Wave Energy Evaluations in Enclosed Seas [601835]

Wave Energy Evaluations in Enclosed Seas
EUGEN RUSU and FLORIN ONEA
Department of Applied Mechanics
Dunarea de Jos University of Galati
47 Domneasca Street, 800008, Galați
ROMANIA
[anonimizat] ,[anonimizat]
Abstract: -The present work has as main objective to evaluate the wave energy patterns in two enclosed seas.
These are the Black and the Caspian s eas.Amedium term analysis of recent satellite data gives a first
perspective of the wave and wi nd climate in the two sea basins . This allowed the identification of the areas that
can be considered as most energetic. Wave prediction systems based on the third generation spectral model
SWAN were i mplemente d and validated in the two seas .A general picture on the wave conditions andthe
wave energy potential, in the basin s of the Black and the Caspian s easis provided by the present work. The
development of the wave energy devices for small amplitude waves is ex pected to be very dynamic in the near
future .Following these tendencies, the problem of extracting this type of renewable energy in the two sea
environments under consideration , most probable coupled in hybrid farms wind -waves, might become of
actuality .
Key-Words: -enclosed seas ,waves, numerical models, renewable energy
1Introduction
Extraction of wave energy became in the last decade
one of the most challenging engineering problems.
Wave energy is abundant and using numerical
models it can be predi cted with a good accuracy in a
time window of a few days. On the other hand,
wave energy is not only more predictable than wind
or solar energy but it has also a higher energetic
density allowing extraction of more energy in
smaller areas.
Various devices to extract this energy have been
designed, but at this moment there is yet no
techno logy that can be considered as being the most
effective. A very important inconvenience that is
usually encountered in the areas traditionally
considered as having high po tential in wave energy
is that very o ften these areas are subjected to strong
wave cond itions that may destroy the devices
operating for e xtracting the wave energy.
From this reason an alternative solution would
be to explore the possibility of implementi ng the
energy farms, eventually together with wind farms,
in areas with smaller wave amplitudes, but where
the extreme wave conditions are usually not so
severe, allowing the functionality of the wave
extraction systems for considerably longer time
interva ls.
In this respect the objective of the present work
is to evaluate in parallel the wave conditions and the
wave energy resources in two enclosed seas usingboth satellite data and results coming from a wave
predi ction system based on numerical models.
The Black Sea is an enclosed sea located deeply
inside the continent that represents the most isolated
part of the World Ocean while the Caspian Sea can
be considered the largest inland body of water in the
world and accounts for 40% to 44% of the total l ake
waters of the world.
The maximal length of the Black Sea (along the
latitude 42o29’N) is 1148 km, while its minimal
width along the meridian from Crimea to the coast
of Turkey is only 258 km. The principal
characteri stics of the Black Sea are: 423000 km2for
the sea area, 555000 km3for its volume, and 1315
and 2258 meters for the mean and maximal depths,
respectiv ely. Three principal structures: the shelf,
the contine ntal slope , and the deep -water basin, can
be clearly distinguished in the bottom topo graphy of
the sea.
The Caspian Sea extends approximately 1200 km
from north to south, with an average width of 325
km east to west, covering a total area of about
400000 km2. Most of the northern Caspian is
shallow, with water depths averaging only 4 m. T he
central Caspian approaches depths of 800 m, while
the southern Caspian has a maximum depth of
slightly over 1000 meters.
The present work points to find out some
preliminary answers to the question of the
availability of the wave energy resources in th ese
two seas.
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
19

2Analysis of Remotely Sensed Data
In the last years more accurate satellite data became
available on various internet sites (as for example
http://las.aviso.oceanobs.com that was the main
source considered in the present work). An altimeter
node gives for each day at zero hours near real time
multi -mission merged non interpolated values of the
significant wave height (Hs) and wind speed (Vw).
These are time (for the last 48 hours) and space (for
1ș squares centered in the node) averaged data sets.
As regards m easuring the sea waves, the
estimates are obtained using empirical models
derived from analyses of the altimeter data. The
algorithm used to deduce the significant wave
height is based on the initial r esults provided in [1]
with the new parametric fits indicated in [2]. For a
typical significant wave height of 2 meters, the error
in the sea state bias correction is approximately 1 -2
cm, i.e., 0.5% to 1.0% of the effect. For measuring
the wind speed, a mathematical relationship that
considers the Ku -band backscatter coefficient
together with the Vandemark and Chapron
algorithm, as in [3]. The wind speed model function
is evaluated for 10 meters above the sea surface.
In order to provide a more r ecent picture of the
characteristics and dynamics of the most relevant
parameters ( HsandVw) concerning the wave and
wind conditions in the two seas, some synthetic
results coming from the analysis of the remotely
sensed d ata, corresponding to t he time in terval
December 2005 -June 2010 arepresented bellow.
The monthly average of the Hs data in eight selected
points (denoted with A1, A2,…A8) from the Black
Sea where this parameter has greater values is
illustrated in Figure 1.The second parameter
analyzed is the wind sp eed (Vw ) and its monthly
average values are illustrated in Figure 2 .
Fig.1.The Black Sea monthly averaged values of
Hs in eight reference points
Fig.2.The Black Sea; monthly averaged values of
Vw (m/s) in eight reference points
The main statistical parameters for three most
energetic points regarding the characteristics of the
wave height and wind speed are presented in Table
1 and Table 2.
Table 1. The Black Sea; overall statistics for the Hs
data in the in three most energe tic points for the
period 2005 to June 2010
Table 2. The Black Sea; overall statistics for the
Vw data in three most energetic points for the
period 2005 to June 2010
For the Caspi an Sea, initial 14 reference points
covering the entire basin of the sea were defined.
They were denoted as P1 to P14 and they are
counted from south to north. The results presented
in Tables 3 and 4 show that the central part of the
Caspian Sea is more e nergetic, especially the
locations corresponding to the refe rence points P8,
P9 and P10. The histograms for these two
parameters ( HsandVw)related to the same data set
(related to the time interval December 2005 -June
2010) are presented in Figure 3 and 4, respectively.
The data from the histograms was stru ctured in total
and winter time respectively, where winter time is
considered here the interval between the beginning
of October until the end of March.Pt PozMean
(m/s)Max
(m/s)Std Kurt Skew
A4 36șE,
42șN0.91 5.1 0.50 7.25 1.92
A8 38șE,
44șN0.86 5.1 0.48 7.82 1.95
A7 36șE,
44șN0.89 5.1 0.48 6.90 1.85
Pt PozMean
(m/s)Max
(m/s)Std Kurt Skew
A4 36șE,
42șN4.76 16.3 2.37 1.22 1.03
A8 38șE,
44șN4.42 16.2 2.11 1.28 1.00
A7 36șE,
44șN4.28 16.2 2.09 1.42 1.03
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
20

Table 3. The Caspian Sea; overall statistics for the
Hs data in three most energetic reference points
Table 4. The Caspian Sea; ove rall statistics for the Vw
data in fourteen reference points
Fig.3.Satellite da ta,Hshistograms for the
reference points P8, P9 and P10. Daily records for
the time interval December 2005 -June 2010; a), c)
and e) total time, b), d) and f) winter time.
Fig.4.Satellite data, Vw histograms for the
reference points P8, P9 and P10 . Daily records for
the time interval December 2005 -June 2010; a), c)
and e) total time, b), d) and f) winter time2 Implementation of the Wave Model
A wave prediction system based on the SWAN
model was implemented and evaluated separately in
each sea. A s regards the Black Sea basin, validation
tests were previously performed against buoy data
as presented in [4, 5].Further validations of the
above mo deling system were performed also in [6]
and the above system was used to provide the
support in the case of the environmental alerts and
to assess the wave -current intera ctions at the mouths
of the Danube [7, 8].
Some results concerning the implementation in
the Caspian Sea of a wave modeling system SWAN
based are presented bellow. The system origin
corresponds to the lower left corner point and has
the coordinates (46.7°E, 36.2°N), where as the
lengths are 8° in x -direction (longitude) and 11.2° in
y-direction (latitude). The computations were
performed in the non stationary mode with a 5
minutes time step. For the three points that were
found as the most energetic in the Caspian Sea (P8,
P9 and P10) Figure 5 illustrates direct comparisons
for Hs, SWAN r esults against remotely sensed data,
while Figure 6 (a, c and e) presents the Hs scatter
plots for the sam e three locations.
Fig.5.The Caspian Sea; Hs direct comparisons,
SWAN r esults against satellite data for the entire
year 2009. a) Refe rence point 8; b) Reference point
9; c) Reference point 10.
Since the accuracy of the input wind field represent
afundamental issue in obtaining be tter results in
wave modeling, Table 5 presents also the statistical
analysis for Vw (ECWMF model wind against the
corresponding remotely sensed data) at the same
five locations (for the reference points P4, P8, P9,Pt PozMean
(m/s)Max
(m/s)Std Kurt Skew
P8 51șE,
42șN0.95 4.18 0.65 2.75 1.51
P9 50șE,
42șN0.95 4.01 0.64 2.65 1.50
P10 49șE,
42șN0.95 3.90 0.65 2.56 1.49
Pt PozMean
(m/s)Max
(m/s)Std Kurt Skew
P8 51șE,
42șN5.32 17.77 2.92 0.24 0.73
P9 50șE,
42șN5.31 17.43 2.92 0.23 0.73
P10 49șE,
42șN5.33 17.19 2.93 0.20 0.72
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
21

P10 an d P12) and Figure 6 (b, d and f) presents the
Vw scatter plots for the points P8, P9 and P10.
Table 5. The Caspian Sea; wave and wind statistics,
remotely sensed data against SWAN and wind
model outputs. Results in five reference points for:
a)Hsstatistics; b) Vwstatistics.
Pt Xm Ym bias rmse si r
a) Hs (m)
P4 0.91 1.04 -0.13 0.37 0.41 0.79
P8 1.03 1.18 -0.14 0.39 0.37 0.83
P9 1.04 1.16 -0.12 0.38 0.37 0.81
P10 1.01 1.02 -0.01 0.34 0.34 0.82
P12 0.60 0.44 0.16 0.33 0.55 0.77
b) Vw (m/s)
P4 4.49 4.37 0.12 1.26 0.28 0.83
P8 5.33 5.60 -0.27 1.40 0.26 0.86
P9 5.29 5.26 0.03 1.28 0.24 0.88
P10 5.20 4.83 0.37 1.29 0.25 0.89
P12 5.08 5.44 -0.36 1.28 0.25 0.90
Fig.6.The Caspian Sea; Scatter plots for the
parameters Hs and Vw model (wave an d wind)
against satellit e data for the entire year 2009,
coresponding for the points P8, P9 and P10.4 Evaluation of two Case Studies
Using the wave modeling systems implemented in
the two seas, two case studies will be considered for
each sea to evaluate and analyze the most relevant
patterns concerning the spatial distribution of the
wave e nergy.
In SWAN, the energy transport components
(expressed in W/m, i.e., energy transport per unit
length of wave front), are computed with the
relationships:
 
  , ,,


        
d d E c g Ed d E c g E
y yTRx xTR
(1)
where: x, y are the problem coordinate system (for
the spherical coordinates x axis corresponds to
longitude and y axis to latitude),   , Ethe wave
energy spectrum, σ the relative wave frequency , θ
the wave direction and c x, cyare the propagation
velocities of the wave energy in the geographical
space d efined as:
  . , U c c cdtx d
g y x
   (2)
Hence the absolute value of the energy transport
(denoted also as wav e power) will be:
.2 2
TRy TRx TR E E E  (3)
The non dimensional normalized wave power is
expressed as:
.
maxTRTR
nTREEE (4)
In the present work ETRmax was defined sep arately
for each individual case study. This is a round off
value approximated the maximum value
corresponding to the computational d omain.
4.1 Black Sea, Cas e study 1 –1997/01/12/h12
This case study, denoted as BS1, provides an
average energy di stribution for the entire Black Sea
basin and is illustrated in Figure 7. This figure
shows the normalized wave power in background
and the ene rgy transport vectors ( in kW/m of wave
front) in the foreground. The locations, for this
computational domain, of the maximum values of
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
22

the wave power are marked with circles. For this
case study the value of ETRmax was set at 20 kW/m.
Although there is an obvious relationship b etween
significant wave height and wave power, the
energetic peak in a comput ational domain is not
necessarily located at the same point as in the
present case.
Fig.7.The Black Sea, Case study 1 –
1997/01/12/h12, average energetic situation,
representa tion for the entire Black Sea.
4.2 Black Sea, Case study 2 -1997/01/21/h18
This second case study presents some storm
cond itions in the Black Sea (BS2) and is illustrated
in Figure 8. It should be noted that this is not an
extreme event, but a regular storm (a typical storm
when the western part of the Black Sea is more
energetic due to the dominant wind patterns). For
this case the value ofmax TRE was set at 150 kW/m.
Fig.8.The Black Sea, Case study 2 –
1997/01/12/h12, average ene rgetic situation,
representation for the entire Black Sea basin
4.3 Caspian Sea, Case study 1 -2009/10/02/h18
The first case study in the Caspian Sea, denoted as
CS1, reflects wave conditions with average energy
(for the winter time period) in the Caspian basin.The background of Figure 9a shows the normalized
wave power (E TR/ETRmax ) for CS1 and the
foreground shows energy transport vectors
(represented with red arrows in kilowatts per meter
of wave front). The location in the computational
domain of the ma ximum value for the wave power
is marked with a red circle. For the present case
study the value ofmax TRE was set to 20 kW/m while
the maximum wave energy in the sea was 22.4
kW/m.
4.3 Caspian Sea, Case study 2 -2009/11/27/h03
The second case study considered in the Caspian
Sea, d enoted as CS2, reflects one of the highest
energetic conditions that were encountered in the
central part of the Caspian Sea for the entire five –
year period analyzed (the time interval December
2005 -June 2010). Thus, although it cannot be
considered as an extreme event this situation can
give a good perspe ctive on the highest energetic
conditions that can be expected in the basin of the
Caspian Sea.
Fig.9.The Caspian Sea; a) Case study 1 –
2009/10/02/h18, av erage energetic situation, b) Case
study 2 -2009/11/27/h03, high energy conditions.
Figure 9b shows in background the normalized
wave power (E TR/ETRmax ) and in foreground the
energy transport vectors (represented with red
arrows in kilowatts per meter of wave front). The
location in the computational domain of the
maximum value for the wave power is marked with
a red circle. For the present situation the value of
ETR max was set to100 kW/m while the maximum
wave energy in the sea was 98.9 kW/m.
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
23

4 Discuss ion of the Results
At this point, a discussion will be employed in
relationship with the acc uracy of the results
provided by the wave prediction system that was
implemented herewith. A comparison of these wave
predictions with some results obtained in simi lar
environments (semi -enclosed or enclosed seas)
coming also from SWAN model simulations will be
made first.
On reference [9] was evaluated a SWAN based
wave forecasting system in the Adriatic Sea forced
with ALADIN wind model (acronym from AROME
Limited Area Decentralized International Network).
For Hsnowcast against buoy data S.I. was in
general between 0.22 -0.3 and rabout 0.9, the
accuracy of the results decreasing in the case of the
forecast pro ducts.
Nevertheless for the case of the wave hindcast
against altimeter data the same system provided Hs
results with lower accuracy ( S.I.about 0.32 -0.34
and rabout 0.72 -78). The accuracy of the wind
velocity was also evaluated against satellite
measurements and there resulted S.I.about 0.34 -36
andrabout 0.78-79.
In the Black Sea using ECMWF wind to force
SWAN, given in [5] was performed a hindcast study
forHsagainst buoy data RMSE in the interval 0.32 –
0.36, S.I.0.36-0.42 and r0.78-0.88. Against satellite
data the same prediction system provided RMSE
0.37-0.4, S.I.0.31-0.34 and r0.77-0.8, as in [8].
Looking at the resul ts presented in Table 5 and
Figure 6 (a, c and e) it can be noticed that from a
statistical point of view the estimations provided by
the wave prediction system implemented in the
Caspian Sea are compatible with those coming from
similar systems in the Adriatic Sea or the Black Sea.
Thus RMSE has values between 0.33 -0.39, S.I.
between 0.34 -0.41 and rabout 0.79 -0.83.
An exception is related with the reference point
P12 that is locat ed in the north. The lower accuracy
concerning S.I.(0.55) and r(0.77) in the conditions
when the wind accuracy is even better in that region
than in other parts of the Caspian Sea ( S.I.= 0.25 and
r=0.90 for Vw) are probably related with the fact
that the entire northern region of the Caspian Sea is
characterized by very shallow water (about 4 m
depth). Thus some process as triad wave -wave
interactions, bottom friction, diffraction, breaking,
etc, that were not accounted in the global
simulations may become rather relevant in this area.
A solution to improve the model predictions in the
north of the Caspian Sea would be to define a higher
resolution computational domain, nested in the area
that covers the entire basin, which would cover the
northern sector o f the sea and where the mainshallow water processes avail able in SWAN would
be also activated (performing eventually a
calibration process).
In relationship with some other factors that may
improve the performances of such a wave prediction
system based on spectral phase averaged models, on
the first place should be probably considered the
quality of the wind fields. Many studies have been
performed on the effects of the wind fields when
modeling waves in enclosed, semi -enclosed or
relatively small basins , as for example in [10, 11].
The most obvious conclusion would be that the
accuracy of the results provided by the wave models
is highly dependent on the accuracy of the
meteorological mo dels that were used to force them.
The most obvious conclusion woul d be that the
accuracy of the results provided by the wave models
is highly dependent on the accuracy of the
meteorological mo dels that were used to force them.
5 Concluding Remarks
The results presented in this work concerning the
evaluation of the w ave energy potential in the Black
and the Caspian seas will be discussed in this final
section in parallel with data from another area that is
considered among the coastal environments richest
in wave energy resources. This is the Iberian coastal
environme nt in relationship with which in depth
analyses concerning the spatial distributions of the
wave energy were carried out in [12, 13]. Thus for
the case of average energetic conditions the
maximum values encountered for the wave power
were between 50 and 10 0 kW/m, in the Ib erian
coastal environment, between 20 and 50 kW/m in
the Black Sea while a little over 20 kW/m for the
Caspian Sea. As regards the highest energetic
conditions in the three environments co nsidered the
differences are even more significant . Thus in the
Iberian near -shore the maximum wave power was
sometimes, even greater than 600 kW/m, in the
Black Sea, was around 300 kW/m, while in the
Caspian Sea less than 100 kW/m.
Finally, it should be also highlighted that in the
perspective of the dev elopment of the wave energy
devices for small amplitude waves that is expected
to be quite dynamic in the near future bringing also
some technological breakthroughs, and most
probable coupled in mixed farms wind -waves, the
problem of extracting this type o f renewable energy
in the Black and the Caspian Seas might become of
actuality.
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
24

Acknowledgments
The work of the second author has been made in the
scope of the project EFICIENT (Management
System for the Fellowships Granted to the PhD
Students) supporte d by the Project SOP HRD –
EFICIENT 61445 /2009 .
References:
[1] Gaspar, P., and Florens, J. -P., Estimation of the
sea state bias in radar altimeter measurements
of sea level: Results from a new nonparametric
method, J. Geophys. Res .,103, 15, 1998, pp.
803-814.
[2] Labroue, F. Ogor, G. Lafitte, L. Marchal, and
Rafanel, M., Improving nonparametric
estimates of the sea state bias in radar altimetry
measurements of sea level, J.Atmos. Oceanic
Technol ., 19, 2002, pp.1690 -1707
[3] Vandemark, D. Gourrion, J. Bailey, S. and
Chapron, B., An Altimeter Wind Speed Model
Using Both Radar Backsca tter and Significant
Wave Height, NASA Technical Report:
20010020922 ,2000.
[4] Guedes Soares, C. and Rusu, E., SWAN
Hindcast in the Black Sea, Fifth International
Symposium WAVES 2005 , 3rd –7th July 2005,
Madrid, Spain, Paper number 155, CD edition,
2005, 11p.
[5] Rusu, E, Wave energy assessments in the Black
Sea, Journal of Marine Science and
Technology , Volume 14, i ssue 3 , 2009, pp.
359-372.
[6] Rusu, L. and Ivan, A., Modelling Wind Waves
in the R omanian Coastal Environment.
Environmental Enginee ring and Management
Journal , 9(4), 2010, pp.547-552.
[7] Rusu, L., A system to evaluate the oil spills
propagation in the coastal environmental of the
Black Sea, Journal of Environmental
Engineering and Lands cape Management , 18
(4),2010, pp.288-295.
[8] Rusu, E, Modeling of wave -current interactions
at the Danube’s mouths, Journal of Marine
Science and Techno logy, Vol. 15, Issue 2,
2010, pp.143-159.
[9] Dykes, J.D., Wang, D.W., Book, J.W., An
evaluation of a high -resolution operational
wave forecasting system in the Adriatic Sea, J.
Mar. Syst. 78,2009, pp.255-271.[10] Rusu, E. & Guedes Soares C., Wave Energy
Assessments in the Coastal Environment of
Portugal Continental, the 27th International
Conference on Offshore Mechanics and Arctic
Engineering OMAE2008 June 15 -20, Estoril,
Portu gal, CD edition, 2008.
[11] Rusu, E. and Guedes Soares C., Numerical
modeling to estimate the spatial distribution of
the wav e energy in the Po rtuguese nearshore,
Renewable Energy ,Elsevier , Volum e 34, Issue
6, 2009, pp. 1501 -1516.
Recent Researches in Environment, Energy Systems and Sustainability
ISBN: 978-1-61804-088-6
25