Content uploaded by Maxime Thiébaut
Author content
All content in this area was uploaded by Maxime Thiébaut on Nov 16, 2020
Content may be subject to copyright.
1 INTRODUCTION
In recent years, tidal energy sector has been growing
rapidly in interest as more countries look for ways to
generate electricity without relying on fossil fuels. In
comparison to other sources of renewable energy,
tidal-stream energy has many perceived benefits, in-
cluding the quality of electricity production because
of its predictability (e.g. Lewis et al. 2019) and the
social acceptance level due to a reduced visual im-
pact. The majority of TECs are designed to be locat-
ed in the bottom boundary layer, a region where the
flow experiences friction from the seabed. Due to
friction effect the current velocity profile in the low-
er layer has a non-linear shape. The 1/7th power law
approximation is typically used to characterize the
velocity profile in a flow and resource assessment at
tidal-stream energy sites (e.g. Batten et al. 2008,
Gooch et al. 2009, Thiebaut & Sentchev 2017).
However, a number of factors, including the acceler-
ation/deceleration of the tidal flow, wave-current in-
teraction, seabed forms, etc., can cause deviations
from the 1/7th power law formulation.
Two national research programs (Thymote and
Hyd2M), funded by the French Research Agency
French (ANR) and by FEM, and focusing on as-
sessment of tidal stream variability in Alderney
Race, wave-current interaction, and turbulence, pro-
vided a large quantity of data and model simulation
results in Alderney Race (e.g. Bennis et al. 2020;
Thiebaut et al., 2020a,b). In the UK, similar investi-
gation at the European Marine Energy Centre site
(EMEC) has been conducted (e.g. Brian G. Sellar et
al., Energies 2018; Osalusi et al., 2009; Hay et al.,
2013).
Towed ADCP surveys performed in Alderney
Race recently in the frame of Hyd2M program re-
vealed spatial variability of the velocity profile
shape as large as temporal variability (Sentchev et
al. 2020a, Thiebaut et al. 2019). The northern sector
of Alderney Race experiences this particularly large
variability which was found to result from the tidal
flow interaction with bathymetric constrictions.
Turbulence characterization at tidal-stream energy site in Alderney Race
A. Sentchev
Univ. Littoral Côte d’Opale, Univ. Lille, CNRS, UMR 8187, LOG, Laboratoire d’Océanologie et de
Géosciences, F-62930 Wimereux, France
M. Thiébaut
France Énergies Marines, Technopôle Brest-Iroise, 525 Avenue Alexis de Rochon, 29280 Plouzané, France
S. Guillou
Normandie Univ, UNICAEN, Laboratoire Universitaire des Sciences Appliquées de Cherbourg, LUSAC,
EA4253, F-50130 Cherbourg-Octeville, France
ABSTRACT: Alderney Race, located between the Alderney Island (UK) and Cotentin Peninsular (France), is
a site with high tidal-stream energy potential. Circulation through Alderney Race is complex, largely domi-
nated by tides with current speed exceeding 5 m/s at spring tide. Current velocity measurements from two
bottom-mounted ADCP are used to assess the variability of velocity and turbulence in the water column. Us-
ing the velocity profiles recorded by ADCPs at 2 Hz, the time variability of the flow and turbulence was
quantified with respect to tidal conditions. Turbulence metrics such as the dissipation rate (ε) of turbulence
kinetic energy and the integral lengthscale (L) were estimated through the spectral approach and used in anal-
ysis of temporal variability of turbulent motions. Turtulence intensity (I) was also estimated and compared to
scaling properties of turbulence. The mooring locations match areas with very different turbulence regime.
This difference is found to be related to large scale bathymetry features which modify the velocity profile
shape and turbulence. All turbulence parameters were found larger at the northern ADCP site where current
velocity is larger. This high level of turbulence is assumed to be caused by the current interaction with irregu-
larities of bathymetry and enhanced friction. At the southern ADCP location, over relatively smooth bathyme-
try, the variability of turbulence was found much smaller and to be depended on the velocity shear. These re-
sults highlight large spatial variability in turbulence characteristics occurring at rather short distance within
the prospective site. The results could provide advanced information to turbine developers.
It was assumed that the seabed topography, with
depth decreasing by more than 20 m over a short
distance (~1 km) and nature of sediments (Furgerot
et al., 2019), generates enhanced turbulent motions
and mixing which could modify the velocity profile
shape (e.g. Ikhennicheu, et al., 2019). Based on the da-
ta from the towed ADCP survey, a comprehensive
analysis of the tidal current structure in the whole
water column was done (Sentchev et al. 2020a).
In the present study, we decided to go further and
to investigate the flow characteristics in the lower
layer, experiencing the effect of friction. High quali-
ty velocity measurements from two moorings were
used. We perform an estimation of scaling parame-
ters of turbulence generated by bottom friction. We
attempt to relate these turbulence metrics to some
basic parameters of the tidal flow, the velocity pro-
file shape and mean shear, and we assess the tem-
poral evolution of all quantities. This allows to rec-
ognize what parameters of the mean flow are
suitable for turbulence quantification. We also assess
the spatial variability of turbulence parameters. The
results are assumed to be useful for tidal stream tur-
bine developers.
2 DATA AND METHODS OF ANALYSIS
2.1 Study site and velocity measurements
Velocity measurements were performed in the east-
ern (French) sector of Alderney Race, northwest of
the Cotentin Peninsula (Fig. 1). The water depth is
less than 40 m in the majority of the domain with an
abrupt increase of depth, to more than 60 m, in the
northern part (Fig. 1). This large fall of water depth
is the major feature of local bathymetry. Another in-
teresting feature, closely associated to the first one,
is the spatial variation in nature of the seabed, which
has been documented in detail by Furgerot et al.
(2019). The sea surface height (SSH) exhibits a wide
range of variations (2.5-7 m) predominantly semi-
diurnal and globally symmetric. Current velocities,
on average, are slightly higher during ebb tide than
during flood tide. However the largest current speed
location varies with respect to tidal conditions. Peak
flood and ebb tidal velocities occur at high and low
water respectively, and the current reversal occurs at
mid-tide. Thus, the tidal current dynamics in Alder-
ney Race is an example of the progressive wave sys-
tem. Depth-averaged peak ebb flow field from the
local model simulations is given in Figure 1 (see
Bailly du Bois et al. 2012 for details of the model).
The maximum current speed is observed at a dis-
tance of ~5 km west of the Cotentin coast, in the ar-
ea where the depth is less than 30 m.
Static point velocity data were collected at two
locations (Fig. 1) matching very different hydrody-
namic conditions and seabed topography. At the
southern location (site A), the seabed topography is
relatively smooth with a majority of cobbles and
blocks and possible corestones (Furgerot et al.,
2019). Peak tidal current velocity, after depth aver-
aging, varies in the range 1.5-2.8 m/s with respect to
tidal conditions. The velocity measurements were
performed during four months, starting from March
2018, with an upward-looking 500 kHz broadband
Teledyne ADCP V50. The data acquired during two
days (28-29 June 2018) were used in this study. This
was spring tide period characterized by largest cur-
rent velocities and also calm weather conditions: the
significant wave height was less than 0.7 m. ADCP
was mounted on the OPENHYDRO experimental
platform sitting motionlessly on the seabed with
depth 34 m at low tide. The profiling range was from
2 to 29 m above the seabed with 1-m vertical spac-
ing. Velocity values in the surface 5 m thick layer
were not considered in analysis. Velocity profiles at
one-hour resolution were derived from ADCP meas-
urements performed at 2 Hz during 20-minute bursts
within each hour.
Figure 1. Depth-averaged current velocity field in Alderney
Race on ebb tide from model results (Bailly du Bois et al.
2012). Triangles show the location of bottom-mounted ADCPs.
Geographic names used in the text are also shown. Bathymetry
contours (20, 30, 40, 50, 60 m) are given in black. High resolu-
tion bathymetry map is given in (Furgerot et al. 2019).
At the northern location (site B), two upward-
looking RDI Workhorse 600 kHz four-beam ADCPs
were deployed simultaneously on the seafloor (31 m
mean water depth) approximately 4 km oshore.
ADCPs operated in master-slave configuration with
eight beams enabling to resolve six components of
the Reynolds stress tensor (see Thiebaut et al., 2020a
for more details). The instruments were mounted on
a specifically designed frame having the following
features: 3.3 m long, 2.5 m wide and 2 100 kg
weight. The ADCPs recorded velocities at 2 Hz with
1.3 m vertical resolution (bin size), starting 2.2 m
from the seafloor (the first bin). The measurements
lasted 38 days, from September 27 to November 03,
2017. Only the data recorded between November 01,
2017 - 01:45 UTC and November 02, 2017 - 13:45
UTC were selected for subsequent analysis of turbu-
lence using the spectral approach. During this 36-
hour long period, the depth averaged flow speed
reached 3.5 m/s and the wave forcing was the weak-
est: waves with significant wave height 0.6 m and
period 6 s were observed. It was found that during
this calm period, the contribution of waves to veloci-
ty variance became insignificant at depth greater
than 10 m (Thiebaut et al., 2020a). For this reason,
only velocities in the range from 2 to 20 m above
bottom (mab) were considered.
The most prominent feature of this measurement
site is large bathymetry gradients observed north-
ward. The seabed decreases rapidly from 30 m to 60
m over a short distance, less than 2 km. Coarse sed-
iments (cobbles, boulders, dropstones) cover the
seabed at the site (Furgerot et al., 2019) and create
larger roughness. As documented in recent studies
(e.g. Bailly de Bois et al. 2012, Sentchev et al.
2020a), the La Hague deep (Fig. 1) plays an im-
portant role in local hydrodynamics by constraining
the current direction and generating large turbulence
which affects the velocity distribution throughout
the water column.
2.2 Data analysis and turbulence characterization
Horizontal velocity components recorded by ADCP
were projected on along- and cross-stream axes (x
and y respectively) of the tidal current ellipse. The
projection angle (20° clockwise at site A and 25° at
site B) matches the orientation of the tidal current el-
lipse with respect to North in the lower layer of the
water column 20-m thick. Time series of the
streamwise velocity component, referred to as Ux,
and spanwise velocity (Uy component), were thus
generated for further analysis.
Standard statistical parameters were estimated us-
ing the velocity time series provided by ADCP: the
time mean, the maximum and the standard deviation
of velocity variations. The turbulence intensity, of-
ten referred to as turbulence level, is defined as:
,
22
U
IN
−
=
where U is the mean flow speed estimated from
three mean velocity components (streamwise Ux,
spanwise Uy and upward Uz) as
222 zyx UUUU ++
, and
( )
222
3
1zyx
++
is the standard deviation from the mean velocity, and
2
N
is the variance of velocity measurements induced
by Doppler noise. The latter was estimated from the
power spectra density (PSD) of velocity components
at different depth levels according to (McMillan et
al. 2017).
The dissipation rate, , of the turbulent kinetic en-
ergy was also estimated from the PSD of velocity,
E(k), assuming the Kolmogorov relationship of the
local isotropic turbulence (Pope, 2000):
,
where C is the Kolmogorov’s constant (C = 1.5) and
k is the wavenumber. Using Taylor’s assumption of
frozen turbulence, the frequency f and wavenumber
k can be related to the mean velocity U such as:
k = 2πf /U. Thus, the dissipation rate can be estimat-
ed from the power spectrum as (Thomson et al.
2012):
where accounts for the height of the PSD slope
which best fits the spectrum in the inertial subrange.
The value of is used to estimate another im-
portant scaling property: the integral lengthscale L,
thought as the size of the most energetic turbulent
eddies, defined by (Pope, 2000):
To characterize the mean tidal flow, one hour av-
eraged velocity profiles in the lower layer of the wa-
ter column (2 – 20.4 mab) were approximated by a
power law
,
with bed roughness coefficient
, vertically averaged
velocity
, and the water column height h (21 m).
The power coefficient
is defined empirically and
can strongly depend on the seabed roughness and
velocity range. Bed roughness coefficient
is relat-
ed directly to grain size of seabed sediments and can
vary in a large range. This formulation, proposed by
Soulsby (1997) on the basis of experiments in flume
tank and in situ measurements, is adopted by the tid-
al energy community for characterizing the techni-
cally exploitable resource at prospective sites (e.g.
Myers and Bahaj 2009, Lewis et al. 2017, Thiebaut
& Sentchev, 2017). The fitting of velocity profiles to
the power law is performed in a least-squares sense
after linearizing the equation by log function. The
resulting power law coefficient
provides the min-
imum approximation error for a given value of
,
which was also varied iteratively in the range 0.05 to
0.5.
3 RESULTS
The tidal stream and turbulence were assessed first
using the data from the southern ADCP mooring
(site A) and the results are compared to that from the
mooring at site B, where the flow conditions are
rougher. Periods used in comparison (28-29 June,
2018 and 1-2 Nov, 2017) are characterized by low
wave forcing thus allowing an accurate estimation of
the scaling properties of the bottom friction generat-
ed turbulence. Two-dimensional spectrum of current
velocities on peak flood flow (Fig. 2) reveals that the
contribution of waves to energy spectrum becomes
visible for frequencies less than 0.1 Hz. Above this
frequency, the level of velocity variations in the up-
per layer is notably lower than in the lower half of
the water column. This confirms a limited contribu-
tion of orbital velocities of wind waves to the overall
velocity variance. Time series of significant wave
heights (Hs) also demonstrated low values of Hs at
periods lower than 10 s for this period of observa-
tions. Thus the frequency 0.1 Hz is adopted as the
lower limit of the inertial subrange spanning up to
0.5 Hz.
Figure 3 shows PSD distribution of the stream-
wise velocity component Ux on flood and ebb tide at
different depth levels. Spectral analysis of velocity
time series clearly demonstrates a locally isotropic
turbulence in a range of depths from 5 to 16 mab.
The spectral slope of PSD distribution varies from -
1.65 to -1.70 at these depth levels (Fig. 3), and the
ratio of standard deviations
XY
/
and
XZ
/
are
0.96 and 0.30 respectively. Despite large difference
in ratios, the mean spectral power level for all three
velocity components appeared very close in the iner-
tial subrange, thus supporting the hypothesis of iso-
tropic turbulence. The turbulent kinetic energy of the
flow is cascading from larger eddies of size L to
smaller eddies of size less than one mm and then
dissipates into heat. It appears convenient to perform
estimation of the major turbulence metrics within
this layer (5- 16 mab) – also corresponding to the
tidal turbines’ hub and rotor location (e.g. Zhou et
al. 2014). More upwards, from 16 to 20 mab, the
spectral slope decreases to -1.46 making the estima-
tion of the scaling properties of turbulence less reli-
able.
Figure 2. Power spectral density distribution at different depth
levels at site A during peak flood tide on 28-Jun-2018.
Figure 3. Power Spectral Density (PSD) of the current velocity
(streamwise component) at depth levels 10 and 16 mab at site
A on 28-Jun-2018. Velocity spectra during peak flood tide (FT)
are shown by gray and black, and during ebb tide (ET) by light
and dark blue. Red dashed line shows the spectral slope
in the inertial subrange.
Velocity profiles during the peak ebb and flood
flows (Ux component) look different (Fig. 4). The
major difference is observed in the lower layer
where the profile appears more sheared on ebb tide
and homogeneous on flood tide. Profile approxima-
tion by a power law provides the exponents 1/5 and
1/7 for peak ebb of flood tide conditions respective-
ly. Figure 4 shows that the shape of profiles during
two other hours of tidal cycle (one hour preceding
and following the peak current) is similar to that of
peak current. The resulting values of the power law
exponent change insignificantly ( 1). The question
arising from this observation is how the turbulence
is affected by the difference in profile shape between
ebb and flood flow? For clarity, we restrict the anal-
ysis by considering the peak current period on ebb
and flood tide. Three tidal cycles under calm weath-
er conditions were used in analysis.
Table 1 summarizes the scaling properties of tur-
bulence for both tidal stages and also gives the tur-
bulent intensity values. Most of similarity in turbu-
lence regime is observed at 10 mab –at typical hub
height. The level of the ambient turbulence (I) is
close to 10%, the dissipation rate of TKE () is with-
in the range 2 – 2.6 x 10-3 m2s-3. The size of the en-
ergy containing eddies is also similar (~7.5 m). At
range 16 mab (i.d. above hub height), the difference
affects only two quantities – and L. The analysis
reveals larger size of eddies and lower dissipation
rate on ebb flow.
The major difference in turbulent parameters is
found below the hub height, at 5 mab. Here, the size
of energy containing eddies is four times larger on
ebb tide than on flood tide. This is the direct conse-
quences of the velocity shear which is 40% larger on
ebb tide (mean shear 0.07 s-1). The level of ambient
turbulence I is also larger on ebb tide (16%) com-
pared to I on flood tide (13%). But this difference
can be explained by lower velocity on ebbing tide
which increases artificially the ratio
/U. The inte-
gral lengthscale L appears to be relevant for charac-
terizing the change in turbulent regime of the flow.
The small size of turbulent eddies on flood flow is a
consequence of mixing which breaks down large ed-
dies. The shape of velocity profiles, characterized by
a power law exponent is another metric useful for
turbulence characterization. Therefore, sheared pro-
files correspond to a different spatial structure of
turbulence, with larger size eddies which can inter-
act with an operating tidal turbine in a different way.
To verify the relationship between the velocity
shear and turbulence, we analyzed the flow regime
and turbulent quantities at site B, located 4 km
northward. Velocity profiles at 1-hour resolution re-
veal more powerful tidal stream with velocities up to
3.4 m/s in the upper layer (Fig. 5). They also show a
significant difference in shape, already revealed for
site A: more homogeneous profile on flood flow and
sheared on ebb flow with power law exponent 1/6
and 1/4 respectively. The optimal bed roughness co-
efficient value was found to be 0.45 (corresponding
to 0.45 10-3 m). Scaling properties of turbulence are
given in Table 1 (site B). They evidence that, in the
sheared velocity layer on ebb tide, the integral
lengthscale is larger and the dissipation rate is
smaller. On the contrary, the level of ambient turbu-
lence, quantified by I is very similar (18%).
Table 1. Turbulence properties of the tidal flow dur-
ing peak flood and ebb tide conditions at sites A and
B: dissipation rate
of the turbulence kinetic energy,
integral lengthscale L, and turbulence intensity I.
Site A
______________________________________________
Dist. from bottom (m2s-3) (m) I(%)
______________________________________________
Ebb tide
____________
16 m 1.1 × 10-3 12.9 9
10 m 2.6 × 10-3 7.5 11
5 m 6.4 × 10-3 5.4 17
Flood tide
____________
16 m 1.5 × 10-3 7.0 8
10 m 2.0 × 10-3 7.8 10
5 m 12.0 × 10-3 1.5 13
_____________________________________________
Site B
______________________________________________
Ebb tide
____________
16 m 2.6 × 10-3 26.6 10
10 m 3.3 × 10-3 27.3 15
5 m 4.7 × 10-3 25.8 18
Flood tide
____________
16 m 3.4 × 10-3 30.0 14
10 m 3.9 × 10-3 26.5 15
5 m 6.0 × 10-3 22.4 18
_____________________________________________
In the upper half of the water column (at range
more than 10 mab) both the integral scale L and dis-
sipation are found larger on flood flow. They appear
to correlate with turbulence intensity, also larger on
flood tide. However the difference in I between ebb
and flood flow is found very large at range 16-20
mab (Fig. 5). This is due to larger velocity variance
on flood tide in the whole water column. We assume
that this enhanced variability of Ux is caused by sea-
bed irregularities upstream of the measurement site.
Figure 1 shows the existence of a promontory at
short distance southward of the mooring location.
Note also that temporal variability of I was fond not
large when considering individually ebb or flood
flow stage. Hourly profiles of I appear to be similar
and the range of variability is limited to 3% in the
bottom layer at site B (Fig. 5).
Comparison of turbulence metrics at two sites
demonstrates higher turbulence level at site B on av-
erage, with larger values of the dissipation rate and
the integral lengthscale L caused by the flow condi-
tions and not only by higher current speed. Large ba-
thymetry gradients, observed northward and south-
ward of the mooring location B, and enhanced
friction, result in higher level of turbulence there. At
the same time, the turbulence intensity , in the up-
per layer (10-20 mab), is similar at both sites (ebb
tide conditions at site B) and does not match the dif-
ference in turbulence characteristics between sites.
This comparison reveals that the turbulence intensity
alone does not capture quantitative changes in the
turbulent regime. Others quantities, such as dissipa-
tion rate, , the integral length-scale, , appear more
appropriate for turbulence characterization.
4 DISCUSSION
Velocity time series used in this study were acquired
as a part of two large observation programs (Thy-
mote and Hyd2M) focusing on assessment of the
tidal stream variability and turbulence in Alderney
Race. The study site is a prospective tidal power site
suitable for a massive deployment of tidal energy
convertors (TECs) there. To be economically feasi-
ble, TECs require large current velocities: typically,
spring tide velocities in excess of 2.5 m/s (Lewis et
al. 2015). This condition is met over a large area
within Alderney Race. As the majority of TECs are
designed to be located in the bottom boundary layer,
a region where the flow experiences friction from
the seabed, the knowledge of vertical shear of cur-
rent velocity is a major consideration. It is assumed
to have important implications for turbine efficiency
and resilience (e.g. Liu et al. 2012, Milne et al.
2016), for assessing the effect of turbulence (e.g.
Blackmore et al. 2016), as well as for the instantane-
ous power available, as described in (Lewis et al.
2019). For this reason, TEC developers look for de-
tailed characteristics of the flow speed across the ar-
ea swept by the turbine blades and the velocity pro-
file shape characterization is an important issue in
the turbine performance research and in the planning
stage of the projects (e.g. Tatum et al. 2016).
Figure 4. (a) Current velocity profiles, 1-hour averaged during
three successive tidal cycles, from ADCP measurements at site
A during flood tide (red) and ebb tide flow (blue). Profiles av-
eraged over 1-hour time interval of peak ebb and flood current
are shown by circles. Solid and dashed lines show average pro-
files for two other 1-hour time intervals: preceding and follow-
ing the peak current. (b) Turbulence intensity profiles for peak
ebb flow (squares) and flood flow (triangles). Profiles for peak
+1h and -1h time intervals are shown by dashed lines.
Figure 5. Same as Figure 4 but for site B.
Therefore when characterizing the velocity pro-
file of a tidal energy site in depth-averaged model
resource studies (e.g. Batten et al. 2008, Myers &
Bahaj, 2010), it appears appropriate to assume a
1/7th power law, which is widely used by oceanog-
raphers (e.g. Thiebaut & Sentchev, 2017, Lewis et
al. 2017). Our results revealed a deviation from the
1/7th power law. On ebb flow, the 1/5th power law
was obtained, matching sheared profile, whereas on
flood flow, the velocity profile is well fitted by 1/7th
power law. It is worth to mention that in the tidal
stream with more sheared profile (i.e. exponent 1/5)
the energy resource is larger by 8% compared to
flow with less sheared profile (e.g. Lewis et al.
2017). At the same time, the increase of velocity
shear generates more turbulence which in turn in-
creases unsteady loading on the blades by tidal flow
instabilities (Ouro & Stoesser 2019, Gaurier et al.
2020). This is a cause of fatigue.
While the major characteristics of the mean tidal
flow (speed, direction, current magnitude asym-
metry, etc) are relatively simple to measure (e.g.,
Guerra & Thomson 2017, Thiébaut & Sentchev
2016, 2017), much less is known about turbulence.
This is indicative of the inherent technical difficul-
ties (i.e., sensor movement, limited sampling rate) in
acquiring measurements of turbulent motions in fast
moving currents (e.g., Milne et al. 2013).
In this study we attempted to use the conventional
measurements of current velocity profiles for turbu-
lence characterization. We also searched for metrics
which are appropriate for quantification of turbu-
lence and its potential impact on TECs.
Among different measurement techniques, em-
ploying ADV, ADCP, VMP for tidal stream and tur-
bulence characterization at prospective tidal sites
(e.g. Thomson et al. 2012, Korotenko et al. 2013
McMillan et al. 2016), ADCP is by far the instru-
ment of choice. Deployed on bottom platforms, it
provides high quality velocity time series with ac-
quisition rate of 1 Hz, 2 Hz or even higher.
Spectral analysis of velocity time series allows
estimating at least four fundamental properties of the
turbulent flow: the dissipation rate of turbulence ki-
netic energy, , the integral lengthscale, , the Kol-
mogorov scale, η, and the Taylor-based Reynolds
number (Sentchev et al. 2020b, Thiebaut et al.
2020a). The integral lengthscale, , provides the
spatial structure of turbulence and quantifies the size
of the most energy containing eddies. This quantity
is useful for understanding the turbulent flow inter-
action with a single device (tidal turbine) or an array
of devices deployed on the seabed.
The dynamics of the large-scale turbulent eddies
was found sensibly different at two measurement
sites, with predominantly larger turbulent motions at
site B and scales comparable with the water depth.
These eddies contain the largest proportion of turbu-
lent energy, and are likely to have large effect on
TECs performance. Recently, a study focusing on
the assessment of the performance of a Darrieus type
turbine operating in real sea conditions demonstrated
that the strongest impact of turbulence on power
generation by the TEC occurred when the length of
the most energetic eddies matches the turbine size
(Sentchev et al., 2020b).
MacEnri et al. (2013) were the first who per-
formed a comprehensive analysis of the SeaGen per-
formance and demonstrated that standard deviation
of the velocity, assumed to be a measure of the tur-
bulence strength, is likely to be one of the significant
factors that contributes to flicker level. Large fluctu-
ations of the output power were found at frequency
~ 0.5 Hz which is fundamental (turbine) frequency.
Similar results were documented by Frost et al.
(2018) for a scaled turbine, tested in the Strangford
Lough (UK). Power fluctuations were related to tur-
bulence effect.
According to our results, it is expected that in the
northern part of Alderney Race, TECs will experi-
ence larger influence of turbulence with high level
of fatigue caused by large size eddies. Another effect
of turbulence eddies on TECs could be the enhanced
fluctuations of the output power caused by turbulent
eddies of size compared to a typical turbine size.
5 CONCLUSIONS
Analysis of ADCP velocity measurements in Al-
derney Race demonstrated large temporal variability
of the velocity profile shape, in the bottom layer ~
10-m thick, which was found correlated with the tid-
al conditions. At both locations, over relatively
smooth bathymetry in the south and rough bathyme-
try in the north, ebb tide flow appeared highly
sheared. Velocity profiles were accurately approxi-
mated at two sites by the power law with exponents
1/5 and 1/4 respectively on ebb tide, and 1/7 on
flood tide. Large velocity shear observed in the low-
er part of the water column is a source of enhanced
turbulence. A relationship between the dissipation
rate
, the size of energy containing eddies and the
velocity shear was demonstrated. Large turbulent
eddies of size comparable with the water depth were
identified in the northern sector and assumed to be
generated by the current interaction with a bathymet-
ric constriction. Turbulence intensity , is a metric
that in general does not alone capture quantitative
changes in the turbulent regime during the flow tran-
sition from ebbing to flooding flow or from lower to
higher velocities. Other quantities, such as dissipa-
tion rate, , the integral lengthscale, , appear more
appropriate for turbulence characterization. The in-
tegral lengthscale is related to the spatial structure of
turbulence. It is assumed to be particularly useful for
TEC developers in assessing a potential interaction
of the turbulent flow with turbines.
Acknowledgments
This work benefitted from the funding support from
France Energies Marines and the French Govern-
ment, operated by the National Research Agency
under the Investments for the Future program: Ref-
erence ANR-10-IEED-0006-07 and ANR-10-IEED-
0006-11. The study represents a contribution to the
projects HYD2M and THYMOTE of the above pro-
gram. We would like to acknowledge Lucile
Furgerot (UNICAEN) for providing the data, and
Anne-Claire Bennis (UNICAEN) - the head of the
research project.
6 REFERENCES
Bailly du Bois, P. B., Dumas, F., Solier, L., & Voiseux, C.
2012. In-situ database toolbox for short-term dispersion
model validation in macro-tidal seas, application for 2D-
model. Continental Shelf Research, 36: 63-82.
Batten, W. M. J., Bahaj, A. S., Molland, A. F., & Chaplin, J. R.
2008. The prediction of the hydrodynamic performance of
marine current turbines. Renewable energy, 33: 1085-1096.
Bennis, A.-C., Furgerot, L., Bailly Du Bois, P., Dumas, F.,
Odaka, T., Lathuilière, C. & Filipot, J.-F. 2020. Numerical
modelling of three-dimensional wave-current interactions in
complex environment: application to Alderney Race. Ap-
plied Ocean Research. 95: 102021
Blackmore T., Myers L.E. and Bahaj A.S. 2016. Effects of tur-
bulence on tidal turbines: Implications to performance,
blade loads, and condition monitoring. International Jour-
nal of Marine Energy. 14: 1–26
Frost, C., Benson, I., Jeffcoate, P., Elsäßer, B., & Whittaker, T.
2018. The Effect of Control Strategy on Tidal Stream Tur-
bine Performance in Laboratory and Field Experiments.
Energies, 11(6) : 1533.
Furgerot, L., Poprawski, Y., Violet, M., Poizot, E., Bailly du
Bois, P., Morillon, M., & Mear, Y. 2019. High-resolution
bathymetry of the Alderney Race and its geological and
sedimentological description (Raz Blanchard, northwest
France). Journal of Maps, 15: 708-718.
Gooch, S., Thomson, J., Polagye, B., & Meggitt, D. 2009. Site
characterization for tidal power. In OCEANS 2009: 1-10.
Guerra, M., & Thomson, J. 2017. Turbulence measurements
from five-beam acoustic Doppler current profilers. Journal
of Atmospheric and Oceanic Technology, 34(6), 1267-1284.
Hay, A. E., McMillan, J., Cheel, R., & Schillinger, D. 2013.
Turbulence and drag in a high Reynolds number tidal pas-
sage targetted for in-stream tidal power. In OCEANS 2013:
1-10.
Ikhennicheu, M., Germain, G., Druault, P., & Gaurier, B. 2019.
Experimental investigation of the turbulent wake past real
seabed elements for velocity variations characterization in
the water column. Int. J. Heat and Fluid Flow, 78: 108426.
Korotenko, K., Sentchev, A., Schmitt, F. G., & Jouanneau, N.
2013. Variability of turbulent quantities in the tidal bottom
boundary layer: Case study in the eastern English Channel.
Continental Shelf Research, 58: 21-31.
Lewis, M., Neill, S. P., Robins P. E., & Hashemi, M. R. 2015
Resource assessment for future generations of tidal-stream
energy arrays. Energy, 83: 403-415.
Lewis, M., Neill, S.P., Robins, P., Hashemi, M.R., & Ward, S.
2017 Characteristics of the velocity profile at tidal-stream
energy sites. Renewable Energy, 114: 258-272.
Lewis, M., McNaughton, J., Márquez-Dominguez, C.,
Todeschini, G., Togneri, M., Masters, I., Allmark, M.,
Stallard, T., Neill, S., Goward-Brown, A. and Robins, P.
2019 Power variability of tidal-stream energy and implica-
tions for electricity supply. Energy, 183: 1061-1074.
Liu, P and Veitch, B. 2012. Design and optimization for
strength and integrity of tidal turbine rotor blades. Energy,
46: 393–404.
MacEnri J., Reed M., Thiringer T. 2013. Influence of tidal pa-
rameters on SeaGen flicker performance. Philosophical
Transactions of the Royal Society of London A: Mathemati-
cal, Physical and Engineering Sciences,
371(1985):20120247.
McMillan, J. M., Hay, A. E., Lueck, R. G., & Wolk, F. 2016.
Rates of dissipation of turbulent kinetic energy in a high
Reynolds number tidal channel. Journal of Atmospheric
and Oceanic Technology, 33(4): 817-837.
McMillan, J. M. & Hay, A. E. 2017. Spectral and structure
function estimates of turbulence dissipation rates in a high-
flow tidal channel using broadband ADCPs. Journal of At-
mospheric and Oceanic Technology, 34(1): 5-20.
Milne, I. A., Sharma, R. N., Flay, R. G., & Bickerton, S. 2013.
Characteristics of the turbulence in the flow at a tidal
stream power site. Phil. Trans. R. Soc. A, 371(1985):
20120196.
Milne, I. A., Day, A. H., Sharma, R. N., & Flay, R. G. J. 2016.
The characterisation of the hydrodynamic loads on tidal
turbines due to turbulence. Renewable and Sustainable En-
ergy Reviews, 56: 851-864.
Myers, L.E., Bahaj, A.S. 2010. Experimental analysis of the
flow field around horizontal axis tidal turbines by use of
scale mesh disk rotor simulators, Ocean Eng, 37 (2): 218-
227.
Osalusi, E., Side, J., Harris, R. 2009. Structure of turbulent
flow in EMEC's tidal energy test site. Int. Communications
in Heat and Mass Transfer, 36: 422-431.
Ouro, P., & Stoesser, T. 2019. Impact of environmental turbu-
lence on the performance and loadings of a tidal stream tur-
bine. Flow, Turbulence and Combustion, 102 : 613-639.
Pope S. B. 2000. Turbulent flows. Cambridge University Press.
Sellar, B. G., Wakelam, G., Sutherland, D. R., Ingram, D. M.,
Venugopal, V. (2018). Characterisation of tidal flows at the
European marine energy centre in the absence of ocean
waves. Energies, 11: 176.
Sentchev, A., Nguyen, T.D., Furgerot, L. & Bailly du Bois, P.
2020a. Underway velocity measurements in Alderney Race:
toward a 3D representation of tidal motions. Phil. Trans.
Royal Soc. A, 2020 (in press)
Sentchev, A., Thiebaut, M. & Schmitt, F.G. 2020b. Impact of
turbulence on power production by a free-stream tidal tur-
bine in real sea conditions. Renewable Energy, 147: 1932–
1940.
Soulsby R. 1997. Dynamics of Marine Sands: a Manual for
Practical Applications, Thomas Telford.
Tatum, S. C., Frost, C. H., Allmark, M., O’Doherty, D. M.,
Mason-Jones, A., Prickett, P. W., ... & O’Doherty, T. 2016.
Wave–current interaction effects on tidal stream turbine
performance and loading characteristics. International
Journal of Marine Energy, 14: 161-179.
Thiebaut, M. & Sentchev, A. 2016. Tidal stream resource as-
sessment in the Dover Strait (eastern English Channel). In-
ternational journal of marine energy, 16: 262-278.
Thiebaut, M. & Sentchev, A. 2017. Asymmetry of tidal cur-
rents off the W. Brittany coast and assessment of tidal ener-
gy resource around the Ushant Island. Renewable energy,
105 : 735-747.
Thiebaut, M., Filipot, J. F., Maisondieu, C., Damblans, G.,
Duarte, R., Droniou, E., ... & Guillou, S. 2020a. A compre-
hensive assessment of turbulence at a tidal-stream energy
site influenced by wind-generated ocean waves. Energy,
191: 116550.
Thiebaut M., Sentchev, A. & Bailly du Bois P. 2020b. Merging
velocity measurements and modeling to improve under-
standing of tidal stream resource in Alderney Race. Energy,
178: 460-470.
Thomson J., Polagye B., Durgesh V., Richmond M. C. 2012.
Measurements of turbulence at two tidal energy sites in Pu-
get Sound,WA. Oceanic Engineering, IEEE J. of Oceanic
Engineering, 37(3):363–374.
Zhou, Z., Scuiller, F., Charpentier, J. F., Benbouzid, M., &
Tang, T. 2014. An up-to-date review of large marine tidal
current turbine technologies. In 2014 International Power
Electronics and Application Conference and Exposition:
480-484.
