor plane and stanchion as well as establishing, in the given context, the use of a yaw mechanism to be superior to a bi-directional system from a performance perspective.
Article
Full-text available
Marine current turbine commercial prototypes have now been deployed and arrays of multiple turbines under design. The tidal flows in which they operate are highly turbulent, but the characteristics of the inflow turbulence have not being considered in present design methods. This work considers the effects of inflow turbulence on the wake behind an actuator disc representation of a marine current turbine. Different turbulence intensities and integral length scales were generated in a large eddy simulation using a gridInlet, which produces turbulence from a grid pattern on the inlet boundary. The results highlight the significance of turbulence on the wake profile, with a different flow regime occurring for the zero turbulence case. Increasing the turbulence intensity reduced the velocity deficit and shifted the maximum deficit closer to the turbine. Increasing the integral length scale increased the velocity deficit close to the turbine due to an increased production of turbulent energy. However, the wake recovery was increased due to the higher rate of turbulent mixing causing the wake to expand. The implication of this work is that marine current turbine arrays could be further optimized, increasing the energy yield of the array when the site-specific turbulence characteristics are considered.
Article
Full-text available
An improved method is developed to couple an inner domain solution of the blade element momentum theory with an outer domain solution of the Reynolds averaged Navier Stokes equations for evaluating performance of tidal current turbines. A mesh sensitivity study shows that a mesh of at least 6M cells with at least 40% of these within the turbine wake is required to ensure satisfactory convergence of the velocity deficit. In addition to the usually applied axial momentum source terms, angular momentum and turbulence intensity source terms are shown to be required to model the near wake evolution. Three different lateral turbine spacing of 2, 4 and 6 turbine diameters are used to demonstrate the influence of the effective channel blockage on the velocity distribution in the turbine bypass region, the rate of spread of the wake and the recovery of velocity distribution. A final study shows that for a fixed number of turbines minimising the lateral spacing within each row, with a small number of staggered rows spaced as longitudinally as far apart as practical, is the most effective strategy for energy capture.
Article
Laboratory scale testing of tidal turbines has generated valuable datasets to support optimised turbine design and numerical model validation. However, tidal sites are highly turbulent with a broad range of length scales and turbulence intensities that are site specific. In this work we describe an experimental campaign using static grids to generate turbulence and investigate its impact on a model tidal turbine in a circulating water flume. Length scales, energy spectra and turbulence dissipation rates are first considered for centre point measurements before full flow characterisation of the ambient conditions across the turbine rotor area. Six different cases were chosen to observe the performance of a 1/20th scale 0.8 m diameter turbine subjected to these flows. The rotor thrust and torque, and flapwise and edgewise blade root bending moments were measured. It was found that the thrust and power coefficients were sensitive to the estimate of ambient velocity. In the most extreme case the Betz limit could be ‘exceeded’ depending on which estimate of ambient velocity was used. Overall variations in the peak power coefficient of over 10% were observed, demonstrating the significance turbulence has on turbine performance. It was also found that there is a strong correlation between fluctuations in blade root bending moments and the rotor loads. As a result we proposed that fatigue loads acting on the blades may be estimated from the fluctuations in power output of the turbine. Therefore maintenance operations maybe optimised from real-time fatigue monitoring of blade loads without the need to install additional instrumentation on the turbine blades. Under this proposed regime the cost of energy will be reduced due to reductions in turbine costs and following optimisation of the maintenance requirements and operational costs. This could also improve turbine reliability which would have significant implications for large multi turbine arrays.
Conference Paper
In this paper, the flow over a marine current turbine is studied. As a test case, the benchmark turbine published in [1, 2] is selected. A bibliography review shows a variety of numerical methods applied to this specific turbine, of which a viscous-flow RANS approach seems to be the best suitable for simulations over a broad range of inflow conditions. Therefore, MARIN’s RANS solver ReFRESCO is used to study the flow over this turbine. ReFRESCO results show a good agreement with the experiments, the calculated results and associated uncertainties overlapping the model-tests results. A numerical procedure is followed to estimate these calculation uncertainties, including an estimation for the numerical, domain and geometrical uncertainties. The flow-field analysis reveals significant viscous effects. Large separation zones at the suction side of the blade are seen in the model-scale results. At model scale, the turbulence level indicates that the turbine is operating in the transitional regime between laminar and turbulent flow, leading to early flow separation. Calculations at full scale show a large scale effect. The separation zones present at model scale are significantly smaller at full scale, resulting in a higher power production and axial loading. This is explained by the fully-turbulent boundary layer.
Article
The west and south coastal region of Korea has very strong tidal current speeds and therefore accommodates many suitable sites for the application of TCP (Tidal Current Power). The maximum current speed recorded in the south is up to 6.5 m/s. Unlike other renewable energy sources, TCP is an extremely reliable, predictable and continuous energy source as the current pattern and speed can be predicted throughout the year. One of the essential components in a TCP device is the rotor converting the inflow current into the rotational energy. The design optimization of the rotor is very important to maximize the power production. The performance of the rotor can be determined by various parameters including the number of blades, shape, sectional size, hub, diameters etc.The blade of the rotor is one of the essential components which can convert tidal current energy into rotational energy to generate electricity. The variable blade properties determine the performance, efficiency and stability of the turbine system. This paper presents the design procedure for a 300 kW tidal current turbine blade. The HAT turbine model was designed based on the wind mill turbine design principles together with known turbine theories. To verify the compatibility of the turbine design method and to analyze the properties of design factors, the 3D CFD model was applied with the ANSYS CFX program. The characteristics and performances of the blades can be applied in the design of 300 KW and larger capacity TCP rotors.