Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas

13
Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas Haralambos S Bagiorgas 1,, Giouli Mihalakakou 1 , Shafiqur Rehman 2,3 and Luai M Al-Hadhrami 2 1 Department of Environmental and Natural Resources Management, University of Ioannina, Seferi 2, 30100 Agrinio, Greece. 2 Center for Engineering Research, Research Institute, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia. 3 Mechanical and Aeronautical Engineering Department, University of Pretoria, Pretoria, South Africa. Corresponding author. e-mail: [email protected] This paper utilizes wind speed data measured at 3 and 10 m above water surface level using buoys at 10 stations in Ionian and Aegean Seas to understand the behaviour of wind and thereafter energy yield at these stations using 5 MW rated power offshore wind turbine. With wind power densities of 971 and 693 W/m 2 at 50 m above water surface level, Mykonos and Lesvos were found to be superb and outstanding windy sites with wind class of 7 and 6, respectively. Other locations like Athos, Santorini and Skyros with wind power density of more than 530 W/m 2 and wind class of 5 were found to be the excellent sites. Around 15–16% higher winds were observed at 10 m compared to that at 3 m. Lower values of wind speed were found during summer months and higher during winter time in most of the cases reported in the present work. Slightly decreasing (2% per year) linear trends were observed in annual mean wind speed at Lesvos and Santorini. These trends need to be verified with more data from buoys or from nearby onshore meteorological stations. At Athos and Mykonos, increasing linear trends were estimated. At all the stations the chosen wind turbine could produce energy for more than 70% of the time. The wind speed distribution was found to be well represented by Weibull parameters obtained using Maximum likelihood method compared to WAsP and Method of Moments. 1. Introduction The power of the wind in good onshore locations has already been transformed to an opportunity for economic profit and is becoming competitive with other power generation methods. The main advan- tage of offshore wind power is that wind speeds are generally higher than over land (Bailey et al. 2002). The other factor, which encourages offshore wind farm development, is the reduced wind tur- bulence. Thus, the offshore turbines are likely to have less fatigue stresses. Musial and Butterfield (2004) reported the cost of offshore wind energy to be $US0.051/kWh based on a generic 5 MW rated capacity wind machine for a hypothetical wind farm of 500 MW installed capacity. The global cumulative offshore wind power installed capacity reached 2940 MW with an addition of 883 MW in 2010 alone, a value almost 52% more than that installed in 2009. According to EWEA 2010 statis- tics, annual additions in offshore global wind power installed capacity present an increasing trend for Keywords. Offshore wind power assessment; Ionian Sea; Aegean Sea; buoys; wind speed; Weibull distribution; frequency distribution. J. Earth Syst. Sci. 121, No. 4, August 2012, pp. 975–987 c Indian Academy of Sciences 975

Transcript of Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas

Offshore wind speed and wind power characteristicsfor ten locations in Aegean and Ionian Seas

Haralambos S Bagiorgas1lowast Giouli Mihalakakou

1Shafiqur Rehman

23 and Luai M Al-Hadhrami2

1Department of Environmental and Natural Resources Management University of Ioannina Seferi 230100 Agrinio Greece

2Center for Engineering Research Research Institute King Fahd University of Petroleum and MineralsDhahran 31261 Saudi Arabia

3Mechanical and Aeronautical Engineering Department University of Pretoria Pretoria South AfricalowastCorresponding author e-mail chbagiorccuoigr

This paper utilizes wind speed data measured at 3 and 10 m above water surface level using buoys at10 stations in Ionian and Aegean Seas to understand the behaviour of wind and thereafter energy yieldat these stations using 5 MW rated power offshore wind turbine With wind power densities of 971and 693 Wm2 at 50 m above water surface level Mykonos and Lesvos were found to be superb andoutstanding windy sites with wind class of 7 and 6 respectively Other locations like Athos Santoriniand Skyros with wind power density of more than 530 Wm2 and wind class of 5 were found to be theexcellent sites Around 15ndash16 higher winds were observed at 10 m compared to that at 3 m Lowervalues of wind speed were found during summer months and higher during winter time in most of thecases reported in the present work Slightly decreasing (sim2 per year) linear trends were observed inannual mean wind speed at Lesvos and Santorini These trends need to be verified with more data frombuoys or from nearby onshore meteorological stations At Athos and Mykonos increasing linear trendswere estimated At all the stations the chosen wind turbine could produce energy for more than 70 ofthe time The wind speed distribution was found to be well represented by Weibull parameters obtainedusing Maximum likelihood method compared to WAsP and Method of Moments

1 Introduction

The power of the wind in good onshore locationshas already been transformed to an opportunity foreconomic profit and is becoming competitive withother power generation methods The main advan-tage of offshore wind power is that wind speedsare generally higher than over land (Bailey et al2002) The other factor which encourages offshorewind farm development is the reduced wind tur-bulence Thus the offshore turbines are likely to

have less fatigue stresses Musial and Butterfield(2004) reported the cost of offshore wind energy tobe $US0051kWh based on a generic 5 MW ratedcapacity wind machine for a hypothetical windfarm of 500 MW installed capacity The globalcumulative offshore wind power installed capacityreached 2940 MW with an addition of 883 MWin 2010 alone a value almost 52 more than thatinstalled in 2009 According to EWEA 2010 statis-tics annual additions in offshore global wind powerinstalled capacity present an increasing trend for

Keywords Offshore wind power assessment Ionian Sea Aegean Sea buoys wind speed Weibull distribution frequency

distribution

J Earth Syst Sci 121 No 4 August 2012 pp 975ndash987ccopy Indian Academy of Sciences 975

976 Haralambos S Bagiorgas et al

almost each year from 2001 to 2010 making thecumulative wind power installed capacity growingrapidly (EWEA 2010)

Back in 1994 Gaudiosi (1994) presented a gen-eral overview of the offshore wind energy activ-ity for the Mediterranean and other EuropeanSeas Emphasis was given to resource assessmentplanning technical development applications eco-nomics and environment Still (2001) presentedthe details of the first offshore wind farm to bebuilt in United Kingdom back in 2001 This windfarm was planned to use wind turbines from VestasManwell et al (2002) presented a summary of theongoing work on the assessment of the wind energyresource off the coast of southern New England inthe United States Their work consisted of a reviewof existing offshore wind data the measurement ofnew data at an offshore site correlation and pre-diction of long-term data at a new offshore site byreference to a longer-term island site and assess-ment of the overall coastal resource through the useof the MesoMap software Rogers et al (2003) andManwell et al (2007) presented an overview of thenational and regional efforts to define the potentialoffshore wind energy resource in the United States

According to NREL (Musial and Ram 2010) theUnited States leads the world in installed land-based wind energy capacity yet has no offshorewind generating capacity to date Although theUnited States has built no offshore wind projectsso far about 20 projects representing more than2000 MW of capacity are in the planning andpermitting process At the state level only Texashas a leasing system for offshore wind Accord-ing to Zhixin et al (2009) the potential of off-shore wind energy is vast and the key techniquesof offshore wind farm includes optimization andestimation electric transmission and connectionsystem and stability operation manufacturingexcursion system investigation and the baseof wind generator three connection projectsbased on light HVDC generator disperse controlEsteban et al (2011) presented a brief revisionof the state-of-the-art of offshore wind power fol-lowed by a critical discussion about the causes ofthe recent growth that is currently taking shapeThe discussion was based on the comparison ofoffshore wind energy with other renewable ener-gies (like onshore wind marine hydrodynamicshydraulic solar etc) and even with conventionalpower According to Bilgili et al (2011) offshorewinds tend to flow at higher speeds than onshorewinds thus it allows turbines to produce more elec-tricity Estimates predict a huge increase in windenergy development over the next 20 years Muchof this development will be offshore wind energy

Dhanju et al (2008) provided the offshore windresources assessment of the US state of Delaware

They found year-round average wind power outputof over 5200 MW or about four times the aver-age electrical consumption of the state Accord-ing to the authors on local wholesale electricitymarkets this could produce over $2 billionyear inrevenue Pimenta et al (2008) used meteorologi-cal station satellite data (QuikSCAT) and boththeoretical and practical measures of wind tur-bine performance to evaluate offshore wind powerpotential of southern coast of Brazil Next theyused bathymetry and the properties of currentwind-electric technology to develop maps of windspeed wind power density and practical turbineoutput in power units (MW) Pimenta et al (2008)found the most favourable conditions along thecoast between 281S and 331S in the shallowerwaters of south Brazil Finally they estimated atotal resource of 102 GW average electrical produc-tion in one coastal area equivalent to the electricdemand of the entire country

Dvorak et al (2010) combined multi-yearmesoscale modelling results validated using off-shore buoys with high resolution bathymetry tocreate a wind energy resource assessment for off-shore California Initial estimates showed that14ndash23 GW 44ndash83 GW and 528ndash649 GW ofdeliverable power could be harnessed from offshoreCalifornia using monopile multi-leg and floatingturbine foundations respectively A single pro-posed wind farm near Cape Mendocino coulddeliver an average 800 MW of gross renewablepower and reduce Californiarsquos current carbon emit-ting electricity generation 4 on an energy basis(Dvorak et al 2010) The electric power genera-tion of co-located offshore wind turbines and waveenergy converters along the California coast wasinvestigated by Stoutenburg et al (2010) Theyused meteorological wind data from the NationalBuoy Data Center to estimate the hourly poweroutput from offshore wind turbine at the sites ofthe buoys The authors estimated the offshore windfarm capacity factors from 30 to 50 MartinMederos et al (2011) used wind speed and direc-tion data from 40 weather stations and satellite todevelop the offshore wind resources atlas The off-shore atlas for Canary Island presented in MartinMederos et al (2011) was based on three combinedmodels The authors have drawn up three mapsnamely one for mean wind speed one for meanwind direction and one for dominant wind direc-tion The study concluded that a great offshorewind power potential exists in the islands

During the last years a lot of discussion hasbeen made concerning the renewable sourcedenergy potential that could be captured in thearea of Eastern Mediterranean and especially inthe islands of Aegean and Ionian seas Variousresearch works and tasks had been demonstrated

Offshore wind speed and wind power characteristics 977

trying to determine the maximum wind sourcedenergy potential in the area (Poulos et al 1997Incecik and Erdogmus 1995 Klaic et al 2009Borhan 1998 Hamad et al 2006) Karamanis et al(2011) assessed the offshore wind power potentialof two sites Zakynthos and Pylos in the Ioniansea by using two years mean wind speed at 10 mThey found the annual mean wind speed at thesetwo sites as 57 and 58 ms respectively Offshorewind capacity factors of up to 48 for energy pro-duction were calculated with the existing offshoreturbines technology at a hub height of 100 m Thepresent study utilizes wind speed and directiondata from 10 buoy stations in Ionian and Aegeanseas to assess the offshore wind power potential inthese areas

The present paper has as main objective tothoroughly examine the wind characteristics formany offshore locations in the areas of Aegean andIonian seas as being areas with large differences inwind characteristics even more between two neigh-bouring sites by using the studied Weibull para-meters of the areas In addition this studyattempts to understand the behaviour of wind andthereafter energy yield at these stations using a5 MW rated power offshore wind turbine

2 Site and data description

The study utilizes measured wind speed and winddirection data at 3-hr intervals from 10 buoy sta-tions namely Athos E1M3A Lesvos MykonosPetrokaravo Santorini and Skyros in Aegean Seaand Kalamata Pylos and Zakynthos in Ionian SeaThe data collection period varied between 31 and11 years as can be seen from 6th column of thetable 1 below The latitude longitude the data col-lection periods are summarized in table 1 and thephysical locations are depicted in figure 1

Two types of surface buoys are used in theselected sites Oceanor Wavescan at AthosE1M3A and Pylos (figure 1 locations 1 2 7) andOceanor SeaWatch at Kalamata Lesvos MykonosPetrokaravo Santorini Zakynthos and Skyros(figure 1 locations 3 4 5 6 8 9 10) Wave-scan buoy1 is a metocean data collection buoy thatprovides wave height wave direction and meteoro-logical parameters sea surface temperature salin-ity and temperature profiles It is the ideal buoyfor deep-water measurements remote locationsand strong current conditions The buoy can alsobe fitted with numerous other sensors to satisfythe scientistrsquos specific configuration needs such as

1httpwwwgeoscoukservicesoceanWavescanasp

oxygen hydrocarbon gamma radiation measure-ment and an optical sensor for algae detectionSeawatch buoy2 is a multi-sensor wave directionaldata buoy capable of measuring wave height anddirection ocean current speed and direction mete-orological parameters sea surface temperaturesalinity and temperature and salinity profiles As avertically stabilised buoy with low pitchroll motionit is ideally suited for wave current and wind mea-surements in shallow water depth up to 500 m Forwater quality measurement the buoy can be fittedwith numerous other sensors satisfying the cus-tomerrsquos specific configuration needs such as oxy-gen hydrocarbon gamma radiation measurementand an optical sensor for algae detection

The data collection buoys used at the aforemen-tioned sites are shown in figure 2 These buoyswere instrumented to measure the air-pressureair-temperature wind speed and direction waveheight period and direction sea surface salin-ity and temperature surface current speed anddirection The anemometer used in both types ofbuoys is the Young 04106 (four-blade propelleranemometer)

The Aegean Sea an arm of the MediterraneanSea is located between the Greek peninsula on thewest and Asia Minor on the east while the islandof Crete can be taken as marking its boundary onthe south The Aegean is connected through thestraits of the Dardanelles the Sea of Marmara andthe Bosporus to the Black Sea It contains morethan 3000 islands with Crete as its largest islandand is considered as the home of the earliest Euro-pean civilization The Aegean Sea covers about214000 km2 in area and measures about 610 kmlongitudinally and 300 km latitudinally The searsquosmaximum depth is 3543 m east of Crete It alsohas a good connection to the Ionian Sea to thewest mainly through the strait lying between thePeloponnese peninsula of Greece and CreteThe rocks making up the floor of the Aegean aremainly limestone though often greatly altered byvolcanic activity that has convulsed the region inrelatively recent geologic times

The Ionian Sea is also an arm of theMediterranean Sea south of the Adriatic Sea Itis bounded by southern Italy including CalabriaSicily and the Salento peninsula to the west and bysouthwestern Albania and Greek peninsula to theeast containing a large number of Greek islands Itis connected with the Adriatic Sea to the north bythe Strait of Taranto The sea is one of the mostseismic areas in the world

North winds prevail in the Aegean Sea althoughfrom the end of September to the end of May

2httpwwwgeoscoukservicesoceanseabuoyasp

978 Haralambos S Bagiorgas et al

Table 1 Summary of site specific coordinates and wind power class

Latitude Longitude Depth DFL Period WPD Wind

Lactation (N) (E) (m) (km) Start date End date (years) (Wm2) class

Athos 3957prime843primeprime 2443prime208primeprime 212 2746 05252000 12312010 11 539 5

E1M3A 3546prime990primeprime 2454prime880primeprime 1440 4031 05282007 01012011 36 493 4

Kalamata 3658prime283primeprime 2205prime620primeprime 340 413 12311999 01012011 11 206 2

Lesvos 3909prime346primeprime 2548prime472primeprime 121 282 12311999 01012011 11 693 6

Mykonos 3730prime690primeprime 2527prime494primeprime 138 441 12311999 01012011 11 971 7

Petrokaravo 3735prime915primeprime 2333prime715primeprime 211 893 08282007 01012011 33 368 3

Pylos 3649prime533primeprime 2135prime762primeprime 1681 817 11092007 01012011 31 444 4

Santorini 3615prime510primeprime 2529prime769primeprime 314 921 12311999 01012011 11 590 5

Skyros 3906prime360primeprime 2427prime568primeprime 117 1392 08282007 01012011 33 540 5

Zakynthos 3756prime927primeprime 2036prime059primeprime 313 681 11082007 01012011 31 469 4

DFL = Distance from land Wind classes 2 = Marginal 3 = Fair 4 = Good 5 = Excellent 6 = Outstanding 7 = Superb

Figure 1 Map of Ionian and Aegean seas with the sites of the buoys 1 Athos 2 E1M3A 3 Kalamata 4 Lesvos 5 Mykonos6 Petrokaravo 7 Pylos 8 Santorini 9 Zakynthos and 10 Skyros

during the mild winter season these winds alter-nate with southwesterlies The Etesians are thestrong dry north winds of the Aegean Sea which

blow from about mid-May to mid-September Thebarometric conditions that cause these prevailingwinds are the existence of both high barometric

Offshore wind speed and wind power characteristics 979

Figure 2 Views of the Oceanor Wavescan (a) and Oceanor SeaWatch (b) instrumented buoys used at these stations

Table 2 Turbulence intensity values for the selected sites

Location WS3 WS10 SD3 SD10 Turbulence 3 Turbulence 10

Athos 473 547 343 396 0837 0724

E1M3A 530 614 275 318 0600 0518

Kalamata 366 424 235 272 0743 0642

Lesvos 584 675 341 393 0673 0582

Mykonos 670 776 356 412 0615 0531

Petrokaravo 467 542 275 319 0683 0589

Pylos 497 577 288 335 0674 0581

Santorini 558 647 302 350 0627 0541

Skyros 5333 618 317 367 0069 0594

Zakynthos 490 568 293 339 0692 0597

pressures over the central Balkans and low pres-sures over Turkey (Akdag et al 2010) During hotsummer days they are at their strongest in theafternoon and often die down at night but some-times these winds last for days without a breakSimilar winds blow in the Adriatic and Ionianregions In the northern Aegean Sea the Etesiansblow as winds of northeasterly to northerly direc-tion Moving south in the central Aegean theyblow as winds of northerly direction while in thesouthern Aegean the Cretan and the CarpathianSea they blow as northwesterlies The same windsblow in Cyprus as westerlies to southwesterliesbeing more humid (Incecik and Erdogmus 1995Poulos et al 1997 Borhan 1998 Hamad et al 2006Klaic et al 2009)

The tides of the Aegean basin seem to followthe movements of those in the eastern Mediter-ranean generally Aegean currents generally are notsmooth whether considered from the viewpoint ofeither speed or direction They are chiefly influ-enced by blowing winds Water temperatures in theAegean are influenced by the cold-water masses oflow temperature that flow in from the Black Sea tothe northeast The sea surface temperature in theAegean ranges from about 60ndash77F (16ndash25C)varying with location and time of year

The turbulence intensity values for the selectedsites are calculated as the quotient of thestandard deviation to the average wind speed(σaverage WS) and are given in table 2 (Belu andKoracin 2009 Yu and Gan Chowdhury 2009) For

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

976 Haralambos S Bagiorgas et al

almost each year from 2001 to 2010 making thecumulative wind power installed capacity growingrapidly (EWEA 2010)

Back in 1994 Gaudiosi (1994) presented a gen-eral overview of the offshore wind energy activ-ity for the Mediterranean and other EuropeanSeas Emphasis was given to resource assessmentplanning technical development applications eco-nomics and environment Still (2001) presentedthe details of the first offshore wind farm to bebuilt in United Kingdom back in 2001 This windfarm was planned to use wind turbines from VestasManwell et al (2002) presented a summary of theongoing work on the assessment of the wind energyresource off the coast of southern New England inthe United States Their work consisted of a reviewof existing offshore wind data the measurement ofnew data at an offshore site correlation and pre-diction of long-term data at a new offshore site byreference to a longer-term island site and assess-ment of the overall coastal resource through the useof the MesoMap software Rogers et al (2003) andManwell et al (2007) presented an overview of thenational and regional efforts to define the potentialoffshore wind energy resource in the United States

According to NREL (Musial and Ram 2010) theUnited States leads the world in installed land-based wind energy capacity yet has no offshorewind generating capacity to date Although theUnited States has built no offshore wind projectsso far about 20 projects representing more than2000 MW of capacity are in the planning andpermitting process At the state level only Texashas a leasing system for offshore wind Accord-ing to Zhixin et al (2009) the potential of off-shore wind energy is vast and the key techniquesof offshore wind farm includes optimization andestimation electric transmission and connectionsystem and stability operation manufacturingexcursion system investigation and the baseof wind generator three connection projectsbased on light HVDC generator disperse controlEsteban et al (2011) presented a brief revisionof the state-of-the-art of offshore wind power fol-lowed by a critical discussion about the causes ofthe recent growth that is currently taking shapeThe discussion was based on the comparison ofoffshore wind energy with other renewable ener-gies (like onshore wind marine hydrodynamicshydraulic solar etc) and even with conventionalpower According to Bilgili et al (2011) offshorewinds tend to flow at higher speeds than onshorewinds thus it allows turbines to produce more elec-tricity Estimates predict a huge increase in windenergy development over the next 20 years Muchof this development will be offshore wind energy

Dhanju et al (2008) provided the offshore windresources assessment of the US state of Delaware

They found year-round average wind power outputof over 5200 MW or about four times the aver-age electrical consumption of the state Accord-ing to the authors on local wholesale electricitymarkets this could produce over $2 billionyear inrevenue Pimenta et al (2008) used meteorologi-cal station satellite data (QuikSCAT) and boththeoretical and practical measures of wind tur-bine performance to evaluate offshore wind powerpotential of southern coast of Brazil Next theyused bathymetry and the properties of currentwind-electric technology to develop maps of windspeed wind power density and practical turbineoutput in power units (MW) Pimenta et al (2008)found the most favourable conditions along thecoast between 281S and 331S in the shallowerwaters of south Brazil Finally they estimated atotal resource of 102 GW average electrical produc-tion in one coastal area equivalent to the electricdemand of the entire country

Dvorak et al (2010) combined multi-yearmesoscale modelling results validated using off-shore buoys with high resolution bathymetry tocreate a wind energy resource assessment for off-shore California Initial estimates showed that14ndash23 GW 44ndash83 GW and 528ndash649 GW ofdeliverable power could be harnessed from offshoreCalifornia using monopile multi-leg and floatingturbine foundations respectively A single pro-posed wind farm near Cape Mendocino coulddeliver an average 800 MW of gross renewablepower and reduce Californiarsquos current carbon emit-ting electricity generation 4 on an energy basis(Dvorak et al 2010) The electric power genera-tion of co-located offshore wind turbines and waveenergy converters along the California coast wasinvestigated by Stoutenburg et al (2010) Theyused meteorological wind data from the NationalBuoy Data Center to estimate the hourly poweroutput from offshore wind turbine at the sites ofthe buoys The authors estimated the offshore windfarm capacity factors from 30 to 50 MartinMederos et al (2011) used wind speed and direc-tion data from 40 weather stations and satellite todevelop the offshore wind resources atlas The off-shore atlas for Canary Island presented in MartinMederos et al (2011) was based on three combinedmodels The authors have drawn up three mapsnamely one for mean wind speed one for meanwind direction and one for dominant wind direc-tion The study concluded that a great offshorewind power potential exists in the islands

During the last years a lot of discussion hasbeen made concerning the renewable sourcedenergy potential that could be captured in thearea of Eastern Mediterranean and especially inthe islands of Aegean and Ionian seas Variousresearch works and tasks had been demonstrated

Offshore wind speed and wind power characteristics 977

trying to determine the maximum wind sourcedenergy potential in the area (Poulos et al 1997Incecik and Erdogmus 1995 Klaic et al 2009Borhan 1998 Hamad et al 2006) Karamanis et al(2011) assessed the offshore wind power potentialof two sites Zakynthos and Pylos in the Ioniansea by using two years mean wind speed at 10 mThey found the annual mean wind speed at thesetwo sites as 57 and 58 ms respectively Offshorewind capacity factors of up to 48 for energy pro-duction were calculated with the existing offshoreturbines technology at a hub height of 100 m Thepresent study utilizes wind speed and directiondata from 10 buoy stations in Ionian and Aegeanseas to assess the offshore wind power potential inthese areas

The present paper has as main objective tothoroughly examine the wind characteristics formany offshore locations in the areas of Aegean andIonian seas as being areas with large differences inwind characteristics even more between two neigh-bouring sites by using the studied Weibull para-meters of the areas In addition this studyattempts to understand the behaviour of wind andthereafter energy yield at these stations using a5 MW rated power offshore wind turbine

2 Site and data description

The study utilizes measured wind speed and winddirection data at 3-hr intervals from 10 buoy sta-tions namely Athos E1M3A Lesvos MykonosPetrokaravo Santorini and Skyros in Aegean Seaand Kalamata Pylos and Zakynthos in Ionian SeaThe data collection period varied between 31 and11 years as can be seen from 6th column of thetable 1 below The latitude longitude the data col-lection periods are summarized in table 1 and thephysical locations are depicted in figure 1

Two types of surface buoys are used in theselected sites Oceanor Wavescan at AthosE1M3A and Pylos (figure 1 locations 1 2 7) andOceanor SeaWatch at Kalamata Lesvos MykonosPetrokaravo Santorini Zakynthos and Skyros(figure 1 locations 3 4 5 6 8 9 10) Wave-scan buoy1 is a metocean data collection buoy thatprovides wave height wave direction and meteoro-logical parameters sea surface temperature salin-ity and temperature profiles It is the ideal buoyfor deep-water measurements remote locationsand strong current conditions The buoy can alsobe fitted with numerous other sensors to satisfythe scientistrsquos specific configuration needs such as

1httpwwwgeoscoukservicesoceanWavescanasp

oxygen hydrocarbon gamma radiation measure-ment and an optical sensor for algae detectionSeawatch buoy2 is a multi-sensor wave directionaldata buoy capable of measuring wave height anddirection ocean current speed and direction mete-orological parameters sea surface temperaturesalinity and temperature and salinity profiles As avertically stabilised buoy with low pitchroll motionit is ideally suited for wave current and wind mea-surements in shallow water depth up to 500 m Forwater quality measurement the buoy can be fittedwith numerous other sensors satisfying the cus-tomerrsquos specific configuration needs such as oxy-gen hydrocarbon gamma radiation measurementand an optical sensor for algae detection

The data collection buoys used at the aforemen-tioned sites are shown in figure 2 These buoyswere instrumented to measure the air-pressureair-temperature wind speed and direction waveheight period and direction sea surface salin-ity and temperature surface current speed anddirection The anemometer used in both types ofbuoys is the Young 04106 (four-blade propelleranemometer)

The Aegean Sea an arm of the MediterraneanSea is located between the Greek peninsula on thewest and Asia Minor on the east while the islandof Crete can be taken as marking its boundary onthe south The Aegean is connected through thestraits of the Dardanelles the Sea of Marmara andthe Bosporus to the Black Sea It contains morethan 3000 islands with Crete as its largest islandand is considered as the home of the earliest Euro-pean civilization The Aegean Sea covers about214000 km2 in area and measures about 610 kmlongitudinally and 300 km latitudinally The searsquosmaximum depth is 3543 m east of Crete It alsohas a good connection to the Ionian Sea to thewest mainly through the strait lying between thePeloponnese peninsula of Greece and CreteThe rocks making up the floor of the Aegean aremainly limestone though often greatly altered byvolcanic activity that has convulsed the region inrelatively recent geologic times

The Ionian Sea is also an arm of theMediterranean Sea south of the Adriatic Sea Itis bounded by southern Italy including CalabriaSicily and the Salento peninsula to the west and bysouthwestern Albania and Greek peninsula to theeast containing a large number of Greek islands Itis connected with the Adriatic Sea to the north bythe Strait of Taranto The sea is one of the mostseismic areas in the world

North winds prevail in the Aegean Sea althoughfrom the end of September to the end of May

2httpwwwgeoscoukservicesoceanseabuoyasp

978 Haralambos S Bagiorgas et al

Table 1 Summary of site specific coordinates and wind power class

Latitude Longitude Depth DFL Period WPD Wind

Lactation (N) (E) (m) (km) Start date End date (years) (Wm2) class

Athos 3957prime843primeprime 2443prime208primeprime 212 2746 05252000 12312010 11 539 5

E1M3A 3546prime990primeprime 2454prime880primeprime 1440 4031 05282007 01012011 36 493 4

Kalamata 3658prime283primeprime 2205prime620primeprime 340 413 12311999 01012011 11 206 2

Lesvos 3909prime346primeprime 2548prime472primeprime 121 282 12311999 01012011 11 693 6

Mykonos 3730prime690primeprime 2527prime494primeprime 138 441 12311999 01012011 11 971 7

Petrokaravo 3735prime915primeprime 2333prime715primeprime 211 893 08282007 01012011 33 368 3

Pylos 3649prime533primeprime 2135prime762primeprime 1681 817 11092007 01012011 31 444 4

Santorini 3615prime510primeprime 2529prime769primeprime 314 921 12311999 01012011 11 590 5

Skyros 3906prime360primeprime 2427prime568primeprime 117 1392 08282007 01012011 33 540 5

Zakynthos 3756prime927primeprime 2036prime059primeprime 313 681 11082007 01012011 31 469 4

DFL = Distance from land Wind classes 2 = Marginal 3 = Fair 4 = Good 5 = Excellent 6 = Outstanding 7 = Superb

Figure 1 Map of Ionian and Aegean seas with the sites of the buoys 1 Athos 2 E1M3A 3 Kalamata 4 Lesvos 5 Mykonos6 Petrokaravo 7 Pylos 8 Santorini 9 Zakynthos and 10 Skyros

during the mild winter season these winds alter-nate with southwesterlies The Etesians are thestrong dry north winds of the Aegean Sea which

blow from about mid-May to mid-September Thebarometric conditions that cause these prevailingwinds are the existence of both high barometric

Offshore wind speed and wind power characteristics 979

Figure 2 Views of the Oceanor Wavescan (a) and Oceanor SeaWatch (b) instrumented buoys used at these stations

Table 2 Turbulence intensity values for the selected sites

Location WS3 WS10 SD3 SD10 Turbulence 3 Turbulence 10

Athos 473 547 343 396 0837 0724

E1M3A 530 614 275 318 0600 0518

Kalamata 366 424 235 272 0743 0642

Lesvos 584 675 341 393 0673 0582

Mykonos 670 776 356 412 0615 0531

Petrokaravo 467 542 275 319 0683 0589

Pylos 497 577 288 335 0674 0581

Santorini 558 647 302 350 0627 0541

Skyros 5333 618 317 367 0069 0594

Zakynthos 490 568 293 339 0692 0597

pressures over the central Balkans and low pres-sures over Turkey (Akdag et al 2010) During hotsummer days they are at their strongest in theafternoon and often die down at night but some-times these winds last for days without a breakSimilar winds blow in the Adriatic and Ionianregions In the northern Aegean Sea the Etesiansblow as winds of northeasterly to northerly direc-tion Moving south in the central Aegean theyblow as winds of northerly direction while in thesouthern Aegean the Cretan and the CarpathianSea they blow as northwesterlies The same windsblow in Cyprus as westerlies to southwesterliesbeing more humid (Incecik and Erdogmus 1995Poulos et al 1997 Borhan 1998 Hamad et al 2006Klaic et al 2009)

The tides of the Aegean basin seem to followthe movements of those in the eastern Mediter-ranean generally Aegean currents generally are notsmooth whether considered from the viewpoint ofeither speed or direction They are chiefly influ-enced by blowing winds Water temperatures in theAegean are influenced by the cold-water masses oflow temperature that flow in from the Black Sea tothe northeast The sea surface temperature in theAegean ranges from about 60ndash77F (16ndash25C)varying with location and time of year

The turbulence intensity values for the selectedsites are calculated as the quotient of thestandard deviation to the average wind speed(σaverage WS) and are given in table 2 (Belu andKoracin 2009 Yu and Gan Chowdhury 2009) For

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 977

trying to determine the maximum wind sourcedenergy potential in the area (Poulos et al 1997Incecik and Erdogmus 1995 Klaic et al 2009Borhan 1998 Hamad et al 2006) Karamanis et al(2011) assessed the offshore wind power potentialof two sites Zakynthos and Pylos in the Ioniansea by using two years mean wind speed at 10 mThey found the annual mean wind speed at thesetwo sites as 57 and 58 ms respectively Offshorewind capacity factors of up to 48 for energy pro-duction were calculated with the existing offshoreturbines technology at a hub height of 100 m Thepresent study utilizes wind speed and directiondata from 10 buoy stations in Ionian and Aegeanseas to assess the offshore wind power potential inthese areas

The present paper has as main objective tothoroughly examine the wind characteristics formany offshore locations in the areas of Aegean andIonian seas as being areas with large differences inwind characteristics even more between two neigh-bouring sites by using the studied Weibull para-meters of the areas In addition this studyattempts to understand the behaviour of wind andthereafter energy yield at these stations using a5 MW rated power offshore wind turbine

2 Site and data description

The study utilizes measured wind speed and winddirection data at 3-hr intervals from 10 buoy sta-tions namely Athos E1M3A Lesvos MykonosPetrokaravo Santorini and Skyros in Aegean Seaand Kalamata Pylos and Zakynthos in Ionian SeaThe data collection period varied between 31 and11 years as can be seen from 6th column of thetable 1 below The latitude longitude the data col-lection periods are summarized in table 1 and thephysical locations are depicted in figure 1

Two types of surface buoys are used in theselected sites Oceanor Wavescan at AthosE1M3A and Pylos (figure 1 locations 1 2 7) andOceanor SeaWatch at Kalamata Lesvos MykonosPetrokaravo Santorini Zakynthos and Skyros(figure 1 locations 3 4 5 6 8 9 10) Wave-scan buoy1 is a metocean data collection buoy thatprovides wave height wave direction and meteoro-logical parameters sea surface temperature salin-ity and temperature profiles It is the ideal buoyfor deep-water measurements remote locationsand strong current conditions The buoy can alsobe fitted with numerous other sensors to satisfythe scientistrsquos specific configuration needs such as

1httpwwwgeoscoukservicesoceanWavescanasp

oxygen hydrocarbon gamma radiation measure-ment and an optical sensor for algae detectionSeawatch buoy2 is a multi-sensor wave directionaldata buoy capable of measuring wave height anddirection ocean current speed and direction mete-orological parameters sea surface temperaturesalinity and temperature and salinity profiles As avertically stabilised buoy with low pitchroll motionit is ideally suited for wave current and wind mea-surements in shallow water depth up to 500 m Forwater quality measurement the buoy can be fittedwith numerous other sensors satisfying the cus-tomerrsquos specific configuration needs such as oxy-gen hydrocarbon gamma radiation measurementand an optical sensor for algae detection

The data collection buoys used at the aforemen-tioned sites are shown in figure 2 These buoyswere instrumented to measure the air-pressureair-temperature wind speed and direction waveheight period and direction sea surface salin-ity and temperature surface current speed anddirection The anemometer used in both types ofbuoys is the Young 04106 (four-blade propelleranemometer)

The Aegean Sea an arm of the MediterraneanSea is located between the Greek peninsula on thewest and Asia Minor on the east while the islandof Crete can be taken as marking its boundary onthe south The Aegean is connected through thestraits of the Dardanelles the Sea of Marmara andthe Bosporus to the Black Sea It contains morethan 3000 islands with Crete as its largest islandand is considered as the home of the earliest Euro-pean civilization The Aegean Sea covers about214000 km2 in area and measures about 610 kmlongitudinally and 300 km latitudinally The searsquosmaximum depth is 3543 m east of Crete It alsohas a good connection to the Ionian Sea to thewest mainly through the strait lying between thePeloponnese peninsula of Greece and CreteThe rocks making up the floor of the Aegean aremainly limestone though often greatly altered byvolcanic activity that has convulsed the region inrelatively recent geologic times

The Ionian Sea is also an arm of theMediterranean Sea south of the Adriatic Sea Itis bounded by southern Italy including CalabriaSicily and the Salento peninsula to the west and bysouthwestern Albania and Greek peninsula to theeast containing a large number of Greek islands Itis connected with the Adriatic Sea to the north bythe Strait of Taranto The sea is one of the mostseismic areas in the world

North winds prevail in the Aegean Sea althoughfrom the end of September to the end of May

2httpwwwgeoscoukservicesoceanseabuoyasp

978 Haralambos S Bagiorgas et al

Table 1 Summary of site specific coordinates and wind power class

Latitude Longitude Depth DFL Period WPD Wind

Lactation (N) (E) (m) (km) Start date End date (years) (Wm2) class

Athos 3957prime843primeprime 2443prime208primeprime 212 2746 05252000 12312010 11 539 5

E1M3A 3546prime990primeprime 2454prime880primeprime 1440 4031 05282007 01012011 36 493 4

Kalamata 3658prime283primeprime 2205prime620primeprime 340 413 12311999 01012011 11 206 2

Lesvos 3909prime346primeprime 2548prime472primeprime 121 282 12311999 01012011 11 693 6

Mykonos 3730prime690primeprime 2527prime494primeprime 138 441 12311999 01012011 11 971 7

Petrokaravo 3735prime915primeprime 2333prime715primeprime 211 893 08282007 01012011 33 368 3

Pylos 3649prime533primeprime 2135prime762primeprime 1681 817 11092007 01012011 31 444 4

Santorini 3615prime510primeprime 2529prime769primeprime 314 921 12311999 01012011 11 590 5

Skyros 3906prime360primeprime 2427prime568primeprime 117 1392 08282007 01012011 33 540 5

Zakynthos 3756prime927primeprime 2036prime059primeprime 313 681 11082007 01012011 31 469 4

DFL = Distance from land Wind classes 2 = Marginal 3 = Fair 4 = Good 5 = Excellent 6 = Outstanding 7 = Superb

Figure 1 Map of Ionian and Aegean seas with the sites of the buoys 1 Athos 2 E1M3A 3 Kalamata 4 Lesvos 5 Mykonos6 Petrokaravo 7 Pylos 8 Santorini 9 Zakynthos and 10 Skyros

during the mild winter season these winds alter-nate with southwesterlies The Etesians are thestrong dry north winds of the Aegean Sea which

blow from about mid-May to mid-September Thebarometric conditions that cause these prevailingwinds are the existence of both high barometric

Offshore wind speed and wind power characteristics 979

Figure 2 Views of the Oceanor Wavescan (a) and Oceanor SeaWatch (b) instrumented buoys used at these stations

Table 2 Turbulence intensity values for the selected sites

Location WS3 WS10 SD3 SD10 Turbulence 3 Turbulence 10

Athos 473 547 343 396 0837 0724

E1M3A 530 614 275 318 0600 0518

Kalamata 366 424 235 272 0743 0642

Lesvos 584 675 341 393 0673 0582

Mykonos 670 776 356 412 0615 0531

Petrokaravo 467 542 275 319 0683 0589

Pylos 497 577 288 335 0674 0581

Santorini 558 647 302 350 0627 0541

Skyros 5333 618 317 367 0069 0594

Zakynthos 490 568 293 339 0692 0597

pressures over the central Balkans and low pres-sures over Turkey (Akdag et al 2010) During hotsummer days they are at their strongest in theafternoon and often die down at night but some-times these winds last for days without a breakSimilar winds blow in the Adriatic and Ionianregions In the northern Aegean Sea the Etesiansblow as winds of northeasterly to northerly direc-tion Moving south in the central Aegean theyblow as winds of northerly direction while in thesouthern Aegean the Cretan and the CarpathianSea they blow as northwesterlies The same windsblow in Cyprus as westerlies to southwesterliesbeing more humid (Incecik and Erdogmus 1995Poulos et al 1997 Borhan 1998 Hamad et al 2006Klaic et al 2009)

The tides of the Aegean basin seem to followthe movements of those in the eastern Mediter-ranean generally Aegean currents generally are notsmooth whether considered from the viewpoint ofeither speed or direction They are chiefly influ-enced by blowing winds Water temperatures in theAegean are influenced by the cold-water masses oflow temperature that flow in from the Black Sea tothe northeast The sea surface temperature in theAegean ranges from about 60ndash77F (16ndash25C)varying with location and time of year

The turbulence intensity values for the selectedsites are calculated as the quotient of thestandard deviation to the average wind speed(σaverage WS) and are given in table 2 (Belu andKoracin 2009 Yu and Gan Chowdhury 2009) For

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

978 Haralambos S Bagiorgas et al

Table 1 Summary of site specific coordinates and wind power class

Latitude Longitude Depth DFL Period WPD Wind

Lactation (N) (E) (m) (km) Start date End date (years) (Wm2) class

Athos 3957prime843primeprime 2443prime208primeprime 212 2746 05252000 12312010 11 539 5

E1M3A 3546prime990primeprime 2454prime880primeprime 1440 4031 05282007 01012011 36 493 4

Kalamata 3658prime283primeprime 2205prime620primeprime 340 413 12311999 01012011 11 206 2

Lesvos 3909prime346primeprime 2548prime472primeprime 121 282 12311999 01012011 11 693 6

Mykonos 3730prime690primeprime 2527prime494primeprime 138 441 12311999 01012011 11 971 7

Petrokaravo 3735prime915primeprime 2333prime715primeprime 211 893 08282007 01012011 33 368 3

Pylos 3649prime533primeprime 2135prime762primeprime 1681 817 11092007 01012011 31 444 4

Santorini 3615prime510primeprime 2529prime769primeprime 314 921 12311999 01012011 11 590 5

Skyros 3906prime360primeprime 2427prime568primeprime 117 1392 08282007 01012011 33 540 5

Zakynthos 3756prime927primeprime 2036prime059primeprime 313 681 11082007 01012011 31 469 4

DFL = Distance from land Wind classes 2 = Marginal 3 = Fair 4 = Good 5 = Excellent 6 = Outstanding 7 = Superb

Figure 1 Map of Ionian and Aegean seas with the sites of the buoys 1 Athos 2 E1M3A 3 Kalamata 4 Lesvos 5 Mykonos6 Petrokaravo 7 Pylos 8 Santorini 9 Zakynthos and 10 Skyros

during the mild winter season these winds alter-nate with southwesterlies The Etesians are thestrong dry north winds of the Aegean Sea which

blow from about mid-May to mid-September Thebarometric conditions that cause these prevailingwinds are the existence of both high barometric

Offshore wind speed and wind power characteristics 979

Figure 2 Views of the Oceanor Wavescan (a) and Oceanor SeaWatch (b) instrumented buoys used at these stations

Table 2 Turbulence intensity values for the selected sites

Location WS3 WS10 SD3 SD10 Turbulence 3 Turbulence 10

Athos 473 547 343 396 0837 0724

E1M3A 530 614 275 318 0600 0518

Kalamata 366 424 235 272 0743 0642

Lesvos 584 675 341 393 0673 0582

Mykonos 670 776 356 412 0615 0531

Petrokaravo 467 542 275 319 0683 0589

Pylos 497 577 288 335 0674 0581

Santorini 558 647 302 350 0627 0541

Skyros 5333 618 317 367 0069 0594

Zakynthos 490 568 293 339 0692 0597

pressures over the central Balkans and low pres-sures over Turkey (Akdag et al 2010) During hotsummer days they are at their strongest in theafternoon and often die down at night but some-times these winds last for days without a breakSimilar winds blow in the Adriatic and Ionianregions In the northern Aegean Sea the Etesiansblow as winds of northeasterly to northerly direc-tion Moving south in the central Aegean theyblow as winds of northerly direction while in thesouthern Aegean the Cretan and the CarpathianSea they blow as northwesterlies The same windsblow in Cyprus as westerlies to southwesterliesbeing more humid (Incecik and Erdogmus 1995Poulos et al 1997 Borhan 1998 Hamad et al 2006Klaic et al 2009)

The tides of the Aegean basin seem to followthe movements of those in the eastern Mediter-ranean generally Aegean currents generally are notsmooth whether considered from the viewpoint ofeither speed or direction They are chiefly influ-enced by blowing winds Water temperatures in theAegean are influenced by the cold-water masses oflow temperature that flow in from the Black Sea tothe northeast The sea surface temperature in theAegean ranges from about 60ndash77F (16ndash25C)varying with location and time of year

The turbulence intensity values for the selectedsites are calculated as the quotient of thestandard deviation to the average wind speed(σaverage WS) and are given in table 2 (Belu andKoracin 2009 Yu and Gan Chowdhury 2009) For

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 979

Figure 2 Views of the Oceanor Wavescan (a) and Oceanor SeaWatch (b) instrumented buoys used at these stations

Table 2 Turbulence intensity values for the selected sites

Location WS3 WS10 SD3 SD10 Turbulence 3 Turbulence 10

Athos 473 547 343 396 0837 0724

E1M3A 530 614 275 318 0600 0518

Kalamata 366 424 235 272 0743 0642

Lesvos 584 675 341 393 0673 0582

Mykonos 670 776 356 412 0615 0531

Petrokaravo 467 542 275 319 0683 0589

Pylos 497 577 288 335 0674 0581

Santorini 558 647 302 350 0627 0541

Skyros 5333 618 317 367 0069 0594

Zakynthos 490 568 293 339 0692 0597

pressures over the central Balkans and low pres-sures over Turkey (Akdag et al 2010) During hotsummer days they are at their strongest in theafternoon and often die down at night but some-times these winds last for days without a breakSimilar winds blow in the Adriatic and Ionianregions In the northern Aegean Sea the Etesiansblow as winds of northeasterly to northerly direc-tion Moving south in the central Aegean theyblow as winds of northerly direction while in thesouthern Aegean the Cretan and the CarpathianSea they blow as northwesterlies The same windsblow in Cyprus as westerlies to southwesterliesbeing more humid (Incecik and Erdogmus 1995Poulos et al 1997 Borhan 1998 Hamad et al 2006Klaic et al 2009)

The tides of the Aegean basin seem to followthe movements of those in the eastern Mediter-ranean generally Aegean currents generally are notsmooth whether considered from the viewpoint ofeither speed or direction They are chiefly influ-enced by blowing winds Water temperatures in theAegean are influenced by the cold-water masses oflow temperature that flow in from the Black Sea tothe northeast The sea surface temperature in theAegean ranges from about 60ndash77F (16ndash25C)varying with location and time of year

The turbulence intensity values for the selectedsites are calculated as the quotient of thestandard deviation to the average wind speed(σaverage WS) and are given in table 2 (Belu andKoracin 2009 Yu and Gan Chowdhury 2009) For

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

980 Haralambos S Bagiorgas et al

overwater wind speed measurements a preliminaryestimation of the gust factor G (which is definedas the ratio of the maximum gust speed to the sus-tained wind speed) could be done via the followingempirical formulation (Hsu 2001 2003a 2003b)

G = 1 + 2P

where G is the gust factor and P is the exponentof the power-law wind profile of a selected siteConsequently the gust factor value for the selectedsites as given by the aforementioned relationshipis G = 1246 at all sites as the wind shear expo-nent was 0123 with roughness length of 00016 mand roughness class of 042 (there is a uniformityin the characteristics of water body and there is nochange in surface roughness)

As preliminary assessment and magnitude ofwind power density calculated at 50 m aboveground level (AGL) Mykonos was classified as thesuperb windy site with class 7 winds and windpower density of 971 Wm2 Petrokaravo andKalamata with wind power densities of 368 and206 Wm2 were classified as the fair and marginalsites with wind power class of 3 and 2 respectivelyas mentioned in the last column of table 1 Allthe other sites were found to be between good andoutstanding

Figure 3 Monthly variation of wind speed at all thestations

3 Results and discussion

The wind speed and direction data analysiswas performed using the state-of-the-art softwarelsquoWindographerrsquo made specifically for wind dataanalysis (Windographer) The annual monthlyand three hourly mean wind speed statisticsthe Weibull parameters k and c using maxi-mum likelihood WAsP and least square methodsthe frequency distribution and finally the windenergy yield and plant capacity factors were deter-mined and are discussed here in the forthcomingparagraphs

31 Annual seasonal and diurnalvariation of wind speed

The annual mean wind speed values are usuallyused to get a rough estimate of wind energy yieldfrom a particular wind turbine and for a particu-lar site The annual mean wind speeds further helpin providing the decreasing or increasing trendsduring the coming years which are required toestimate the energy yield and thus the economicsof the wind power development in that region orarea In the present case the annual wind speedvalues were available for 11 years at only four

Figure 4 Diurnal variation of 3 hourly mean wind speeds atall the stations

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 981

locations out of the 10 used in the present work AtAthos the annual mean wind speeds were alwaysgreater than 45 ms at 3 and 10 m above watersurface with the exception of year 2003 The linearbest fit trends show increasing trends from years2000 to 2010 This implies that the wind powerdevelopment and utilization at Athos will be agood option for profit making looking from theinvestment point of view

In Lesvos the annual mean wind speed wasalways above 5 ms except in the year 2005 and2007 when it was a bit less Overall the linear

trend showed a slightly decreasing pattern whichmay be due to some operational and maintenanceprocedures but need to be verified with more yearsof data from nearby historical meteorological sta-tion The annual mean wind speed was above 6 msin all years at 10 m At Mykonos the annual meanwind speed was most of the times above 65 msat measurement heights of 3 and 10 m above watersurface level and hence is expected to be muchhigher at 100 to 120 m above the water surfacelevel In 2001 and 2006 annual mean wind speedsof more than 8 ms were noticed at this site The

Table 3 Summary of Weibull parameters fits and the R2 va1ues for different methods at 10 m

Weibull Weibull Mean R2

Algorithm k c (ms) (ms)

(a) Athos

Maximum likelihood 124 581 542 090

Least squares 093 641 663 077

WAsP 151 621 560 086

(b) E1M3A

Maximum likelihood 199 691 612 094

Least squares 186 702 623 093

WAsP 173 651 580 087

(c) Kalamata

Maximum likelihood 149 466 421 094

Least squares 123 490 458 090

WAsP 188 505 449 089

(d) Lesvos

Maximum likelihood 162 745 667 079

Least squares 114 825 788 059

WAsP 274 852 758 069

(e) Mykonos

Maximum likelihood 178 859 764 070

Least squares 120 953 896 043

WAsP 385 978 884 061

(f) Petrokaravo

Maximum likelihood 172 606 541 092

Least squares 163 611 547 093

WAsP 248 677 600 073

(g) Pylos

Maximum likelihood 177 648 576 095

Least squares 171 651 580 095

WAsP 185 657 583 094

(h) Santorini

Maximum likelihood 182 721 641 093

Least squares 132 787 724 073

WAsP 233 771 683 093

(i) Skyros

Maximum likelihood 170 691 616 094

Least squares 157 702 631 094

WAsP 202 733 649 088

(j) Zakynthos

Maximum likelihood 171 636 567 095

Least squares 164 640 572 096

WAsP 205 678 601 088

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

982 Haralambos S Bagiorgas et al

annual mean wind speeds showed increasing trendsin the coming years At Santorini a visible decreas-ing trend of around 25 per year was observedin the annual mean wind speeds but the annualmean wind speed was always greater than 6 msand even above 7 ms in some years

The seasonal trends at Athos showed higherwind speeds of about 5 ms from January toApril and then decreasing towards June and againincreasing towards end of the year reaching above6 ms at both measurement heights as shown infigure 3(a) This implies that a wind turbine with

Figure 5 Frequency distribution and Weibull fits using different methods

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 983

cut-in-speed of 4 ms can produce energy duringall the months except from May to July Higherwinds of more than 5 ms were observed dur-ing all the months except in May at E1M3A sta-tion (figure 3b) At all the measurement stationshigher winds were observed during winter time andrelatively lower during summer time at both theheights Clearly higher winds were observed at10 m compared to those at 3 m above water surfacelevel

The diurnal variation of 3-hourly mean windspeed at 3 and 10 m above water surface level forall the buoy stations is shown in figure 4 At all ofthese stations the mean wind speeds at 10 m werearound 15ndash16 higher than those at 3 m mea-surement height At Athos Petrokaravo and Syrosstationsrsquo lower values of 3-hourly mean wind speedswere observed 12 to 15 hours and higher at otherhours of the day At all the remaining stations themean wind speeds were higher from 12 to 18 hoursin general and lower during remaining hours of theday as can be seen from figure 4

32 Weibull and frequency distribution analysis

The speed of the wind is well represented byWeibull distribution function and it is widely usedfor wind power estimation and resource assess-ment In the present scope of the work Weibull

parameters k and c were determined using Max-imum likelihood least squares and WAsP meth-ods and for entire dataset at both 3 and 10 mmeasurement heights For the sake of space opti-mization the k and c values are presented for mea-surements made at 10 m In most of the casesWAsP method provides highest values of shapeparameter k and scale parameter c as can beobserved from table 3 The Weibull parameterswere used to find the best fit curves and comparedwith the measured wind speed in different windspeed bins are shown in figure 5(andashj) At Athosthe most close agreement between the Weibull dis-tribution function and the measured wind speedwas obtained by maximum likelihood method withR2 = 090 while the least close agreement by leastsquare method with R2 = 077 The comparisonis shown in figure 5(a) Almost at all the stationsmaximum likelihood method gave the best compa-rable values of k and c while least square methodproved the next best and WAsP at number three

At Athos the wind speed was above cut-in-speedof 4 ms for 524 of the times during the data col-lection period and the predominantly blown fromNS (gt45) At E1M3A the remained above cut-in-speed was appeared for more than 71 of thetimes and came from WN directions for 50 of thetimes At Kalamata Mykonos and Petrokaravothe wind remained above cut-in-speed for morethan 48 67 and 60 of the times and was

Table 4 Effect of hub height on wind energy yield and plant capacity factor

Athos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 744 737 728 719

Zero output () 2659 269 2709 2753

Rated output () 1561 1531 1463 1439

Mean power (kW) 141260 139170 136880 134360

Annual energy (MWh) 123745 121909 119907 117695

PCF () 283 278 274 269

Mykonos

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 1054 1043 1031 1018

Zero output () 1286 1298 1308 1319

Rated output () 3598 354 3396 3336

Mean power (kW) 24110 23884 23635 23351

Annual energy (MWh) 211200 209225 207039 204559

PCF () 482 478 473 467

Santorini

Hub height (HH) 120 m 110 m 100 m 90 m

WS at HH (ms) 879 87 86 848

Zero output () 1352 1368 1381 14

Rated output () 1814 1763 167 1615

Mean power (kW) 18564 18272 17952 17596

Annual energy (MWh) 162623 160063 157260 154141

PCF () 371 365 359 352

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

984 Haralambos S Bagiorgas et al

found to be blowing predominantly from NE NndashNSndashNE and NW respectively At remaining sta-tions the wind speed remained above cut-in-speedfor more than 62 of the times The wind directionwas mostly found to be from N

33 Wind energy yield estimation andeffect of hub height

The energy yield was obtained using an offshorewind turbine of 5 MW rated power from REPowerwith electrical blade angle adjustment ndash pitchand speed control (Windographer) The technicalspecifications of the chosen wind turbine wererotor diameter=126 m cut-in-speed=4 ms rated

speed = 13 ms cut-out-speed = 30 ms and dif-ferent hub heights (HH) ranging from 90 to 120 mwith an increment of 10 m The overall losses weretaken as 177 which include down time lossesof 6 array losses of 5 icingsoiling losses ofanother 4 and finally the other unaccountablelosses of 4 Of the 10 buoy stations the energyyield was estimated for all the stations but detailsare given only for Athos Mykonos and Santorinistations in table 4

At Athos the chosen turbine could produce12375 MWh of electricity annually with a plantcapacity factor (PCF) of 283 and with zero out-put of 266 and rated output of 1561 of thetimes during the year with 120 m hub height (HH)as summarized in table 4 At 120 m HH the wind

Figure 6 Annual mean wind energy yield at HH = 90 m

Figure 7 Annual mean wind speed and PCF at HH = 90

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 985

Figure 8 Percent rated and zero energy output

speed was 744 ms Slight increases of 19 17and 15 were found in energy yield while increas-ing the HH from 90 to 100 m 100 to 110 m andfrom 110 to 120 m while the corresponding changesin PCF were 05 04 and 05 At Mykonos sta-tion which is the best location from energy har-nessing point of view the increases in energy yieldwere 02 11 and 09 with changes in HHheights from 90 to 100 m 100 to 110 m and 110to 120 m respectively However the PCFrsquos were467 473 478 and 482 corresponding toHHrsquos of 90 100 110 and 120 m as listed intable 4 At Mykonos as given in the aforemen-tioned table the rated output was obtained formore than 35 of the times and the zero out-put was about 13 of the times during the yearThe annual mean wind speeds at respective HHrsquoswere always greater than 10 ms At Santorini sta-tion the annual mean wind speeds at HHrsquos of 90to 120 m were found to be varying from 848 to879 ms as can be seen from table 4 The chosenwind turbine was able to produce 15414 1572616006 and 16262 MWh of electricity annually cor-responding to different HHrsquos The correspondingPCFrsquos were found to be 352 359 365 and371 respectively The rated output at this sitewas achieved more than 16 of the times duringthe year while the zero output was observed tobe around 135 of the times as summarized intable 4

The annual energy output at all the sites iscompared in figure 6 and PCFrsquos in figure 7 forthe designed HH of 90 m Maximum energy of2046 GWh was obtained at Mykonos station andthe minimum of 717 GWh at Kalamata as shownin figure 6 The second and third best sites withhigher energy yields of 1682 and 1551 GWh were

Lesvos and Santorini respectively Similarly thecorresponding higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini The annual wind speeds at 90 m HH arealso depicted in this figure On the other hand thepercent maximum and minimum zero and ratedenergy output was observed at Kalamata at 90 mHH as can be seen from figure 8 The maximumrated output was observed at Mykonos at this HH

4 Conclusions

The study found that most of the stations couldbe used for the development of technically fea-sible wind farms but consideration of the depthof the sea and the distance from the land to thesites needed to complete the benefit of the studySpecifically following main findings could be of useto readers and wind power seekers in Ionian andAegean seas

bull Maximum mean wind speed of 78 ms was foundat Mykonos and the minimum of 42 ms at Kala-mata At all stations the overall mean wind speedwas more than 5 ms

bull The seasonal trends were quite visible withhigher values in winter time and lower in summerseason with minimum in May and June

bull The diurnal pattern of 3-hourly mean windspeed showed usually higher valued during 12 to21 hours and relatively during rest of the hoursof the day

bull The wind was found to be flowing predominantlyfrom N NE NW W WN E and EN directions

bull Based on the availability of wind speed data ofaround 11 years the annual wind speed trends

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

986 Haralambos S Bagiorgas et al

were studied for Athos Lesvos Mykonos andSantorini Slightly decreasing trends were foundat Lesvos and Santorini but increasing at Athosand Mykonos

bull The wind speed distribution was found to bewell represented by Weibull parameters obtainedusing Maximum likelihood method compared toWAsP and Method of Moments

bull At all stations the chosen wind turbine could pro-duce energy for more than 70 of the time Thewind speed distribution was found to be well rep-resented by Weibull parameters obtained usingMaximum likelihood method compared to WAsPand Method of Moments

bull Maximum energy of 2046 GWh was obtainedat Mykonos station and the minimum of717 GWh at Kalamata The second and thirdbest sites with higher energy yields of 1682and 1551 GWh were Lesvos and Santorinirespectively

bull Similarly the higher PCFrsquos of 467 384 and352 were obtained at Mykonos Lesvos andSantorini

bull The effect of HH was very minimal At all thesites studied the wind energy was found toincrease only by less than 1 to little more than1 for each increment of 10 m This necessitatesthe measurements of wind speed at least at threesites viz Mykonos Lesvos and Santorini at dif-ferent heights and up to 120 m above the watersurface These measurements will provide moreaccurate local wind shear exponents and hencethe exact wind speed at hub height

Acknowledgements

The authors wish to acknowledge the support ofthe Research Institute of King Fahd Universityof Petroleum and Minerals Dhahran Saudi Ara-bia and the Hellenic Centre for Marine Research ndashPoseidon Team for the access to the Centrersquos buoysdatabase

References

Akdag S A Bagiorgas H S and Mihalakakou G 2010 Useof two-component Weibull mixtures in the analysis ofwind speed in the Eastern Mediterranean Appl Energ87 2566ndash2573

Bailey B Brennan S Kinal B Markus M andKreiselman J 2002 Long islands offshore wind energydevelopment potential Phase 1 Preliminary AssessmentAWS Scientific Inc

Belu R and Koracin D 2009 Wind characteristics and windenergy potential in western Nevada Renew Energ 342246ndash2251

Bilgili M Yasar A and Simsek E 2011 Offshore wind powerdevelopment in Europe and its comparison with onshorecounterpart Renew Sust Energ Rev 15 905ndash915

Borhan Y 1998 Mesoscale interactions on wind energy poten-tial in the northern Aegean region A case study RenewSust Energ Rev 2(4) 353ndash360

Dhanju A Whitaker P and Kempton W 2008 Assessing off-shore wind resources An accessible methodology RenewEnerg 33 55ndash64

Dvorak M J Archer C L and Jacobson M Z 2010 Californiaoffshore wind energy potential Renew Energ 35 1244ndash1254

Esteban M D Diez J J Lopez J S and Negro V 2011 Whyoffshore wind energy Renew Energ 36 444ndash450

EWEA 2010 statistics Offshore and eastern Europenew growth drivers for wind power in Europehttpwwweweaorgindexphpid=60ampno cache=1amptxttnews[tt news]=1896amptx ttnews[backPid]=1ampcHash=8b64626e4bf6996eea71ec3c08994b0a (accessed onFebruary 2 2011)

Gaudiosi G 1994 Offshore wind energy in the Mediterraneanand other European seas Renew Energ 5(1) 675ndash691

Hamad N Millot C and Taupier-Letage I 2006 The surfacecirculation in the eastern basin of the Mediterranean SeaSci Mar 70(3) 457ndash503

Hsu S A 2001 Spatial variations in gust factor across thecoastal zone during Hurricane Opal in 1995 Natl WeaDig 25(1ndash2) 21ndash23

Hsu S A 2003a Estimating overwater friction velocity andexponent of power-law wind profile from gust factorduring storms J Waterw Port C-ASCE 129(4) 174ndash177

Hsu S A 2003b A gust-factor criterion for rapid deter-mination of atmospheric stability and mixing heightfor overwater dispersion estimates Natl Wea Dig 2770ndash74

Incecik S and Erdogmus F 1995 An investigation of the windpower potential on the western coast of Anatolia RenewEnerg 6(7) 863ndash865

Karamanis D Tsabaris C Stamoulis K and GeorgopoulosD 2011 Wind energy resources in the Ionian Sea RenewEnerg 36 815ndash822

Klaic Z B Pasaric Z and Tudor M 2009 On the inter-play between sea-land breezes and Etesian winds over theAdriatic J Marine Syst 78(1) S101ndashS118

Manwell J F Rogers A L McGowan J G and Bailey B H2002 An offshore wind resource assessment study for NewEngland Renew Energ 27 175ndash187

Manwell J F Elkinton C N Rogers A L and McGowan J G2007 Review of design conditions applicable to offshorewind energy systems in the United States Renew SustEnerg Rev 11 210ndash234

Martin Mederos A C Medina Padron J F and Feijoo LorenzoA E 2011 An offshore wind atlas for the Canary IslandsRenew Sust Energ Rev 15 612ndash620

Musial W and Butterfield S 2004 Future for offshorewind energy in the United States NRELCP-500ndash36313Energy Ocean 2004 Palm Beach Florida June 28ndash29

Musial W and Ram B 2010 Large-scale offshore wind powerin the United States ndash assessment of opportunities andbarriers September 2010 NRELTP-500-40745

Pimenta F Kempton W and Garvine R 2008 Combin-ing meteorological stations and satellite data to evaluatethe offshore wind power resource of southeastern BrazilRenew Energ 33 2375ndash2387

Poulos S E Drakopoulos P G and Collins M B 1997 Sea-sonal variability in sea surface oceanographic conditionsin the Aegean Sea (Eastern Mediterranean) An overviewJ Marine Syst 13(1ndash4) 225ndash244

Rogers A L Manwell J F and McGowan J G 2003 A year2000 summary of offshore wind development in the UnitedStates Energ Convers Manage 44 215ndash229

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References

Offshore wind speed and wind power characteristics 987

Still D 2001 Offshore wind at Blyth Renew Energ 24 545ndash551Stoutenburg E D Jenkins N and Jacobson M Z 2010 Power

output variations of co-located offshore wind turbines andwave energy converters in California Renew Energ 352781ndash2791

Windographer Version 221 httpwwwwindographercom

Yu B and Gan Chowdhury A 2009 Gust factors and turbu-lence intensities for the tropical cyclone environment JAppl Meteorol Clim 48 534ndash552

Zhixin W Chuanwen J Qian A and Chengmin W 2009The key technology of offshore wind farm and its newdevelopment in China Renew Sust Energ Rev 13216ndash222

MS received 26 September 2011 revised 6 January 2012 accepted 12 March 2012

  • Offshore wind speed and wind power characteristics for ten locations in Aegean and Ionian Seas
    • Abstract
      • Introduction
      • Site and data description
      • Results and discussion
        • Annual seasonal and diurnal variation of wind speed
        • Weibull and frequency distribution analysis
        • Wind energy yield estimation and effect of hub height
          • Conclusions
          • References