Assessment of the wave potential at selected hydrology and coastal environments around a tropical island, case study: Mauritius

Assessment of the wave potential at selected hydrology and coastal environments around a tropical... Waves are the dominant influence on coastal morphology and ecosystem structure of tropical islands. The geographical positioning of Mauritius near to the Tropic of Capricorn ensures that the eastern regions benefit from the persistent southeast trade winds which influence the incoming surface waves. In this study, we present the high dependence of the wave regimes of windward offshore site on the trade winds. The higher occurrence of incoming waves in the winter season directed in the southeast direction indicates that the trade winds are more prevalent in the winter season. Storms within the extratropical South Atlantic, Indian and Pacific oceans generally propagate towards the east such that extratropical South Atlantic swell energy spreads through the entire Indian Ocean. Since waves are very directional and tend to get shadowed by land masses, Mauritius situated in the line of sight from those sources end up in the shadow region due to the geographical location of Reunion island. In this study, we support the explanation on how the western region of the island gets influenced by episodic swell events. A detailed wave energy resource assessment is provided for different targeted coastal environments around the island. It is revealed that the mean wave power observed in the summer season at one of the sites can attain 28.8 kW/m and is found to be lower as compared to the winter season (31.7 kW/m). Keywords Wave climate · Wave energy · Tropical Island · Statistical analysis · Resource assessment · Indian Ocean Introduction representing a growth of 0.1% as compared to the preced- ing year [1]. The increasing population dynamics fuels the Mauritius, a small island located in the tropical belt of the need for increasing energy requirement. Consequently, this southwestern Indian Ocean basin, enjoys a tropical maritime has led to an increase of 1% in total primary energy require- climate that spans over two seasons: summer and winter. ment during this 1 year interval to reach a value of 1550 ktoe The population of the island in 2016 stood at 1,263,820, [2]. Typical of small island developing states (SIDS), Mau- ritius relies heavily on imported fossil fuels (with a share of 85%), reflecting on the vulnerable and volatile economy Electronic supplementary material The online version of this of the country. The electricity generation for the year 2016 article (http s://doi.org/10.1007 /s400 95-018-0259 -7) contains amounted to 3042 GWh, with a share of 78.2% attributed supplementary material, which is available to authorized users. to imported fuel oil and coal and the remaining percentage * Jay Rovisham Singh Doorga (21.8%) coming from renewable energy sources compris- jay.doorga927@gmail.com ing bagasse (16.3%), hydro (3.3%), photovoltaic (1%), wind (0.6%) and Landfill gas (0.6%) [ 2]. The huge renewable Physical Oceanography Unit, Mauritius Oceanography energy potential of the island for sustainable energy produc- Institute, Avenue des Anchois, Morcellement de Chazal, Albion, Mauritius tion has been recognized and bold measures are being taken by policy makers in the form of international partnerships Physical Oceanography/Marine Geoscience Unit, Department of Continental Shelf and Maritime Zones Administration and supportive regulatory frameworks to encourage inde- and Exploration, Prime Minister’s Office, Port -Louis, pendent power producers (IPPs) invest in the green sector. Mauritius The long-term vision of the Government of Mauritius is to Technical Unit, Mauritius Oceanography Institute, Avenue des Anchois, Morcellement de Chazal, Albion, Mauritius Vol.:(0123456789) 1 3 136 International Journal of Energy and Environmental Engineering (2018) 9:135–153 increase the percentage of electricity generation from renew- wave climate imposed by estuarine environment on waves able energy resources to 35% by 2025 [3]. coming from deep water is worth exploring. The authors Presently, the generation of electricity in Mauritius from believe that information presented in this paper will be help- wave resources is inexistent. Wave energy is acknowledged ful to the Government or any relevant organization in mak- as being a reliable renewable energy source whose output ing an informed decision with regard to major investments can be predicted in both spatial and temporal scales [4]. In for harnessing the wave resource for domestic use at the addition to displacing fossil fuels, which inadvertently help investigated sites. reduce greenhouse gas emissions and improve energy secu- rity, wave energy holds the benefit of being relatively less intermittent in nature than solar and wind. The heterogeneity Material and methods associated with cloud systems and prevailing wind condi- tions results in fluctuations from solar and wind power gen- Geology and climatology erations. Consequently, the implementation of wave energy ◦ ◦ extraction devices in the near-shore and off-shore regions of Mauritius (20 10 S, 57 30E), located in the south western the island would help reduce instabilities occurring in the Indian Ocean near the southern end of the Mascarene Ridge, national grid by conditioning power and smoothing fluctua- has an approximated surface area of 1859 km (Fig. 1). The tions. The success of renewable energy lies in its diversi- island has an Exclusive Economic Zone (EEZ) of about 2.3 2 2 fication and the inclusion of wave energy can be a strong million km (including approx. 400,000 km jointly man- component in the renewable energy mix which would benefit aged with the Seychelles) that extends over the islands of to the local community at large. Moreover, the proximity of Rodrigues, Agalega, Cargados Carajos shoals, Chagos inland and coastal regions to the sea (typically less than 30 Archipelago and Tromelin. Of volcanic origin, starting with km radius) holds the benefit of electricity transmission and the Breccia Series 10–7.8 million years (my) B.P., the geo- access even to remote locations on the island. morphological features of the island comprise an elevated Conversion of wave resources can contribute substantially central plateau (about 500 m above sea level) surrounded by to the electricity demand of several islands. The Orkney a chain of mountains and some isolated peaks. The eleva- archipelago, situated in the north of Scotland is estimated tion of the terrain decreases gently from the central plateau to have a mean wave power of 10–25 kW/m and has been towards the coastal edge. The coastline measures about 200 earmarked to contribute 550 MW to the total 1.6 GW wave km long and is composed of different shore types (catego- and tidal energy capacity from twelve leased sites by 2020 rized as sandy shores, rocky shores, muddy shores, mixed [5]. An assessment of the near-shore wave energy potential shores, calcareous limestone shores, cliffs and coastal wet- for the southern pacific islands of Fiji showed that mean lands) whose landforms are related to the coastal geomor- wave power of 9.81 kW/m was detected at a depth of 15 m phologic features. The shoreline is ringed by fringing coral in the west of the main island while energy flux of around reefs which enclose a lagoon having an approximated area 28.78 kW/m was identified near Kadavu island at a depth of 243 km [11]. With an annual growth rate of about 2 mm, of 18 m [6]. The Cape Verde Islands, located off the coast living coral reef formations resist destruction from sea action of West Africa in the central Atlantic Ocean, witnessed a by constantly growing due to the preference for light and wave power exceeding 7 kW/m in coastal areas neighbor- sediment-free waters [11]. ing the islands, with the summer season recording around The climatological regime of the island, characterized as half the wave potential of the winter season [7]. The island mild tropical maritime, is dictated by alterations of the two of El Hierro, situated south west of the Canary Islands in seasons: Summer (November–April) and winter (May–Octo- the Atlantic Ocean estimates its wave potential to be of the ber). The summer season is distinguished as warm and order of 25 kW/m, ree fl cting on the possibility of harnessing humid while the winter season is generally cool and dry. energy from waves [8]. Microclimatic conditions exist spatially in different locali- The aim of the present study is to investigate the wave ties attributed to the varying atmospheric conditions on the resource potential of open water sites, not too distant for island. The wave patterns on the coast of Mauritius is influ- electricity transmission inland and located outside of the enced all year-round by the westerly and persistent south- coral reef platform of the island. Despite the fact that easterly trade winds which is ensured by the geographical refracted ocean wave energies may be great at estuarine location of the island near to the Inter Tropical Convergence sites near wide entrances to the sea, significant energy loss Zone (ITCZ). occurs when entering estuaries, either by refraction around split platforms or through breaking on entrance shoals [9]. Waves entering basins may dominate the spectral estimates at sites close to the ocean [10]. Consequently, the complex 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 137 Fig. 1 a Location of three earmarked sites considered in Mauritius nified view of the two stations located just outside the coral reef plat- for the assessment of wave potential. b Geographical location of the form on the western side of the island tropical island of Mauritius, southwest of Indian Ocean basin. c Mag- west of the island by the presence of the mountains located Coastal hydrodynamics in the south western region. Typically, wind waves in open sea are between 0.5 m during summer and 3 m during winter The geographical position of the island ensures that it ben- with a period of 3–11 s [12]. Significant wave heights in efits from the South Equatorial Current (SEC) flowing from south, southwest (SSW) and southeast (SE) of the island west to east throughout the duration of the year. Addition- are of the order of 1.5–2.5 m in summer and 2.5–3.5 m in ally, the trade winds from anticyclones of the south Indian winter [13]. High waves reaching 3–5 m with a wave period Ocean which moves from east to west influence the coastal of 12–20 s traveling long distances, reaching the southern hydrodynamics and offshore current systems of the island. region of the island as swells with little loss in energy are These exhibit seasonal variations with the winter season generated by the deep low pressure systems in the high lati- experiencing relatively stronger currents in magnitude due to tudes [12]. the strong prevailing South Easterly trade winds. The speeds of the current increase as it passes through the channels situ- ated within the Mascarene Plateau resulting in the forma- Site selection tion of strong gyres on the leeward side. When it reaches Madagascar, part of the current flows to the north to feed the Three study sites (Fig. 2) were selected based on diversified Agulhas current in the Mozambique Channel and the East hydrologic and coastal environments, reflecting the distinct Africa current while the other part flows southwards along wave regimes prevailing at these localities. The impact of the Madagascan coast forming an anticyclone gyre further three different types of coastal hydrological environments south [12]. During the South west monsoon, the SEC feeds (Sandy shore with reef, estuary and mixed shore with reef), into the Somali current along the East coast of Africa. geographical location in connection with the prevailing The west coast of the island is relatively more protected southeast trade winds and swells coming from the south west from wind and wave action than the east as the body of the on the wave climate around the island are explored. island creates a lee whose effect is strengthened in the south 1 3 138 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Fig. 2 Aerial view depicting the coastal environments of the three sites under investigation in this study. The exact location for the deployment of the Wave and Tide Recorder (WTR) on the western flanks of the island and the waverider buoy on the eastern side are presented in the figure Flic en Flac is a sandy shore (with reef) composed essen- good balance of sands and rocks, generally interspersed. tially of sediments. The sediments, of carbonate origin are Located in the northeastern part of the island of Mauritius, made up of coral fragments, crustaceans and molluscan Roches Noires is one region where the highest waves are shells among others. The sandy beach that it makes from the usually observed off the reefs. This eastern region of the adjacent coral reefs constitutes one of the most economically island is under the influence of the southeast trade winds important and environmentally sensitive coastal habitats of throughout the year which provide the most substantial the Republic of Mauritius. Indeed, with the effect of climate wave action that dominate the depositional dynamics [18]. change, this fragile ecosystem is at risk of damage due to The deployment of the instruments at those exact geo- the impacts of sea level rise, warming of air and sea, and graphical coordinates was based on ease of access to the increased storminess [14]. sites, human activity level near these regions and sensitiv- The wave characteristics on the estuarine environment ity of the apparatuses in the water depths placed. Other of Tamarin is also studied. The influence of ocean waves locations were also considered but dropped primarily due may extend several kilometers into an estuary [15] but its to safety issues arising from the fact that divers would not effectiveness generally decreases with distance from the be able to access the devices easily. For the regions of opening to the ocean [16]. The location of Tamarin bay on Flic-en-Flac and Tamarin which have significant human the southwest coast is angled so that the winds blow offshore activity levels, we chose to investigate sites relatively away such that incoming swells have to bend sharply around the from these influences for ease of deployment and to get coast before hitting the reef [17]. We choose to study the minimum disturbance of wave records from anthropogenic wave characteristics in the Tamarin estuary near the center activities. A description of the four wave measurement of the wide entrance to the sea due to high refracted wave sites considered for deployment is provided in Table  1. energies converging at this site. More details on the regions of interest are provided in the Roches Noires is a mixed shore (reef) composed essen- following sub-section. tially of sediments and pebbles. It constitutes a fairly 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 139 Table 1 Description of the four wave measurement sites in terms of period of deployment, radial distance from the shore (d ) , water depth shore (Dep) and human activity levels Site Period d (m) Dep (m) Human activity shore 1 Flic-en-Flac 2 Apr–20 Apr 2007 490 12 Popular tourist region 2 Tamarin 28 Mar–22 Apr 2014 935 10 Intermediate human activity 3 Roches Noires 9 Feb 2012–10 Nov 2014 1228 46.2 Some human activity apparatus were detected. Also, selection of years was based Materials and methods on a number of factors including the wave climate prevailing during the considered period. It was ensured that the selected The Seawatch mini II Waverider buoy, designed by Fugro years for which the WTR was deployed contained no data Oceanor, is intended to follow the motion of the sea surface recorded in cyclonic weather conditions, which would inevi- waves. This results in exact measurement of the water sur- tably lead to biased results when it comes to selecting the face and can be employed to derive directional measurement most energetic region for tapping in surface wave energy. of the surface wave field [19]. Solid-state accelerometers are The WTR and Waverider buoy were configured with a sam- used to represent accelerations on three orthogonal axes. pling frequency of 1 Hz to sample data for 1024 s. A 1 Hz Integration with simultaneous heading and tilt readings frequency ensures that waves of up to 0.3 Hz (3.33 s) can be performed on-board using the ’wavesense’ computer unit accurately recorded. Owing to the fact that waves resulting installed in the buoy, results in accelerations being resolved from meteorological events lie in the period range of 7–15 s, to heave (vertical), east and north axes. The design ensures the selected sampling frequency would make optimum use that it does not influence the active frequency range for sur - of memory and batteries for such a long period deployment. face waves, also minimizing any damping or phase shift on the signals. As stated by Joosten [20], the mooring design Theoretical background for floating wave buoys must make sure that they allow the latter to respond unimpeded to wave frequencies while pro- Significant wave height H and wave power P are two viding the sufficient restoring force to resist tidal flow and m0 w important parameters when it comes to assessing the wave drift forces. In this study, the buoy was moored using a 600 energy potential of a particular sea state. The significant kg bottom weight attached to a 45 m polypropylene braided wave height is defined as the average wave height of one- rope, having its end connected to a 15 m rubber cable. The third of the highest waves in a wave record. For two regions: waverider has a lead acid battery bank 12 V, 62 Ah × 4 Flic-en-Flac and Tamarin, spectral data have been collected (248 Ah) coupled to an optional lithium battery bank rated using the WTR. The spectral estimate of significant wave 1088 Ah. Also included in the design are solar panels with height can be estimated from the zeroth moment of variance a rating of 12 W × 6 (72 W). The buoy was secured in water spectrum m as follows: depth of 46.2 m at a distance of 1228 m from the shore of Roches Noires. Measurements started at this site on the 9th H = 4 m . (1) m0 0 of February 2012 and lasted for 32 months. The buoy was For the non-directional variance density spectrum, the spec- retrieved on the 10th of November 2014. Cyclonic weather tral moment of nth order, m , is calculated using the follow- conditions resulting from the passage of cyclone ‘Imelda’ n ing equation. near the vicinity of the coast of Mauritius from the 6th to 16th of April 2013 resulted in extreme wave values recorded. Technical issue was encountered by the device in the period n m = f E Δf , (2) n i i from 16th October 2013 to 7th November 2013. i=1 A Valeport Wave and Tide recorder (WTR) was deployed where f is the i th frequency corresponding to the n th order, in the month of April at two sites around the island: Flic- i Δf is the frequency increment and E is the sea surface vari- en-Flac and Tamarin. WTR sensor has a precision resonant ance density over a range of wave frequencies i. quartz crystal and optional strain gauge with an accuracy The energy period T which is the variance-weighted of ± 0.01 m for water level measurements. Archived wave mean period of the non-directional variance density spec- measurements taken for the month of April by the Mauritius trum is calculated from the spectral moments as follows: Oceanography Institute has been chosen owing to important wave characteristics taking place in that specific month of −1 T = . the year. The selection of the years for analysis was primarily (3) based on the fact that no missing wave records due to faulty 1 3 140 International Journal of Energy and Environmental Engineering (2018) 9:135–153 The wave power (kW/m) in deep water can be calculated Procedure adopted using significant wave height and energy period as follows: The flowchart portraying the adopted strategy in this paper 2 2 is presented in Fig.  3. The intention of the current study P = H T = 0.491H T . (4) w m0 e m0 e involves assessing the wave potential at selected hydrol- ogy and coastal environments around Mauritius. Conse- Since the points investigated lies outside the coral reef bar- rier in open water and are found in water depths exceeding quently, two sites having fairly distinct wave climate and located in the western part of the island have been selected. 10 and 200 m from the shore, the use of Eq. 4 is justified. For the Roches Noires dataset, information on spectral These include the sandy shore with reef region of Flic-en- Flac and the estuarine environment of Tamarin. Significant moments or spectral shape was not provided, and therefore, T must be estimated from available dataset and one approach is wave height ( H ) and energy period ( T ) measur ements m0 e have been recorded at the two sites through the deploy- to assume the following: ment of a Wave and Tide recorder for the time frames: 2 T = T , e p (5) April–20 April 2007 (Flic-en-Flac) and 28 March 22 April where  is a coefficient whose value depends on the shape 2014 (Tamarin). The selected time interval for the data as of the wave spectrum (0.86 for a Pierson–Moskowitz spec- mentioned previously is based on suitability of data available trum and increasing towards unity with decreasing spectral (no missing wave records and no data selected in cyclonic width) [21]. In the spectrum formulation, it is assumed that weather conditions), safety issues for divers to access instru- the wind is blowing steadily for a long time and over a large ment, human activity level which may influence the wave area such that the waves attain a point of equilibrium with measurements and important wave processes taking place the wind. The conversion from peak period (T ) to energy p in that specific month of the year. On the eastern region of period (T ) was enable through the approximation: the island, the mixed shore with reef site of Roches Noires has been chosen for further investigation. Significant wave T ∼ 0.86T . (6) e p height and peak period ( T ) measurements have been By combining Eqs. 4 and 6, the total wave energy resource at acquired from the wave buoy for the time frame 9 February Roches Noires can be estimated. However, besides estimat- 2012–10 November 2014. ing the wave energy potential at the sites, it is of interest to Due to the temporal variability of the phenomena under analyze the temporal variability of the distributions. Steady study, we attempt to compare the wave regimes of the three wave energy flux are preferred to those having unsteady selected sites on the same time scale. Consequently, April wave regimes due to the fact that they are more reliable and datasets for the years 2012–2014 recorded at Roches Noires display greater efficiency [22]. have been extracted and averaged (corresponding to day and Two coefficients that were proposed by Cornett [ 23] are time of record) to obtain a single ’significant wave height’ employed to assess the temporal variability in wave power for dataset from 1 April to 30 April for the region. Using the selected sites around the island. They are: coefficient of vari- extracted and averaged dataset alongside with the ’signifi- ation (CV) and seasonal variability index (SV). cant wave height’ measurements conducted mostly for the The CV is obtained through the ratio of standard deviation month of April in the two western sites, statistical analysis ( ) of the wave power time series [P(t)] and the mean power is performed to probe on the differences among the wave ( ). A lower CV value is preferable to ease the sizing of a wave regimes of the three sites investigated. We then proceed conversion system and to reduce the possible dumping of the to investigate the dependence of western wave climate on converted wave energy system. The CV is given by: Indian Ocean swells and the eastern wave climate on the trade winds in an attempt to characterize eastern and western [f (t)] CV(P)= . (7) wave regimes. [f (t)] Next, we proceed in a case study to investigate the The SV gives an indication of the seasonal variation that wave energy potential of the western sites of interest. exists within the analysed time frame. It is obtained by sub- Spectral wave measurements taken by the Wave and Tide tracting the mean wave power for highest ( P ) and lowest s(max) recorder on the two western sites are used to produce ( P ) energy seasons and diving it by the annual mean s(min) a time series of the wave power flux prevailing for the wave power ( P ). The equation describing SV is given by: year month of April. Bivariate histograms showing the occur- rence and energy distribution of the two sites are derived. P − P s(max) s(min) Owing to the promising wave climate of the eastern site SV = . (8) year of Roches Noires, a long-term wave analysis (3 years) is performed which includes monthly, seasonal as well as 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 141 Fig. 3 Flowchart illustrating the main procedures adopted in the paper wave energy potential computations. Wave power flux have Results and discussions been calculated using significant wave height measure- ments alongside with energy period computations, derived Statistical differences among wave datasets from peak period dataset and assuming Pierson–Moskow- itz spectrum. The occurrence and wave energy power flux In view of determining whether significant differences matrices are produced to portray the wave energy regimes exist among the significant wave height measurements of the region of Roches Noires. As a final step, we con- recorded at the three regions, an ANalysis Of VAriance duct an economical evaluation, technological assessment (ANOVA) test is performed for the selected month of as well as devise maintenance operations for a proposed April. Table 2 summarizes the main results of the analysis. wave energy farm to be located at Roches Noires. Initially there are two hypotheses: 1 3 142 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Table 2 Analysis of variance Sum of squares df Mean square F P value (ANOVA) of significant wave height measurements for Roches Between groups 4228.17 3 1409.39 19310.26 0 Noires, Tamarin and Flic-en- Within groups 772.78 10588 0.07 flac for the month of April Total 5000.95 10591 Table 3 Post-hoc analysis through the Tukey–Kramer multiple comparisons test performed to reveal significant differences among the pair-wise combinations of regions investigated Site (U) Site (T) Mean difference (U–T) P value Lower limit Upper limit Flic-en-flac Roches Noires – 1.7878 0.0000 – 1.8143 – 1.7613 Tamarin 0.1478 0.0000 0.1254 0.1702 Roches Noires Tamarin 1.9356 0.0000 1.9139 1.9572 Also included is the mean difference, P value and 95% confidence intervals (Lower and upper limits) • H : No significant difference exists among the significant wave height datasets of spatially distant locations around the island. • H : Significant differences exist among the significant wave height datasets of spatially distant locations around the island. Results from Table 2 indicates the fact that the differences in significant wave height measurements taken at these three sites are highly significant (F = 19310.26, P value =0). This, therefore, leads to the rejection of the null hypothesis in favor of the alternative hypothesis. Having detected significant differences among the data - sets, a post-hoc analysis is performed in view of revealing Fig. 4 Multiple comparisons of group means test to check for differ - ences among wave regimes of the three sites. Also included are the those regions which differed significantly. Comparison tests 95% confidence intervals are employed through the Tukey–Kramer method and the main results are highlighted in Table 3. Through this analy- sis it is revealed that all sites studied differ statistically in of Flic-en-Flac and Roches Noires or Tamarin and Roches pair-wise comparison from one another. This demonstrates Noires than for the regions of Flic-en-Flac and Tamarin. the fairly different wave regimes, attributed to the different This reflects the difference in wave regimes that exists hydrologic and coastal environments which have been cho- between eastern and western waters of Mauritius. sen for investigation. Since the ANOVA test cannot determine which particular pairs are ’significantly different’, a multiple comparisons of Distributions of wave measurements group means test is carried out to support the Tukey–Kramer test previously performed. The multiple comparisons of The main distributions of the significant wave height data- group means test is used to check for differences among the sets are shown in the boxplots diagram of Fig. 5. It is evident wave regimes prevailing at the three sites investigated as from the figure that the region of Roches Noires witnesses illustrated in Fig. 4. Also included are the lower and upper a wave climate characterized by higher wave regimes with limits for the 95% confidence intervals for the true mean a mean value of 2.28 m and larger spread in the distribu- difference. The comparison intervals for the wave datasets tion than the other regions identified. The estuarine region corresponding to the three different sites do not intersect. of Tamarin records the lowest spread with a mean value of This lack of intersection indicates that the means in signifi- 0.34 m. This site detects a significant number of particularly cant wave height of the three regions investigated are signifi- high extreme values (outliers), reflecting on the unusually cantly different from each other. However, as observed from high occasional wave regimes of the estuarine structure. Fig. 4, the significant difference is greater for the regions The region of Flic-en-flac observes a mean values of 0.49 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 143 regions also play a major role in the hydrodynamic processes taking place by surface waves. The region of Tamarin offers the least likely environ- ment to place a wave energy extraction due to its low wave regimes depicted by the low mean (0.34 m) and high coef- ficient of variation (79%), delineating the high variations in significant wave height throughout the temporal scale investigated. It can be observed from spectral time series of Fig. 6Ab that the occurrence of swells in the estuary of Tamarin is spasmodic in nature, characterized by sudden outbursts of energy which occurs infrequently. This char- acteristic is due to the relatively less exposed geographi- cal location of the region pertaining to the incoming swells coming from south, southwest of the island. The second western site of Flic-en-Flac (Fig. 6A) investi- Fig. 5 Boxplot distributions of significant wave height records for gated demonstrated a higher mean in significant wave height three spatially distant sites around the coastal structure of Mauritius (0.49 m) with higher quartiles than the region of Tamarin. for the month of April. Also included is the presence of outliers, indi- cated by red crosses Moreover, a lower coefficient of variation (53.1%) was reg- istered at this site, indicating a lower variability and higher stability of the wave regimes for energy extraction as com- m with occasionally high waves, inherent of its hydrologic pared to Tamarin. On a comparative basis, the sandy shore and coastal structure. with reef site of Flic-en-Flac has higher wave energy poten- tial than the estuarine region of Tamarin despite the influ - ence of both hydrological systems from the incoming swells Variability and statistical interpretations of wave on the western regions of the island. data Interpretation of the distributions of the significant wave height datasets of the three sites reveal that the western sites Table 4 gives a descriptive statistics of the main distribu- of Flic-en-Flac and Tamarin with kurtosis of 6.41 and 7.14, tions for significant wave height (m) at the three sites of respectively, describe a sharper than normal distribution, investigation. The region of Roches Noires portrays the best characterizing them as leptokurtic. On the other hand, the wave climate for surface wave energy extraction with high eastern site of Roches Noires (kurtosis = 2.92) follow a plat- waves having a mean of 2.28 m that can reach a peak value ykurtic distribution with a flatter than normal distribution. of 3.42 m and the least variation as characterized by the low- All three sites demonstrate positively skewed distributions. est coefficient of variation (14.8%), indicative of the stability of the wave regime in the time frame considered. A low CV Dependence of eastern wave climate on trade winds is preferable to facilitate the sizing of a wave conversion system and decrease the likelihood of possible dumping of The wave regime of Roches Noires (Fig. 7a) is dependent converted wave energy system. on the trade winds originating from the subtropical high- It can be noted from the significant wave height measure- pressure zone and blowing in a southeasterly direction ments of Fig. 6 that the swell signatures of the regions of throughout the year. The predominance of the southeast Tamarin and Flic-en-Flac differ to some extent. This can trade winds influences mostly the eastern and southern be explained through the geographical location of the sites coastal microclimate wave systems of the island. Fig. 7b with respect to the incoming swells. The bathymetry of these shows the incoming wave direction at Roches Noires Table 4 Descriptive statistics which includes the mean, standard (Q1), median (Med.), upper quartile (Q3), maximum (Max.), skew- error mean (SE Mean), standard deviation (SD), coefficient of vari- ness (Skew.) and kurtosis (Kurt.) of significant wave height measured ation (CV) measured in percentage, minimum (Min.), lower quartile in meters at the three sites of investigation for the month of April Site Mean SE Mean SD CV Min. Q1 Med. Q3 Max. Skew. Kurt. Flic-en-Flac 0.493 0.0073 0.262 53.1 0.150 0.306 0.434 0.614 2.02 1.55 6.41 R. Noires 2.276 0.0089 0.336 14.8 1.575 2.031 2.214 2.513 3.42 0.65 2.92 Tamarin 0.341 0.0045 0.270 79.0 0.069 0.170 0.230 0.409 1.92 1.98 7.14 1 3 144 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Fig. 6 A Analysis of irregular offshore wave data at Flic-en-Flac for B Analysis of irregular offshore wave data at Tamarin for the month the month of April: a time series of significant wave height; b spec- of April: a time series of significant wave height; b spectral time tral time series of surface waves recorded; c zoomed images of ener- series of surface waves recorded; c zoomed images of energetic wave getic wave events occurring during the 18 days period at Flic-en-Flac. events occurring during the 26 days period at Tamarin Fig. 7 Irregular offshore wave records at Roches Noires for the month cant wave height approaching in an ESE direction under the influence of April which has been averaged over 3 years (2012–2014): a time of trade winds series of significant wave height; b wave rose representing the signifi- for the month of April, averaged over 3-years period is mostly due to the prevailing trade winds. It has been (2012–2014). The directional rose shows that the incom- reported that these waves typically possess peak wave ing waves generally approach in an east, southeast direc- periods of less than 12 s [24]. tion with the higher occurrence in the southeast direction with significant wave height peaking at 4 m, supporting the idea that the wave conditions prevailing at this site 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 145 among southern ocean basins permits extratropical south Dependence of western wave climate on Indian Ocean swells Indian Ocean swells originating from extratropical systems from adjacent extratropical areas to freely penetrate adjacent Figure 8 depicts the evolution of the eastward propagating ocean basins. Storms within the extratropical south Atlantic, Indian and Pacific oceans generally propagate towards the swells from the Atlantic Ocean, after travelling beyond the coasts of South Africa towards the Indian Ocean. Since east, below 40 S, with maximum storm densities occurring on the western flanks of these ocean basins near 50 S [25]. waves are very directional and tend to get shadowed by land masses, Mauritius situated in the line of sight from those The geographic location of Reunion Island reduces the wave energy potential of the coasts of Mauritius. Swell sources end up in the shadow region due to the geographi- cal location of Reunion island. This results in swell height heights of the order of 15.0 m, at the generation site is reduced to 6.0 m near Reunion Island [26]. It is revealed that peaking on average for the 3 days near the southern border of Mauritius at about 4.5 m and wave period of 18 s. Further the eastward propagating extratropical South Atlantic swell energy spreads through the entire Indian Ocean, attaining movement of the swell eastwards caused the shadow zone area to reduce, thereby resulting in swells to be observed the coasts of Thailand, Indonesia and southwestern Australia and penetrating as far as the Tasman Sea [25]. Owing to the at higher latitudes on the western flank of the island. Also, upon hitting Reunion Island, much of the wave energy gets directional nature of waves, the wave climate of the south and southwestern region of Mauritius changes according to dissipated. It is of interest to note that the southern region will witness the first swells followed progressively by the variations in swell events due to being exposed to the incoming swells. Consequently, as revealed in this study, regions of higher latitudes. Ocean swells are wind-waves generated by intense storms both western regions of Flic-en-Flac and Tamarin are influ- enced by swell events. that journeys long distances as they propagate away from their generation zone. Unlike the Northern Hemisphere Wave energy case study (Northern Atlantic and Pacific Ocean) where land masses provide a barrier which hinders the connection to other From the previous subsection, analysis led us to consider ocean basins, the Southern Hemisphere is open to the propa- gation of southern swell towards all ocean basins due to the the regions of Roches Noires as potential site for harness- ing wave energy. The higher waves and lower coefficient of fact that a circumpolar oceanic zone free of land barriers connects them together [25]. The wide connection existing variation of this site gives it an advantage on the other sites Fig. 8 Swell charts delineating the evolution of swell height across the Indian Ocean in temporal scale with zoomed imagery for the scenario close to Mauritius occurring for three selected dates 1 3 146 International Journal of Energy and Environmental Engineering (2018) 9:135–153 considered. It is of interest, however, to also probe into the variation (CV = 1.68) and standard deviation (SD = 2.30) wave energy potential of selected sites on the western part of indicates the highly variable wave energy regimes of this the island, whose wave regimes are dictated by the incoming locality. The quartiles of distribution were calculated and swells. Having compared the wave regimes of the two sites results show that the lower quartile, median and upper on the western flanks of the island, we provide a case study quartiles are 0.206, 0.644 and 1.52 kW/m, respectively. for the wave energy potential at the western sites of Flic-en- The distribution itself is positively skewed (4.67) and lep- Flac and Tamarin during the same above-mentioned periods tokurtic (38.1). as recorded from spectral measurements of the WTR. These The wave variability regimes of a particular sea state at a two sites provide contrasting wave regimes due to different specific location is often conveyed on a histogram of wave coastal environments, bathymetry and geographical location height and energy period. Figure 9b gives the occurrence to incoming swells. and wave energy scatters of significant wave height and The power flux for the region of Flic-en-Flac was com- energy period combinations for the 18 days records taken puted and the time series presented in Fig. 9a delineates at Flic-en-Flac. The significant wave height is arranged in the variations in temporal scales. Statistical interpreta- 0.05 m bins while energy period is organized in 0.5 s inter- tion of the time series shows that the mean power flux at vals. The color ramp represents the number of wave records this site is 1.369 kW/m with a peak value of 33.7 kW/m falling within the bins, giving an indication on the occur- recorded for the month of April. A high coefficient of rence of a particular wave state. That of Fig. 9c gives the Fig. 9 a Time series of wave power flux at Flic-en-Flac. Bivariate quency of wave records falling within a particular wave state while histograms showing the occurrence and energy distribution during that of figure (c) represent the total wave energy flux of a particular the 18 days period. The contour colors of figure (b) indicate the fre- wave state 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 147 total wave energy flux of the total number of points falling the relatively stable wave energy conditions prevailing at within the bins. Tamarin as compared to that of Flic-en-Flac. The lower, From Fig. 9b, c, it can be observed that the wave state of median and upper quartiles of distribution was found to be maximum occurrence does not coincide with the wave state 0.07, 0.133 and 0.426 kW/m, respectively, showing that the of maximum total energy flux. The most commonly occur - major portion of the distribution for wave energy computa- ring wave state has significant wave height in the interval tions at Tamarin lies significantly lower than that at Flic-en- 0.25–0.45 m and energy period in the interval 11.2–12.2s Flac. Similar to the distribution observed at Flic-en-Flac, while the most energetic wave state occurs for significant the region of Tamarin shows positively skewed distribution wave height in the interval of 0.58 and 0.75m and energy (3.86) and is characterized as leptokurtic (23.3). period in the interval 11.8 and 12.8s. For the case of Tamarin, the significant wave height is Variations in power flux for the region of Tamarin is arranged in 0.05 m bins while energy period is organized in presented in Fig. 10a and statistical analysis shows a mean 0.1 s intervals. It can be observed again that the wave state of power flux of around 0.484 kW/m with a peak value of 9.54 maximum occurrence does not coincide with the wave state kW/m, indicative of the lower potential of this site as com- of maximum total energy flux. The most commonly occur - pared to that of Flic-en-Flac. The coefficient of variation ring wave state has significant wave height in the interval (CV = 1.85) and standard deviation (SD = 0.894) reflects 0.15–0.25m and energy period in the interval 5.05–5.15s Fig. 10 a Time series of wave power flux at Tamarin. Bivariate histo- wave records falling within a particular wave state while that of figure grams showing the occurrence and energy distribution during the 26 (c) represent the total wave energy flux of a particular wave state days period. The contour colors of figure b indicate the frequency of 1 3 148 International Journal of Energy and Environmental Engineering (2018) 9:135–153 while the most energetic wave state occurs for significant of significant wave height were recorded from March to July wave height in the interval of 0.75–0.85m and energy period with a bimodal peak in April (3.42 m) and in July (3.38 m). in the interval 5.25 and 5.30s. It is evident from the plot that the peak summer months of An understanding of the performance of wave energy November and December record the lowest average signifi- converters in different wave states is essential for the con- cant wave height. The drop in significant wave height dur - structional details, cost and efficiency of the installed device. ing the summer period is due to lower southeast trade wind It would be preferable to install a wave energy converters in speeds observed during that same time scale [13]. a wave state having a maximum total energy flux rather than The wave roses showing the summer, winter and annual in one having maximum occurrence. directional significant wave height measurements averaged over the years 2012–2014 is presented in Fig. 12. It can be Long‑term analysis observed that the southeast trade winds influences the sig- nificant wave height to a greater extent in the winter period Figure 11 shows the annual distributions of significant wave with greater occurrence of southeasterly significant wave height on a monthly basis with standard deviation. It can be height measurements. However, on a comparative note, it observed that the winter half extending from May through can be observed that during summer, the magnitude of sig- October experiences the higher average significant wave nificant wave height measurements is greater with values height, reflecting on the greater wave power flux as com - peaking at 5–5.5 m as compared to the winter half where the pared to the relatively calm summer half. The highest values magnitude attains a lower value of 3.5–4.0 m. This reflects on the higher intensity of trade winds in the summer period. The boxplots representing the yearly mean evolution of significant wave height measurements are presented in Fig. 13. Data filtering processes have been performed, with the objective to remove missing data due to faulty instrument prior to boxplot representation. Of the 3 years investigated, the year 2013 records the larger spread in significant wave height measurements. This can be explained on the basis of the passage of cyclone ’Imelda’ within the vicinity of the island. Owing to relatively stable atmospheric conditions, the year 2014 recorded the lower spread of the distributions over the 3-year period. The greater number of high outliers is observed for the year 2012 while the median value over the 3-year period at Roches Noires seem to deviate not by much from the 2 m significant wave height mark. Taking into consideration the temporal variability of sig- nificant wave height measurements, an average has been per - Fig. 11 Average monthly significant wave height distributions for formed over the 3 years on the temporal basis (corresponding Roches Noires Fig. 12 Wave rose showing the magnitude and direction in which the incoming significant wave height is approaching the region of Roches Noires during (a) summer (b) winter (c) annual, averaged over the years 2012–2014 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 149 Figure  14a represents a scatter table which displays a better characterization of the composition of the wave energy resource throughout the 3 years. Each power flux was calculated with intervals of T = 2 s and H = p m0 0.5 m. From Fig. 14a, wave energy is mostly distributed between peak period T = 6 s and T = 26 s and signifi- p p cant wave height of H = 1.5 m and H = 5.5 m. From m0 m0 this distribution, the optimal energy is obtained at T = 16 s and H = 5.5 m. When further evaluating the total m0 power flux distribution by including the occurrence for each wave, its optimal distribution shifts to the right as shown in Fig. 14b. This implies that the highest amount of waves which contribute to a maximum energy output lies at T = 26 s and between H = 1.5 m and H = 4.0 m. p m0 m0 Fig. 13 Boxplots representing yearly distributions in significant wave Seasonal variation analysis at Roches Noires shows height at Roches Noires that the summer season with a SV index value of 34.3, is highly variable as compared to the winter season whose to timestamp records). Each of the yearly significant wave SV value approximates to 5.97. Moreover, the mean wave height dataset contained 15,948 records and when averag- power observed in the summer season is 28.8 kW/m and ing the 3 years corresponding to the day and time of the is found to be lower as compared to the winter season year, results in the average 2012–2014 dataset also contain- (31.7 kW/m) averaged over 3 years interval. The winter ing 15,948 records. From the ’average 2012–2014’ box plot, season with higher observed wave power flux and lower the median lies near the 2 m significant wave height mark, variability demonstrates a higher potential for extracting with upper and lower quartiles corresponding to 2.3 and 1.7 wave energy for this site. m, respectively. Fig. 14 Scatter table represent- ing 3 years (a) occurrence wave power matrix (kW/m) (b) total wave energy flux power matrix at Roches Noires from peak period (T ) and significant wave height (H ) m0 1 3 150 International Journal of Energy and Environmental Engineering (2018) 9:135–153 hand, in the post-construction stage, lower than expected Feasibility of setting up a wave energy farm energy output and high operations and maintenance costs may pose an economic risk for implementing the project in Any area with yearly averages of 15 kW/m has the potential to generate wave energy [27]. As recognized, the yearly aver- the waters of Mauritius. Besides, proper infrastructure costs need to be taken into consideration to build a rigid system age wave resource potential of the region of Roches Noires found to be 29.7 kW/m is much higher than this threshold capable of withstanding cyclonic weather conditions and storm surges to avoid excess expenditures for the repair of value and is characterized by low variations throughout the year due to persistent trade winds. A global technical damaged systems. For economic feasibility, wave power plants must be potential of 500 GW is expected for offshore wave energy devices having an efficiency of 40% and installed near coast- manufactured for an operating lifetime of at least 20 years [33]. The budget allocated to wave energy plant is spent as: lines with wave climates of the order of 30 kW/m [28]. This highly encourages the placement of a wave energy converter pre-operating cost, capital expenditure, operational expendi- ture and decommissioning costs [34]. Pre-operating costs at that location. includes expenditures encountered in pre-construction stage, environmental impact assessment among others. Capital Maintenance operations on wave energy farm expenditure costs includes the main structure as well as installation expenses. The operational and maintenance costs Operation and maintenance are needed to maintain the opti- mum mechanical performance of the wave farm through an are crucial for plant lifetime. The whole plant is supposed to be dismantled after 20 years and the decommissioning increase in productivity and reduction in running costs of wave energy converters [29]. The design for reliability and cost is estimated to be 0.5–1% of the initial investment [35]. According to a study, a 500 MW wave power plant set up in maintenance schedules need to be cautious due to the fact that system failures can be aggravated and there may be Oregon would entail a combined construction cost estimate of about $750,000/MW [36]. According to the same study, no appropriate weather window immediately available to retrieve the devices in case of failure [30]. Consequently, the implementation of such a project could produce over $90 million annually in new tax revenues on a statewide the plan to install a wave energy device in the offshore areas of Mauritius should come with an appropriate maintenance basis [36]. The introduction of a wave energy farm in the vicinity of strategy. There exists several maintenance strategies for wave energy converters which includes corrective strategy the island would entail numerous economic benefits in the long run. Integrating wave energy resources in the energy performed after failure or preventive strategy done before a breakdown. mix of Mauritius would ensure security of energy supply and the transition to greener alternatives. Consequently, in A systematic maintenance strategy exists which is based on systematic maintenance actions performed every 6 the long run, the island may become less dependent on high costs fossil fuel imports. The higher wave energy potential months combined with corrective maintenance action per- formed in case of breakdown. Despite the fact that system- site of Roches Noires identified, coupled to appropriate maintenance strategies planned are important aspects that atic maintenance strategy presents a wide variety of main- tenance actions, it is identified as the most costly strategy contribute to the longevity and economic yield of a typical wave farm to be implemented in that region. [31]. Nonetheless, it would suit the climate of Mauritius since as observed in the results section, the summer season Technological assessment experiences lower average significant wave height, reflecting a lower wave energy flux. Consequently, it would be appro- Several technologies that are capable of absorbing energy priate to plan proper maintenance strategies in the summer half to be undertaken at intervals of 6 months. from waves and converting it into electrical energy have been developed. However, the selection of appropriate wave Economic evaluation device depends on the water depth and the location (shore- line, near-shore, offshore) considered for construction [37]. Wave energy developers target high economic return with Classification of wave energy technologies is based on work - ing principles and are mainly grouped into oscillating water the intention to satisfy investor’s vested interests and gener- ate profits. In addition to expected rate-of-return, investment column (with air turbine), oscillating bodies (with hydraulic motor, hydraulic turbine and linear electrical generator) and risks that exist before and after project implementation is an important factor that cannot be neglected [32]. In the overtopping systems (with low-head hydraulic turbine) [37]. Each of them can be mounted on a fixed structure or placed pre-construction stage, developers may experience friction with local inhabitants and marine conservation groups per- on a floating platform. The Mutriku Breakwater Wave Plant located in Spain, having an installed capacity of 296 kW taining to the implementation of the project. On the other 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 151 from 16 turbines and 16 oscillating water columns, produced be influenced by episodic swell events originating from and exported over 1.3 GWh of power to the Basque grid the Atlantic Ocean with the region of Flic-en-Flac more during its first 5 years [38]. Nevertheless, the efficiency of exposed and the estuarine environment of Tamarin less energy capture is a major limitation of wave energy convert- exposed to the SSW incoming swells. On the eastern side ers. According to one study, out of 2700 GW power gener- of the island, the region of Roches Noires is observed to ated from waves, only about 500 GW can be captured using have higher wave resource potential with lower variabil- present technologies [39]. ity and higher wave power flux. Wave rose plots depicted the fact that the incoming waves from Roches Noires are International standardization influenced by the prevailing southeast trade winds, with the winter season displaying a greater influence owing to For implementing the wave farm in the waters of Roches the greater occurrence of incoming waves from the south Noires, the IEC standard guideline (62600-101, 2015) east during that period. The mean high wave power flux which aims at defining good practices in the field of wave registered at Roches Noires (28.8 kW/m in summer and resource assessment needs to be taken into consideration. 31.7 kW/m in winter) demonstrates the relatively higher Having identified the region of Roches Noires as potentially potential of exploiting surface wave energy from this site suitable site for installing the wave energy converters, the as compared to the western regions of Flic-en-Flac and next phase involves performing a spatial analysis of wave Tamarin. Consequently, for the tropical island of Mauri- energy resources in that region. The IEC standard guideline tius, efficient harnessing of wave energy will be performed establishes a system for estimating, analyzing and report- through the installation of wave energy converters around ing the wave energy resource at sites identified having good the eastern and southeastern coasts of the island, where wave energy resources [40]. The report provides methods the system is influenced by wind–wave generated system regarding data collection and resolution alongside with the as compared to the western coast which is under the influ- development of numerical model for wave energy resource ence of extratropical South Indian Ocean swells. Having estimation. In the data collection section, the need for prepa- identified the eastern region of the island as most prom- ration of a bathymetry contour map is highlighted and will ising for harnessing energy from waves, further orienta- be used to construct a Digital Elevation Model for the wave tions of this research involve performing a spatial analysis propagation model. Moreover, the section includes the fact of the wave resource potential of this study area through that archived data may be used as primary data source while numerical modelling techniques. Numerical simulations wave measurements can be used to validate the numerical coupled to a feasibility study on wave energy penetration model. Additional data which may be of use includes wind, in the eastern coasts of Roches Noires will be investigated tide, current and water density. in the next phase. In the numerical model section, production of at least 10 years of sea state data, with wave resource taken as station- Data availability The data that support the findings of this ary, is required. The minimum frequency for sea state data is study are available from the Mauritius Oceanography Insti- one dataset recorded at intervals of 3 h. Guidelines regarding tute but restrictions apply to the availability of these data, the configuration of boundary conditions, data processing which were used under approval for the current study, and so and validation of numerical modelling as well as dealing are not publicly available. Data are, however, available from with missing data are mentioned. The practical implications the author upon reasonable request and with permission of of the report has been evaluated through comparison with the Mauritius Oceanography Institute. the Biscay Marine Energy Platform (BiMEP) facility which is situated on the Basque coast. Consequently, the case study Acknowledgements We would like to acknowledge the African Moni- may be of relevance to the Government of Mauritius or any toring of the Environment for Sustainable Development (AMESD) pro- organizations considering the implementation of a wave ject and Mr Eric Martial for grating permission to use the wave buoy farm in the identified site of Roches Noires. data. We extend our gratitude to Magicseaweed.com who has cordially agreed to provide us with the swell charts. Special thanks to Arnaud Nicolas and Khishma Modoosoodun–Nicolas for their assistance in the study. We are grateful to the Mauritius Oceanography Institute Conclusions (MOI) for providing the adequate facilities for carrying out research presented in this paper This paper presents an assessment of the wave energy potential of three sites having distinct coastal environ- Compliance with ethical standards ments and wave regimes. Two sites located on the west- ern (leeward) side region of the island are observed to Conflict of interest On behalf of all authors, the corresponding author states that there is no conflict of interest. 1 3 152 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Open Access This article is distributed under the terms of the Creative 16. Nordstrom, K.F.: Bayside beach dynamics: implications for Commons Attribution 4.0 International License (http ://crea tive comm simulation modeling on eroding sheltered tidal beaches. Mar. ons.org/licenses /b y/4.0/), which permits unrestricted use, distribution, Geol. 25(4), 333–342 (1977) and reproduction in any medium, provided you give appropriate credit 17. Warshaw, M.: The encyclopedia of surfing. Houghton Mifflin to the original author(s) and the source, provide a link to the Creative Harcourt, Boston (2005) Commons license, and indicate if changes were made. 18. IOC: IOC-UNEP-WMO-SAREC Planning Workshop on an Integrated Approach to Coastal Erosion, Sea Level Changes and their Impacts. http://unes doc.unes co.or g/images/0010 /0010 01/1001 38eo .pdf (1994). Accessed 15 May 2017 References 19. Ashton, I.G.C., Johanning, L.: On errors in low frequency wave measurements from wave buoys. Ocean Eng. 95, 11–22 (2015) 1. SM: Population and Vital Statistics—Year 2016. http: //stats mau 20. Joosten, H.: Directional wave buoys and their elastic mooring. riti us.govm u.org/Engl ish/Publ icat ions /Page s/Pop_and_Vita l_ Int. Ocean Syst. 10(4), 18–21 (2006) Stat s_Yr16 .aspx (2017a). Accessed 7 April 2017 21. Sierra, J.P., Martn, C., Msso, C., Mestres, M., Jebbad, R.: Wave 2. SM: Energy and Water Statistics—Year 2016. http://s tatsmau r iti energy potential along the Atlantic coast of Morocco. Renew. us.govm u.org/Engl ish/Stat sbyS ubj/Page s/Ener gy_Wate r_Stat Energy 96, 20–32 (2016) s_Year _2016 .aspx (2017b). Accessed 9 April 2017 22. Rusu, E., Soares, C.G.: Wave energy pattern around the Madeira 3. MPU: Long-Term Energy Strategy 2009–2025. https://sus taina Islands. Energy 45(1), 771–785 (2012) bled evel opme nt.un.org/cont ent/docu ment s/1245 maur itiu sEne 23. Cornett, A. M.: A global wave energy resource assessment. In rgy%20St rate gy.pdf (2009). Accessed 10 April 2017 The Eighteenth International Offshore and Polar Engineering 4. Lannoye, E.: Renewable energy integration: practical manage- Conference. In: International Society of Offshore and Polar ment of variability, uncertainty, and flexibility in power grids Engineers (2008) [book reviews]. IEEE Power Energy Mag. 13(6), 106–107 24. Baird Report: Proposal for Controlling Severe Erosion and (2015) Beach Improvement works in front of The Grand Mauritian 5. Neill, S.P., Lewis, M.J., Hashemi, M.R., Slater, E., Lawrence, Hotel, Balaclava. http://environment.govmu.org/English/ J., Spall, S.A.: Inter-annual and inter-seasonal variability of eia/Documents/Repor ts/blueocean/grand/chapt2.pdf(2003). the Orkney wave power resource. Appl. Energy 132, 339–348 Accessed 20 June 2017 (2014) 25. Alves, J.H.G.: Numerical modeling of ocean swell contributions 6. Ram, K., Narayan, S., Ahmed, M.R., Nakavulevu, P., Lee, Y.H.: to the global wind-wave climate. Ocean Model. 11(1), 98–122 In situ near-shore wave resource assessment in the Fiji Islands. (2006) Energy Sustain. Dev. 23, 170–178 (2014) 26. Samiksha, S. V., Vethamony, P., Aboobacker, V. M., Rashmi, R.: 7. Bernardino, M., Rusu, L., Soares, C.G.: Evaluation of the wave Propagation of Atlantic Ocean swells in the north Indian Ocean: energy resources in the Cape Verde Islands. Renew. Energy 101, a case study. Nat. Haz. Ear. Sys. Sci. 12, 3605–3615 (2012) 316–326 (2017) 27. Nelson, V.C., Starcher, K.L.: Introduction to renewable energy. 8. Iglesias, G., Carballo, R.: Wave resource in El Hierroan island CRC press, Boca Roton (2015) towards energy self-sufficiency. Renew. Energy 36(2), 689–698 28. Sims, R. E., Schock, R. N., Adegbululgbe, A., Fenhann, J. V., (2011) Konstantinaviciute, I., Moomaw, W., Uchiyama, Y.: Energy sup- 9. Nordstrom, K.F.: Estuarine Beaches: an introduction to the ply. In Climate change 2007: Mitigation. In: Contribution of physical and human factors affecting use and management of Working Group III to the fourth assessment report of the Inter- beaches in estuaries, lagoons, bays and fjords. Springer Science governmental Panel on Climate Change. Cambridge University & Business Media, Netherlands (1992) Press (2007) 10. Ludwick, J. C.: Mechanisms of sand loss from an estuarine 29. Rinaldi, G., Thies, P.R., Walker, R., Johanning, L.: On the anal- groin system following an artificial sand fill. In: Technical ysis of a wave energy farm with focus on maintenance opera- Report 87-2 Department of Oceanography, Old Dominion Uni- tions. J. Mar. Sci. Eng. 4(3), 51 (2016) versity, Norfolk, VA (1987) 30. Cruz, J.: Ocean wave energy: current status and future prespec- 11. Lutz, W., Wils, A. B.: People on Mauritius: 1638–1991. In Pop- tives. Springer Science & Business Media, Berlin (2007) ulationDevelopmentEnvironment (pp. 75-97). Springer, Berlin, 31. Sahnoun, M. H., Baudry, D., Mustafee, N., Louis, A., Smart, P. Heidelberg(1994) A., Godsiff, P., Mazari, B.: Modelling and simulation of opera- 12. ASCLME: National Marine Ecosystem Diagnostic Analysis, tion and maintenance strategy for offshore wind farms based on Mauritus. http://www .asclme.or g/repor ts2013/N ational %20ME multi-agent system. J. Intell. Manuf. 1, 1–17 (2015) DAs/MEDA %20Ma urit ius/_Maur itiu sMED APro of.pdf(2012). 32. Van Haaren, R., Fthenakis, V.: GIS-based wind farm site selec- Accessed 19 April 2017 tion using spatial multi-criteria analysis (SMCA): evaluating the 13. JICA: The Project for Capacity Development on Coastal Protec- case for New York State. Renew. Sustain. Energy Rev. 15(7), tion NIPPON KOEI CO., LTD. and Rehabilitation in the Repub- 3332–3340 (2011) lic of Mauritius. http://open_jicareport.jica.go.jp/pdf/12237327. 33. Wengenmayr, R., Buhrke, T.: Renewable energy: sustainable pdf (2015). Accessed 2 May 2017 energy concepts for the future. Wiley, Hoboken (2011) 14. Anthony, A., Atwood, J., August, P., Byron, C., Cobb, S., Foster, 34. Astariz, S., Iglesias, G.: Enhancing wave energy competitive- C., Fry, C., Gold, A., Hagos, K., Heffner, L., Kellog, Q., Lel- ness through co-located wind and wave energy farms. A review lis, K.,Opaluch, J., Oviatt, C., Pfeiffer-Herbert, A., Rohr, N., on the shadow effect. Energies 8(7), 7344–7366 (2015) Smith, L.,Smyth, T., Swift, J., Vinhateiro, N.: Coastal lagoons 35. Castano, M.: Sistema de Monitorización y Supervisión de Una and climate change: ecological and social ramifications in US Boya Para Generación de Energía Undimotriz. Universidad Atlantic and Gulf Coast ecosystems. Ecol. Soc. 14(1), 8 (2009) Politécnica de Cataluña, Barcelona, Spain (2011) 15. Nordstrom, K.F., Jackson, N.L.: Migration of swash zone, 36. OWET: Economic Impact Analysis of Wave Energy: Phase One. step and microtopographic features during tidal cycles on an https://oregonwave.org/oceanic/wp-content/uploads/2013/09/ estuarine beach, Delaware Bay, New Jersey, USA. Mar. Geol. Economic-Im pact- Anal y sis-of- Wa v e-Ener gy -Phase- 92(1–2), 147–154 (1990) 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 153 One%E2%80%94September-2009.pdf(2009). Accessed 15 Dec 40. OPERA: Wave Energy Measurement Methodologies for IEC 2017 Technical Specifications.http://oper a-h202 0.eu/wp-cont ent/uplo 37. Antonio, F.D.O.: Wave energy utilization: a review of the tech-ads/2016 /03/OPER A_D5.1_IEC_TS_Appl icat ion_UCC_2016 nologies. Renew. Sustain. Energy Rev. 14(3), 899–918 (2010)1129 _v2.1.pdf (2016). Accessed 18 Dec 2017 38. TidalEnergyToday: Mutriku wave plant generates over 1 GWh of clean power. http s://tida lene rgyt oday .com/2016 /07/19/mutr Publisher’s Note Springer Nature remains neutral with regard to iku-wave -plan t-gene rate s-over -1gwh -of-clea n-powe r/ (2016). jurisdictional claims in published maps and institutional affiliations. Accessed 16 Dec 2017 39. McCormick, M.E., Ertekin, R.C.: Renewable sea power. Mech. Eng. 131(5), 36 (2009) 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Energy and Environmental Engineering Springer Journals

Assessment of the wave potential at selected hydrology and coastal environments around a tropical island, case study: Mauritius

Free
19 pages

Loading next page...
 
/lp/springer_journal/assessment-of-the-wave-potential-at-selected-hydrology-and-coastal-MOSuGlY6KD
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Engineering; Renewable and Green Energy
ISSN
2008-9163
eISSN
2251-6832
D.O.I.
10.1007/s40095-018-0259-7
Publisher site
See Article on Publisher Site

Abstract

Waves are the dominant influence on coastal morphology and ecosystem structure of tropical islands. The geographical positioning of Mauritius near to the Tropic of Capricorn ensures that the eastern regions benefit from the persistent southeast trade winds which influence the incoming surface waves. In this study, we present the high dependence of the wave regimes of windward offshore site on the trade winds. The higher occurrence of incoming waves in the winter season directed in the southeast direction indicates that the trade winds are more prevalent in the winter season. Storms within the extratropical South Atlantic, Indian and Pacific oceans generally propagate towards the east such that extratropical South Atlantic swell energy spreads through the entire Indian Ocean. Since waves are very directional and tend to get shadowed by land masses, Mauritius situated in the line of sight from those sources end up in the shadow region due to the geographical location of Reunion island. In this study, we support the explanation on how the western region of the island gets influenced by episodic swell events. A detailed wave energy resource assessment is provided for different targeted coastal environments around the island. It is revealed that the mean wave power observed in the summer season at one of the sites can attain 28.8 kW/m and is found to be lower as compared to the winter season (31.7 kW/m). Keywords Wave climate · Wave energy · Tropical Island · Statistical analysis · Resource assessment · Indian Ocean Introduction representing a growth of 0.1% as compared to the preced- ing year [1]. The increasing population dynamics fuels the Mauritius, a small island located in the tropical belt of the need for increasing energy requirement. Consequently, this southwestern Indian Ocean basin, enjoys a tropical maritime has led to an increase of 1% in total primary energy require- climate that spans over two seasons: summer and winter. ment during this 1 year interval to reach a value of 1550 ktoe The population of the island in 2016 stood at 1,263,820, [2]. Typical of small island developing states (SIDS), Mau- ritius relies heavily on imported fossil fuels (with a share of 85%), reflecting on the vulnerable and volatile economy Electronic supplementary material The online version of this of the country. The electricity generation for the year 2016 article (http s://doi.org/10.1007 /s400 95-018-0259 -7) contains amounted to 3042 GWh, with a share of 78.2% attributed supplementary material, which is available to authorized users. to imported fuel oil and coal and the remaining percentage * Jay Rovisham Singh Doorga (21.8%) coming from renewable energy sources compris- jay.doorga927@gmail.com ing bagasse (16.3%), hydro (3.3%), photovoltaic (1%), wind (0.6%) and Landfill gas (0.6%) [ 2]. The huge renewable Physical Oceanography Unit, Mauritius Oceanography energy potential of the island for sustainable energy produc- Institute, Avenue des Anchois, Morcellement de Chazal, Albion, Mauritius tion has been recognized and bold measures are being taken by policy makers in the form of international partnerships Physical Oceanography/Marine Geoscience Unit, Department of Continental Shelf and Maritime Zones Administration and supportive regulatory frameworks to encourage inde- and Exploration, Prime Minister’s Office, Port -Louis, pendent power producers (IPPs) invest in the green sector. Mauritius The long-term vision of the Government of Mauritius is to Technical Unit, Mauritius Oceanography Institute, Avenue des Anchois, Morcellement de Chazal, Albion, Mauritius Vol.:(0123456789) 1 3 136 International Journal of Energy and Environmental Engineering (2018) 9:135–153 increase the percentage of electricity generation from renew- wave climate imposed by estuarine environment on waves able energy resources to 35% by 2025 [3]. coming from deep water is worth exploring. The authors Presently, the generation of electricity in Mauritius from believe that information presented in this paper will be help- wave resources is inexistent. Wave energy is acknowledged ful to the Government or any relevant organization in mak- as being a reliable renewable energy source whose output ing an informed decision with regard to major investments can be predicted in both spatial and temporal scales [4]. In for harnessing the wave resource for domestic use at the addition to displacing fossil fuels, which inadvertently help investigated sites. reduce greenhouse gas emissions and improve energy secu- rity, wave energy holds the benefit of being relatively less intermittent in nature than solar and wind. The heterogeneity Material and methods associated with cloud systems and prevailing wind condi- tions results in fluctuations from solar and wind power gen- Geology and climatology erations. Consequently, the implementation of wave energy ◦ ◦ extraction devices in the near-shore and off-shore regions of Mauritius (20 10 S, 57 30E), located in the south western the island would help reduce instabilities occurring in the Indian Ocean near the southern end of the Mascarene Ridge, national grid by conditioning power and smoothing fluctua- has an approximated surface area of 1859 km (Fig. 1). The tions. The success of renewable energy lies in its diversi- island has an Exclusive Economic Zone (EEZ) of about 2.3 2 2 fication and the inclusion of wave energy can be a strong million km (including approx. 400,000 km jointly man- component in the renewable energy mix which would benefit aged with the Seychelles) that extends over the islands of to the local community at large. Moreover, the proximity of Rodrigues, Agalega, Cargados Carajos shoals, Chagos inland and coastal regions to the sea (typically less than 30 Archipelago and Tromelin. Of volcanic origin, starting with km radius) holds the benefit of electricity transmission and the Breccia Series 10–7.8 million years (my) B.P., the geo- access even to remote locations on the island. morphological features of the island comprise an elevated Conversion of wave resources can contribute substantially central plateau (about 500 m above sea level) surrounded by to the electricity demand of several islands. The Orkney a chain of mountains and some isolated peaks. The eleva- archipelago, situated in the north of Scotland is estimated tion of the terrain decreases gently from the central plateau to have a mean wave power of 10–25 kW/m and has been towards the coastal edge. The coastline measures about 200 earmarked to contribute 550 MW to the total 1.6 GW wave km long and is composed of different shore types (catego- and tidal energy capacity from twelve leased sites by 2020 rized as sandy shores, rocky shores, muddy shores, mixed [5]. An assessment of the near-shore wave energy potential shores, calcareous limestone shores, cliffs and coastal wet- for the southern pacific islands of Fiji showed that mean lands) whose landforms are related to the coastal geomor- wave power of 9.81 kW/m was detected at a depth of 15 m phologic features. The shoreline is ringed by fringing coral in the west of the main island while energy flux of around reefs which enclose a lagoon having an approximated area 28.78 kW/m was identified near Kadavu island at a depth of 243 km [11]. With an annual growth rate of about 2 mm, of 18 m [6]. The Cape Verde Islands, located off the coast living coral reef formations resist destruction from sea action of West Africa in the central Atlantic Ocean, witnessed a by constantly growing due to the preference for light and wave power exceeding 7 kW/m in coastal areas neighbor- sediment-free waters [11]. ing the islands, with the summer season recording around The climatological regime of the island, characterized as half the wave potential of the winter season [7]. The island mild tropical maritime, is dictated by alterations of the two of El Hierro, situated south west of the Canary Islands in seasons: Summer (November–April) and winter (May–Octo- the Atlantic Ocean estimates its wave potential to be of the ber). The summer season is distinguished as warm and order of 25 kW/m, ree fl cting on the possibility of harnessing humid while the winter season is generally cool and dry. energy from waves [8]. Microclimatic conditions exist spatially in different locali- The aim of the present study is to investigate the wave ties attributed to the varying atmospheric conditions on the resource potential of open water sites, not too distant for island. The wave patterns on the coast of Mauritius is influ- electricity transmission inland and located outside of the enced all year-round by the westerly and persistent south- coral reef platform of the island. Despite the fact that easterly trade winds which is ensured by the geographical refracted ocean wave energies may be great at estuarine location of the island near to the Inter Tropical Convergence sites near wide entrances to the sea, significant energy loss Zone (ITCZ). occurs when entering estuaries, either by refraction around split platforms or through breaking on entrance shoals [9]. Waves entering basins may dominate the spectral estimates at sites close to the ocean [10]. Consequently, the complex 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 137 Fig. 1 a Location of three earmarked sites considered in Mauritius nified view of the two stations located just outside the coral reef plat- for the assessment of wave potential. b Geographical location of the form on the western side of the island tropical island of Mauritius, southwest of Indian Ocean basin. c Mag- west of the island by the presence of the mountains located Coastal hydrodynamics in the south western region. Typically, wind waves in open sea are between 0.5 m during summer and 3 m during winter The geographical position of the island ensures that it ben- with a period of 3–11 s [12]. Significant wave heights in efits from the South Equatorial Current (SEC) flowing from south, southwest (SSW) and southeast (SE) of the island west to east throughout the duration of the year. Addition- are of the order of 1.5–2.5 m in summer and 2.5–3.5 m in ally, the trade winds from anticyclones of the south Indian winter [13]. High waves reaching 3–5 m with a wave period Ocean which moves from east to west influence the coastal of 12–20 s traveling long distances, reaching the southern hydrodynamics and offshore current systems of the island. region of the island as swells with little loss in energy are These exhibit seasonal variations with the winter season generated by the deep low pressure systems in the high lati- experiencing relatively stronger currents in magnitude due to tudes [12]. the strong prevailing South Easterly trade winds. The speeds of the current increase as it passes through the channels situ- ated within the Mascarene Plateau resulting in the forma- Site selection tion of strong gyres on the leeward side. When it reaches Madagascar, part of the current flows to the north to feed the Three study sites (Fig. 2) were selected based on diversified Agulhas current in the Mozambique Channel and the East hydrologic and coastal environments, reflecting the distinct Africa current while the other part flows southwards along wave regimes prevailing at these localities. The impact of the Madagascan coast forming an anticyclone gyre further three different types of coastal hydrological environments south [12]. During the South west monsoon, the SEC feeds (Sandy shore with reef, estuary and mixed shore with reef), into the Somali current along the East coast of Africa. geographical location in connection with the prevailing The west coast of the island is relatively more protected southeast trade winds and swells coming from the south west from wind and wave action than the east as the body of the on the wave climate around the island are explored. island creates a lee whose effect is strengthened in the south 1 3 138 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Fig. 2 Aerial view depicting the coastal environments of the three sites under investigation in this study. The exact location for the deployment of the Wave and Tide Recorder (WTR) on the western flanks of the island and the waverider buoy on the eastern side are presented in the figure Flic en Flac is a sandy shore (with reef) composed essen- good balance of sands and rocks, generally interspersed. tially of sediments. The sediments, of carbonate origin are Located in the northeastern part of the island of Mauritius, made up of coral fragments, crustaceans and molluscan Roches Noires is one region where the highest waves are shells among others. The sandy beach that it makes from the usually observed off the reefs. This eastern region of the adjacent coral reefs constitutes one of the most economically island is under the influence of the southeast trade winds important and environmentally sensitive coastal habitats of throughout the year which provide the most substantial the Republic of Mauritius. Indeed, with the effect of climate wave action that dominate the depositional dynamics [18]. change, this fragile ecosystem is at risk of damage due to The deployment of the instruments at those exact geo- the impacts of sea level rise, warming of air and sea, and graphical coordinates was based on ease of access to the increased storminess [14]. sites, human activity level near these regions and sensitiv- The wave characteristics on the estuarine environment ity of the apparatuses in the water depths placed. Other of Tamarin is also studied. The influence of ocean waves locations were also considered but dropped primarily due may extend several kilometers into an estuary [15] but its to safety issues arising from the fact that divers would not effectiveness generally decreases with distance from the be able to access the devices easily. For the regions of opening to the ocean [16]. The location of Tamarin bay on Flic-en-Flac and Tamarin which have significant human the southwest coast is angled so that the winds blow offshore activity levels, we chose to investigate sites relatively away such that incoming swells have to bend sharply around the from these influences for ease of deployment and to get coast before hitting the reef [17]. We choose to study the minimum disturbance of wave records from anthropogenic wave characteristics in the Tamarin estuary near the center activities. A description of the four wave measurement of the wide entrance to the sea due to high refracted wave sites considered for deployment is provided in Table  1. energies converging at this site. More details on the regions of interest are provided in the Roches Noires is a mixed shore (reef) composed essen- following sub-section. tially of sediments and pebbles. It constitutes a fairly 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 139 Table 1 Description of the four wave measurement sites in terms of period of deployment, radial distance from the shore (d ) , water depth shore (Dep) and human activity levels Site Period d (m) Dep (m) Human activity shore 1 Flic-en-Flac 2 Apr–20 Apr 2007 490 12 Popular tourist region 2 Tamarin 28 Mar–22 Apr 2014 935 10 Intermediate human activity 3 Roches Noires 9 Feb 2012–10 Nov 2014 1228 46.2 Some human activity apparatus were detected. Also, selection of years was based Materials and methods on a number of factors including the wave climate prevailing during the considered period. It was ensured that the selected The Seawatch mini II Waverider buoy, designed by Fugro years for which the WTR was deployed contained no data Oceanor, is intended to follow the motion of the sea surface recorded in cyclonic weather conditions, which would inevi- waves. This results in exact measurement of the water sur- tably lead to biased results when it comes to selecting the face and can be employed to derive directional measurement most energetic region for tapping in surface wave energy. of the surface wave field [19]. Solid-state accelerometers are The WTR and Waverider buoy were configured with a sam- used to represent accelerations on three orthogonal axes. pling frequency of 1 Hz to sample data for 1024 s. A 1 Hz Integration with simultaneous heading and tilt readings frequency ensures that waves of up to 0.3 Hz (3.33 s) can be performed on-board using the ’wavesense’ computer unit accurately recorded. Owing to the fact that waves resulting installed in the buoy, results in accelerations being resolved from meteorological events lie in the period range of 7–15 s, to heave (vertical), east and north axes. The design ensures the selected sampling frequency would make optimum use that it does not influence the active frequency range for sur - of memory and batteries for such a long period deployment. face waves, also minimizing any damping or phase shift on the signals. As stated by Joosten [20], the mooring design Theoretical background for floating wave buoys must make sure that they allow the latter to respond unimpeded to wave frequencies while pro- Significant wave height H and wave power P are two viding the sufficient restoring force to resist tidal flow and m0 w important parameters when it comes to assessing the wave drift forces. In this study, the buoy was moored using a 600 energy potential of a particular sea state. The significant kg bottom weight attached to a 45 m polypropylene braided wave height is defined as the average wave height of one- rope, having its end connected to a 15 m rubber cable. The third of the highest waves in a wave record. For two regions: waverider has a lead acid battery bank 12 V, 62 Ah × 4 Flic-en-Flac and Tamarin, spectral data have been collected (248 Ah) coupled to an optional lithium battery bank rated using the WTR. The spectral estimate of significant wave 1088 Ah. Also included in the design are solar panels with height can be estimated from the zeroth moment of variance a rating of 12 W × 6 (72 W). The buoy was secured in water spectrum m as follows: depth of 46.2 m at a distance of 1228 m from the shore of Roches Noires. Measurements started at this site on the 9th H = 4 m . (1) m0 0 of February 2012 and lasted for 32 months. The buoy was For the non-directional variance density spectrum, the spec- retrieved on the 10th of November 2014. Cyclonic weather tral moment of nth order, m , is calculated using the follow- conditions resulting from the passage of cyclone ‘Imelda’ n ing equation. near the vicinity of the coast of Mauritius from the 6th to 16th of April 2013 resulted in extreme wave values recorded. Technical issue was encountered by the device in the period n m = f E Δf , (2) n i i from 16th October 2013 to 7th November 2013. i=1 A Valeport Wave and Tide recorder (WTR) was deployed where f is the i th frequency corresponding to the n th order, in the month of April at two sites around the island: Flic- i Δf is the frequency increment and E is the sea surface vari- en-Flac and Tamarin. WTR sensor has a precision resonant ance density over a range of wave frequencies i. quartz crystal and optional strain gauge with an accuracy The energy period T which is the variance-weighted of ± 0.01 m for water level measurements. Archived wave mean period of the non-directional variance density spec- measurements taken for the month of April by the Mauritius trum is calculated from the spectral moments as follows: Oceanography Institute has been chosen owing to important wave characteristics taking place in that specific month of −1 T = . the year. The selection of the years for analysis was primarily (3) based on the fact that no missing wave records due to faulty 1 3 140 International Journal of Energy and Environmental Engineering (2018) 9:135–153 The wave power (kW/m) in deep water can be calculated Procedure adopted using significant wave height and energy period as follows: The flowchart portraying the adopted strategy in this paper 2 2 is presented in Fig.  3. The intention of the current study P = H T = 0.491H T . (4) w m0 e m0 e involves assessing the wave potential at selected hydrol- ogy and coastal environments around Mauritius. Conse- Since the points investigated lies outside the coral reef bar- rier in open water and are found in water depths exceeding quently, two sites having fairly distinct wave climate and located in the western part of the island have been selected. 10 and 200 m from the shore, the use of Eq. 4 is justified. For the Roches Noires dataset, information on spectral These include the sandy shore with reef region of Flic-en- Flac and the estuarine environment of Tamarin. Significant moments or spectral shape was not provided, and therefore, T must be estimated from available dataset and one approach is wave height ( H ) and energy period ( T ) measur ements m0 e have been recorded at the two sites through the deploy- to assume the following: ment of a Wave and Tide recorder for the time frames: 2 T = T , e p (5) April–20 April 2007 (Flic-en-Flac) and 28 March 22 April where  is a coefficient whose value depends on the shape 2014 (Tamarin). The selected time interval for the data as of the wave spectrum (0.86 for a Pierson–Moskowitz spec- mentioned previously is based on suitability of data available trum and increasing towards unity with decreasing spectral (no missing wave records and no data selected in cyclonic width) [21]. In the spectrum formulation, it is assumed that weather conditions), safety issues for divers to access instru- the wind is blowing steadily for a long time and over a large ment, human activity level which may influence the wave area such that the waves attain a point of equilibrium with measurements and important wave processes taking place the wind. The conversion from peak period (T ) to energy p in that specific month of the year. On the eastern region of period (T ) was enable through the approximation: the island, the mixed shore with reef site of Roches Noires has been chosen for further investigation. Significant wave T ∼ 0.86T . (6) e p height and peak period ( T ) measurements have been By combining Eqs. 4 and 6, the total wave energy resource at acquired from the wave buoy for the time frame 9 February Roches Noires can be estimated. However, besides estimat- 2012–10 November 2014. ing the wave energy potential at the sites, it is of interest to Due to the temporal variability of the phenomena under analyze the temporal variability of the distributions. Steady study, we attempt to compare the wave regimes of the three wave energy flux are preferred to those having unsteady selected sites on the same time scale. Consequently, April wave regimes due to the fact that they are more reliable and datasets for the years 2012–2014 recorded at Roches Noires display greater efficiency [22]. have been extracted and averaged (corresponding to day and Two coefficients that were proposed by Cornett [ 23] are time of record) to obtain a single ’significant wave height’ employed to assess the temporal variability in wave power for dataset from 1 April to 30 April for the region. Using the selected sites around the island. They are: coefficient of vari- extracted and averaged dataset alongside with the ’signifi- ation (CV) and seasonal variability index (SV). cant wave height’ measurements conducted mostly for the The CV is obtained through the ratio of standard deviation month of April in the two western sites, statistical analysis ( ) of the wave power time series [P(t)] and the mean power is performed to probe on the differences among the wave ( ). A lower CV value is preferable to ease the sizing of a wave regimes of the three sites investigated. We then proceed conversion system and to reduce the possible dumping of the to investigate the dependence of western wave climate on converted wave energy system. The CV is given by: Indian Ocean swells and the eastern wave climate on the trade winds in an attempt to characterize eastern and western [f (t)] CV(P)= . (7) wave regimes. [f (t)] Next, we proceed in a case study to investigate the The SV gives an indication of the seasonal variation that wave energy potential of the western sites of interest. exists within the analysed time frame. It is obtained by sub- Spectral wave measurements taken by the Wave and Tide tracting the mean wave power for highest ( P ) and lowest s(max) recorder on the two western sites are used to produce ( P ) energy seasons and diving it by the annual mean s(min) a time series of the wave power flux prevailing for the wave power ( P ). The equation describing SV is given by: year month of April. Bivariate histograms showing the occur- rence and energy distribution of the two sites are derived. P − P s(max) s(min) Owing to the promising wave climate of the eastern site SV = . (8) year of Roches Noires, a long-term wave analysis (3 years) is performed which includes monthly, seasonal as well as 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 141 Fig. 3 Flowchart illustrating the main procedures adopted in the paper wave energy potential computations. Wave power flux have Results and discussions been calculated using significant wave height measure- ments alongside with energy period computations, derived Statistical differences among wave datasets from peak period dataset and assuming Pierson–Moskow- itz spectrum. The occurrence and wave energy power flux In view of determining whether significant differences matrices are produced to portray the wave energy regimes exist among the significant wave height measurements of the region of Roches Noires. As a final step, we con- recorded at the three regions, an ANalysis Of VAriance duct an economical evaluation, technological assessment (ANOVA) test is performed for the selected month of as well as devise maintenance operations for a proposed April. Table 2 summarizes the main results of the analysis. wave energy farm to be located at Roches Noires. Initially there are two hypotheses: 1 3 142 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Table 2 Analysis of variance Sum of squares df Mean square F P value (ANOVA) of significant wave height measurements for Roches Between groups 4228.17 3 1409.39 19310.26 0 Noires, Tamarin and Flic-en- Within groups 772.78 10588 0.07 flac for the month of April Total 5000.95 10591 Table 3 Post-hoc analysis through the Tukey–Kramer multiple comparisons test performed to reveal significant differences among the pair-wise combinations of regions investigated Site (U) Site (T) Mean difference (U–T) P value Lower limit Upper limit Flic-en-flac Roches Noires – 1.7878 0.0000 – 1.8143 – 1.7613 Tamarin 0.1478 0.0000 0.1254 0.1702 Roches Noires Tamarin 1.9356 0.0000 1.9139 1.9572 Also included is the mean difference, P value and 95% confidence intervals (Lower and upper limits) • H : No significant difference exists among the significant wave height datasets of spatially distant locations around the island. • H : Significant differences exist among the significant wave height datasets of spatially distant locations around the island. Results from Table 2 indicates the fact that the differences in significant wave height measurements taken at these three sites are highly significant (F = 19310.26, P value =0). This, therefore, leads to the rejection of the null hypothesis in favor of the alternative hypothesis. Having detected significant differences among the data - sets, a post-hoc analysis is performed in view of revealing Fig. 4 Multiple comparisons of group means test to check for differ - ences among wave regimes of the three sites. Also included are the those regions which differed significantly. Comparison tests 95% confidence intervals are employed through the Tukey–Kramer method and the main results are highlighted in Table 3. Through this analy- sis it is revealed that all sites studied differ statistically in of Flic-en-Flac and Roches Noires or Tamarin and Roches pair-wise comparison from one another. This demonstrates Noires than for the regions of Flic-en-Flac and Tamarin. the fairly different wave regimes, attributed to the different This reflects the difference in wave regimes that exists hydrologic and coastal environments which have been cho- between eastern and western waters of Mauritius. sen for investigation. Since the ANOVA test cannot determine which particular pairs are ’significantly different’, a multiple comparisons of Distributions of wave measurements group means test is carried out to support the Tukey–Kramer test previously performed. The multiple comparisons of The main distributions of the significant wave height data- group means test is used to check for differences among the sets are shown in the boxplots diagram of Fig. 5. It is evident wave regimes prevailing at the three sites investigated as from the figure that the region of Roches Noires witnesses illustrated in Fig. 4. Also included are the lower and upper a wave climate characterized by higher wave regimes with limits for the 95% confidence intervals for the true mean a mean value of 2.28 m and larger spread in the distribu- difference. The comparison intervals for the wave datasets tion than the other regions identified. The estuarine region corresponding to the three different sites do not intersect. of Tamarin records the lowest spread with a mean value of This lack of intersection indicates that the means in signifi- 0.34 m. This site detects a significant number of particularly cant wave height of the three regions investigated are signifi- high extreme values (outliers), reflecting on the unusually cantly different from each other. However, as observed from high occasional wave regimes of the estuarine structure. Fig. 4, the significant difference is greater for the regions The region of Flic-en-flac observes a mean values of 0.49 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 143 regions also play a major role in the hydrodynamic processes taking place by surface waves. The region of Tamarin offers the least likely environ- ment to place a wave energy extraction due to its low wave regimes depicted by the low mean (0.34 m) and high coef- ficient of variation (79%), delineating the high variations in significant wave height throughout the temporal scale investigated. It can be observed from spectral time series of Fig. 6Ab that the occurrence of swells in the estuary of Tamarin is spasmodic in nature, characterized by sudden outbursts of energy which occurs infrequently. This char- acteristic is due to the relatively less exposed geographi- cal location of the region pertaining to the incoming swells coming from south, southwest of the island. The second western site of Flic-en-Flac (Fig. 6A) investi- Fig. 5 Boxplot distributions of significant wave height records for gated demonstrated a higher mean in significant wave height three spatially distant sites around the coastal structure of Mauritius (0.49 m) with higher quartiles than the region of Tamarin. for the month of April. Also included is the presence of outliers, indi- cated by red crosses Moreover, a lower coefficient of variation (53.1%) was reg- istered at this site, indicating a lower variability and higher stability of the wave regimes for energy extraction as com- m with occasionally high waves, inherent of its hydrologic pared to Tamarin. On a comparative basis, the sandy shore and coastal structure. with reef site of Flic-en-Flac has higher wave energy poten- tial than the estuarine region of Tamarin despite the influ - ence of both hydrological systems from the incoming swells Variability and statistical interpretations of wave on the western regions of the island. data Interpretation of the distributions of the significant wave height datasets of the three sites reveal that the western sites Table 4 gives a descriptive statistics of the main distribu- of Flic-en-Flac and Tamarin with kurtosis of 6.41 and 7.14, tions for significant wave height (m) at the three sites of respectively, describe a sharper than normal distribution, investigation. The region of Roches Noires portrays the best characterizing them as leptokurtic. On the other hand, the wave climate for surface wave energy extraction with high eastern site of Roches Noires (kurtosis = 2.92) follow a plat- waves having a mean of 2.28 m that can reach a peak value ykurtic distribution with a flatter than normal distribution. of 3.42 m and the least variation as characterized by the low- All three sites demonstrate positively skewed distributions. est coefficient of variation (14.8%), indicative of the stability of the wave regime in the time frame considered. A low CV Dependence of eastern wave climate on trade winds is preferable to facilitate the sizing of a wave conversion system and decrease the likelihood of possible dumping of The wave regime of Roches Noires (Fig. 7a) is dependent converted wave energy system. on the trade winds originating from the subtropical high- It can be noted from the significant wave height measure- pressure zone and blowing in a southeasterly direction ments of Fig. 6 that the swell signatures of the regions of throughout the year. The predominance of the southeast Tamarin and Flic-en-Flac differ to some extent. This can trade winds influences mostly the eastern and southern be explained through the geographical location of the sites coastal microclimate wave systems of the island. Fig. 7b with respect to the incoming swells. The bathymetry of these shows the incoming wave direction at Roches Noires Table 4 Descriptive statistics which includes the mean, standard (Q1), median (Med.), upper quartile (Q3), maximum (Max.), skew- error mean (SE Mean), standard deviation (SD), coefficient of vari- ness (Skew.) and kurtosis (Kurt.) of significant wave height measured ation (CV) measured in percentage, minimum (Min.), lower quartile in meters at the three sites of investigation for the month of April Site Mean SE Mean SD CV Min. Q1 Med. Q3 Max. Skew. Kurt. Flic-en-Flac 0.493 0.0073 0.262 53.1 0.150 0.306 0.434 0.614 2.02 1.55 6.41 R. Noires 2.276 0.0089 0.336 14.8 1.575 2.031 2.214 2.513 3.42 0.65 2.92 Tamarin 0.341 0.0045 0.270 79.0 0.069 0.170 0.230 0.409 1.92 1.98 7.14 1 3 144 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Fig. 6 A Analysis of irregular offshore wave data at Flic-en-Flac for B Analysis of irregular offshore wave data at Tamarin for the month the month of April: a time series of significant wave height; b spec- of April: a time series of significant wave height; b spectral time tral time series of surface waves recorded; c zoomed images of ener- series of surface waves recorded; c zoomed images of energetic wave getic wave events occurring during the 18 days period at Flic-en-Flac. events occurring during the 26 days period at Tamarin Fig. 7 Irregular offshore wave records at Roches Noires for the month cant wave height approaching in an ESE direction under the influence of April which has been averaged over 3 years (2012–2014): a time of trade winds series of significant wave height; b wave rose representing the signifi- for the month of April, averaged over 3-years period is mostly due to the prevailing trade winds. It has been (2012–2014). The directional rose shows that the incom- reported that these waves typically possess peak wave ing waves generally approach in an east, southeast direc- periods of less than 12 s [24]. tion with the higher occurrence in the southeast direction with significant wave height peaking at 4 m, supporting the idea that the wave conditions prevailing at this site 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 145 among southern ocean basins permits extratropical south Dependence of western wave climate on Indian Ocean swells Indian Ocean swells originating from extratropical systems from adjacent extratropical areas to freely penetrate adjacent Figure 8 depicts the evolution of the eastward propagating ocean basins. Storms within the extratropical south Atlantic, Indian and Pacific oceans generally propagate towards the swells from the Atlantic Ocean, after travelling beyond the coasts of South Africa towards the Indian Ocean. Since east, below 40 S, with maximum storm densities occurring on the western flanks of these ocean basins near 50 S [25]. waves are very directional and tend to get shadowed by land masses, Mauritius situated in the line of sight from those The geographic location of Reunion Island reduces the wave energy potential of the coasts of Mauritius. Swell sources end up in the shadow region due to the geographi- cal location of Reunion island. This results in swell height heights of the order of 15.0 m, at the generation site is reduced to 6.0 m near Reunion Island [26]. It is revealed that peaking on average for the 3 days near the southern border of Mauritius at about 4.5 m and wave period of 18 s. Further the eastward propagating extratropical South Atlantic swell energy spreads through the entire Indian Ocean, attaining movement of the swell eastwards caused the shadow zone area to reduce, thereby resulting in swells to be observed the coasts of Thailand, Indonesia and southwestern Australia and penetrating as far as the Tasman Sea [25]. Owing to the at higher latitudes on the western flank of the island. Also, upon hitting Reunion Island, much of the wave energy gets directional nature of waves, the wave climate of the south and southwestern region of Mauritius changes according to dissipated. It is of interest to note that the southern region will witness the first swells followed progressively by the variations in swell events due to being exposed to the incoming swells. Consequently, as revealed in this study, regions of higher latitudes. Ocean swells are wind-waves generated by intense storms both western regions of Flic-en-Flac and Tamarin are influ- enced by swell events. that journeys long distances as they propagate away from their generation zone. Unlike the Northern Hemisphere Wave energy case study (Northern Atlantic and Pacific Ocean) where land masses provide a barrier which hinders the connection to other From the previous subsection, analysis led us to consider ocean basins, the Southern Hemisphere is open to the propa- gation of southern swell towards all ocean basins due to the the regions of Roches Noires as potential site for harness- ing wave energy. The higher waves and lower coefficient of fact that a circumpolar oceanic zone free of land barriers connects them together [25]. The wide connection existing variation of this site gives it an advantage on the other sites Fig. 8 Swell charts delineating the evolution of swell height across the Indian Ocean in temporal scale with zoomed imagery for the scenario close to Mauritius occurring for three selected dates 1 3 146 International Journal of Energy and Environmental Engineering (2018) 9:135–153 considered. It is of interest, however, to also probe into the variation (CV = 1.68) and standard deviation (SD = 2.30) wave energy potential of selected sites on the western part of indicates the highly variable wave energy regimes of this the island, whose wave regimes are dictated by the incoming locality. The quartiles of distribution were calculated and swells. Having compared the wave regimes of the two sites results show that the lower quartile, median and upper on the western flanks of the island, we provide a case study quartiles are 0.206, 0.644 and 1.52 kW/m, respectively. for the wave energy potential at the western sites of Flic-en- The distribution itself is positively skewed (4.67) and lep- Flac and Tamarin during the same above-mentioned periods tokurtic (38.1). as recorded from spectral measurements of the WTR. These The wave variability regimes of a particular sea state at a two sites provide contrasting wave regimes due to different specific location is often conveyed on a histogram of wave coastal environments, bathymetry and geographical location height and energy period. Figure 9b gives the occurrence to incoming swells. and wave energy scatters of significant wave height and The power flux for the region of Flic-en-Flac was com- energy period combinations for the 18 days records taken puted and the time series presented in Fig. 9a delineates at Flic-en-Flac. The significant wave height is arranged in the variations in temporal scales. Statistical interpreta- 0.05 m bins while energy period is organized in 0.5 s inter- tion of the time series shows that the mean power flux at vals. The color ramp represents the number of wave records this site is 1.369 kW/m with a peak value of 33.7 kW/m falling within the bins, giving an indication on the occur- recorded for the month of April. A high coefficient of rence of a particular wave state. That of Fig. 9c gives the Fig. 9 a Time series of wave power flux at Flic-en-Flac. Bivariate quency of wave records falling within a particular wave state while histograms showing the occurrence and energy distribution during that of figure (c) represent the total wave energy flux of a particular the 18 days period. The contour colors of figure (b) indicate the fre- wave state 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 147 total wave energy flux of the total number of points falling the relatively stable wave energy conditions prevailing at within the bins. Tamarin as compared to that of Flic-en-Flac. The lower, From Fig. 9b, c, it can be observed that the wave state of median and upper quartiles of distribution was found to be maximum occurrence does not coincide with the wave state 0.07, 0.133 and 0.426 kW/m, respectively, showing that the of maximum total energy flux. The most commonly occur - major portion of the distribution for wave energy computa- ring wave state has significant wave height in the interval tions at Tamarin lies significantly lower than that at Flic-en- 0.25–0.45 m and energy period in the interval 11.2–12.2s Flac. Similar to the distribution observed at Flic-en-Flac, while the most energetic wave state occurs for significant the region of Tamarin shows positively skewed distribution wave height in the interval of 0.58 and 0.75m and energy (3.86) and is characterized as leptokurtic (23.3). period in the interval 11.8 and 12.8s. For the case of Tamarin, the significant wave height is Variations in power flux for the region of Tamarin is arranged in 0.05 m bins while energy period is organized in presented in Fig. 10a and statistical analysis shows a mean 0.1 s intervals. It can be observed again that the wave state of power flux of around 0.484 kW/m with a peak value of 9.54 maximum occurrence does not coincide with the wave state kW/m, indicative of the lower potential of this site as com- of maximum total energy flux. The most commonly occur - pared to that of Flic-en-Flac. The coefficient of variation ring wave state has significant wave height in the interval (CV = 1.85) and standard deviation (SD = 0.894) reflects 0.15–0.25m and energy period in the interval 5.05–5.15s Fig. 10 a Time series of wave power flux at Tamarin. Bivariate histo- wave records falling within a particular wave state while that of figure grams showing the occurrence and energy distribution during the 26 (c) represent the total wave energy flux of a particular wave state days period. The contour colors of figure b indicate the frequency of 1 3 148 International Journal of Energy and Environmental Engineering (2018) 9:135–153 while the most energetic wave state occurs for significant of significant wave height were recorded from March to July wave height in the interval of 0.75–0.85m and energy period with a bimodal peak in April (3.42 m) and in July (3.38 m). in the interval 5.25 and 5.30s. It is evident from the plot that the peak summer months of An understanding of the performance of wave energy November and December record the lowest average signifi- converters in different wave states is essential for the con- cant wave height. The drop in significant wave height dur - structional details, cost and efficiency of the installed device. ing the summer period is due to lower southeast trade wind It would be preferable to install a wave energy converters in speeds observed during that same time scale [13]. a wave state having a maximum total energy flux rather than The wave roses showing the summer, winter and annual in one having maximum occurrence. directional significant wave height measurements averaged over the years 2012–2014 is presented in Fig. 12. It can be Long‑term analysis observed that the southeast trade winds influences the sig- nificant wave height to a greater extent in the winter period Figure 11 shows the annual distributions of significant wave with greater occurrence of southeasterly significant wave height on a monthly basis with standard deviation. It can be height measurements. However, on a comparative note, it observed that the winter half extending from May through can be observed that during summer, the magnitude of sig- October experiences the higher average significant wave nificant wave height measurements is greater with values height, reflecting on the greater wave power flux as com - peaking at 5–5.5 m as compared to the winter half where the pared to the relatively calm summer half. The highest values magnitude attains a lower value of 3.5–4.0 m. This reflects on the higher intensity of trade winds in the summer period. The boxplots representing the yearly mean evolution of significant wave height measurements are presented in Fig. 13. Data filtering processes have been performed, with the objective to remove missing data due to faulty instrument prior to boxplot representation. Of the 3 years investigated, the year 2013 records the larger spread in significant wave height measurements. This can be explained on the basis of the passage of cyclone ’Imelda’ within the vicinity of the island. Owing to relatively stable atmospheric conditions, the year 2014 recorded the lower spread of the distributions over the 3-year period. The greater number of high outliers is observed for the year 2012 while the median value over the 3-year period at Roches Noires seem to deviate not by much from the 2 m significant wave height mark. Taking into consideration the temporal variability of sig- nificant wave height measurements, an average has been per - Fig. 11 Average monthly significant wave height distributions for formed over the 3 years on the temporal basis (corresponding Roches Noires Fig. 12 Wave rose showing the magnitude and direction in which the incoming significant wave height is approaching the region of Roches Noires during (a) summer (b) winter (c) annual, averaged over the years 2012–2014 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 149 Figure  14a represents a scatter table which displays a better characterization of the composition of the wave energy resource throughout the 3 years. Each power flux was calculated with intervals of T = 2 s and H = p m0 0.5 m. From Fig. 14a, wave energy is mostly distributed between peak period T = 6 s and T = 26 s and signifi- p p cant wave height of H = 1.5 m and H = 5.5 m. From m0 m0 this distribution, the optimal energy is obtained at T = 16 s and H = 5.5 m. When further evaluating the total m0 power flux distribution by including the occurrence for each wave, its optimal distribution shifts to the right as shown in Fig. 14b. This implies that the highest amount of waves which contribute to a maximum energy output lies at T = 26 s and between H = 1.5 m and H = 4.0 m. p m0 m0 Fig. 13 Boxplots representing yearly distributions in significant wave Seasonal variation analysis at Roches Noires shows height at Roches Noires that the summer season with a SV index value of 34.3, is highly variable as compared to the winter season whose to timestamp records). Each of the yearly significant wave SV value approximates to 5.97. Moreover, the mean wave height dataset contained 15,948 records and when averag- power observed in the summer season is 28.8 kW/m and ing the 3 years corresponding to the day and time of the is found to be lower as compared to the winter season year, results in the average 2012–2014 dataset also contain- (31.7 kW/m) averaged over 3 years interval. The winter ing 15,948 records. From the ’average 2012–2014’ box plot, season with higher observed wave power flux and lower the median lies near the 2 m significant wave height mark, variability demonstrates a higher potential for extracting with upper and lower quartiles corresponding to 2.3 and 1.7 wave energy for this site. m, respectively. Fig. 14 Scatter table represent- ing 3 years (a) occurrence wave power matrix (kW/m) (b) total wave energy flux power matrix at Roches Noires from peak period (T ) and significant wave height (H ) m0 1 3 150 International Journal of Energy and Environmental Engineering (2018) 9:135–153 hand, in the post-construction stage, lower than expected Feasibility of setting up a wave energy farm energy output and high operations and maintenance costs may pose an economic risk for implementing the project in Any area with yearly averages of 15 kW/m has the potential to generate wave energy [27]. As recognized, the yearly aver- the waters of Mauritius. Besides, proper infrastructure costs need to be taken into consideration to build a rigid system age wave resource potential of the region of Roches Noires found to be 29.7 kW/m is much higher than this threshold capable of withstanding cyclonic weather conditions and storm surges to avoid excess expenditures for the repair of value and is characterized by low variations throughout the year due to persistent trade winds. A global technical damaged systems. For economic feasibility, wave power plants must be potential of 500 GW is expected for offshore wave energy devices having an efficiency of 40% and installed near coast- manufactured for an operating lifetime of at least 20 years [33]. The budget allocated to wave energy plant is spent as: lines with wave climates of the order of 30 kW/m [28]. This highly encourages the placement of a wave energy converter pre-operating cost, capital expenditure, operational expendi- ture and decommissioning costs [34]. Pre-operating costs at that location. includes expenditures encountered in pre-construction stage, environmental impact assessment among others. Capital Maintenance operations on wave energy farm expenditure costs includes the main structure as well as installation expenses. The operational and maintenance costs Operation and maintenance are needed to maintain the opti- mum mechanical performance of the wave farm through an are crucial for plant lifetime. The whole plant is supposed to be dismantled after 20 years and the decommissioning increase in productivity and reduction in running costs of wave energy converters [29]. The design for reliability and cost is estimated to be 0.5–1% of the initial investment [35]. According to a study, a 500 MW wave power plant set up in maintenance schedules need to be cautious due to the fact that system failures can be aggravated and there may be Oregon would entail a combined construction cost estimate of about $750,000/MW [36]. According to the same study, no appropriate weather window immediately available to retrieve the devices in case of failure [30]. Consequently, the implementation of such a project could produce over $90 million annually in new tax revenues on a statewide the plan to install a wave energy device in the offshore areas of Mauritius should come with an appropriate maintenance basis [36]. The introduction of a wave energy farm in the vicinity of strategy. There exists several maintenance strategies for wave energy converters which includes corrective strategy the island would entail numerous economic benefits in the long run. Integrating wave energy resources in the energy performed after failure or preventive strategy done before a breakdown. mix of Mauritius would ensure security of energy supply and the transition to greener alternatives. Consequently, in A systematic maintenance strategy exists which is based on systematic maintenance actions performed every 6 the long run, the island may become less dependent on high costs fossil fuel imports. The higher wave energy potential months combined with corrective maintenance action per- formed in case of breakdown. Despite the fact that system- site of Roches Noires identified, coupled to appropriate maintenance strategies planned are important aspects that atic maintenance strategy presents a wide variety of main- tenance actions, it is identified as the most costly strategy contribute to the longevity and economic yield of a typical wave farm to be implemented in that region. [31]. Nonetheless, it would suit the climate of Mauritius since as observed in the results section, the summer season Technological assessment experiences lower average significant wave height, reflecting a lower wave energy flux. Consequently, it would be appro- Several technologies that are capable of absorbing energy priate to plan proper maintenance strategies in the summer half to be undertaken at intervals of 6 months. from waves and converting it into electrical energy have been developed. However, the selection of appropriate wave Economic evaluation device depends on the water depth and the location (shore- line, near-shore, offshore) considered for construction [37]. Wave energy developers target high economic return with Classification of wave energy technologies is based on work - ing principles and are mainly grouped into oscillating water the intention to satisfy investor’s vested interests and gener- ate profits. In addition to expected rate-of-return, investment column (with air turbine), oscillating bodies (with hydraulic motor, hydraulic turbine and linear electrical generator) and risks that exist before and after project implementation is an important factor that cannot be neglected [32]. In the overtopping systems (with low-head hydraulic turbine) [37]. Each of them can be mounted on a fixed structure or placed pre-construction stage, developers may experience friction with local inhabitants and marine conservation groups per- on a floating platform. The Mutriku Breakwater Wave Plant located in Spain, having an installed capacity of 296 kW taining to the implementation of the project. On the other 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 151 from 16 turbines and 16 oscillating water columns, produced be influenced by episodic swell events originating from and exported over 1.3 GWh of power to the Basque grid the Atlantic Ocean with the region of Flic-en-Flac more during its first 5 years [38]. Nevertheless, the efficiency of exposed and the estuarine environment of Tamarin less energy capture is a major limitation of wave energy convert- exposed to the SSW incoming swells. On the eastern side ers. According to one study, out of 2700 GW power gener- of the island, the region of Roches Noires is observed to ated from waves, only about 500 GW can be captured using have higher wave resource potential with lower variabil- present technologies [39]. ity and higher wave power flux. Wave rose plots depicted the fact that the incoming waves from Roches Noires are International standardization influenced by the prevailing southeast trade winds, with the winter season displaying a greater influence owing to For implementing the wave farm in the waters of Roches the greater occurrence of incoming waves from the south Noires, the IEC standard guideline (62600-101, 2015) east during that period. The mean high wave power flux which aims at defining good practices in the field of wave registered at Roches Noires (28.8 kW/m in summer and resource assessment needs to be taken into consideration. 31.7 kW/m in winter) demonstrates the relatively higher Having identified the region of Roches Noires as potentially potential of exploiting surface wave energy from this site suitable site for installing the wave energy converters, the as compared to the western regions of Flic-en-Flac and next phase involves performing a spatial analysis of wave Tamarin. Consequently, for the tropical island of Mauri- energy resources in that region. The IEC standard guideline tius, efficient harnessing of wave energy will be performed establishes a system for estimating, analyzing and report- through the installation of wave energy converters around ing the wave energy resource at sites identified having good the eastern and southeastern coasts of the island, where wave energy resources [40]. The report provides methods the system is influenced by wind–wave generated system regarding data collection and resolution alongside with the as compared to the western coast which is under the influ- development of numerical model for wave energy resource ence of extratropical South Indian Ocean swells. Having estimation. In the data collection section, the need for prepa- identified the eastern region of the island as most prom- ration of a bathymetry contour map is highlighted and will ising for harnessing energy from waves, further orienta- be used to construct a Digital Elevation Model for the wave tions of this research involve performing a spatial analysis propagation model. Moreover, the section includes the fact of the wave resource potential of this study area through that archived data may be used as primary data source while numerical modelling techniques. Numerical simulations wave measurements can be used to validate the numerical coupled to a feasibility study on wave energy penetration model. Additional data which may be of use includes wind, in the eastern coasts of Roches Noires will be investigated tide, current and water density. in the next phase. In the numerical model section, production of at least 10 years of sea state data, with wave resource taken as station- Data availability The data that support the findings of this ary, is required. The minimum frequency for sea state data is study are available from the Mauritius Oceanography Insti- one dataset recorded at intervals of 3 h. Guidelines regarding tute but restrictions apply to the availability of these data, the configuration of boundary conditions, data processing which were used under approval for the current study, and so and validation of numerical modelling as well as dealing are not publicly available. Data are, however, available from with missing data are mentioned. The practical implications the author upon reasonable request and with permission of of the report has been evaluated through comparison with the Mauritius Oceanography Institute. the Biscay Marine Energy Platform (BiMEP) facility which is situated on the Basque coast. Consequently, the case study Acknowledgements We would like to acknowledge the African Moni- may be of relevance to the Government of Mauritius or any toring of the Environment for Sustainable Development (AMESD) pro- organizations considering the implementation of a wave ject and Mr Eric Martial for grating permission to use the wave buoy farm in the identified site of Roches Noires. data. We extend our gratitude to Magicseaweed.com who has cordially agreed to provide us with the swell charts. Special thanks to Arnaud Nicolas and Khishma Modoosoodun–Nicolas for their assistance in the study. We are grateful to the Mauritius Oceanography Institute Conclusions (MOI) for providing the adequate facilities for carrying out research presented in this paper This paper presents an assessment of the wave energy potential of three sites having distinct coastal environ- Compliance with ethical standards ments and wave regimes. Two sites located on the west- ern (leeward) side region of the island are observed to Conflict of interest On behalf of all authors, the corresponding author states that there is no conflict of interest. 1 3 152 International Journal of Energy and Environmental Engineering (2018) 9:135–153 Open Access This article is distributed under the terms of the Creative 16. Nordstrom, K.F.: Bayside beach dynamics: implications for Commons Attribution 4.0 International License (http ://crea tive comm simulation modeling on eroding sheltered tidal beaches. Mar. ons.org/licenses /b y/4.0/), which permits unrestricted use, distribution, Geol. 25(4), 333–342 (1977) and reproduction in any medium, provided you give appropriate credit 17. Warshaw, M.: The encyclopedia of surfing. Houghton Mifflin to the original author(s) and the source, provide a link to the Creative Harcourt, Boston (2005) Commons license, and indicate if changes were made. 18. IOC: IOC-UNEP-WMO-SAREC Planning Workshop on an Integrated Approach to Coastal Erosion, Sea Level Changes and their Impacts. http://unes doc.unes co.or g/images/0010 /0010 01/1001 38eo .pdf (1994). Accessed 15 May 2017 References 19. Ashton, I.G.C., Johanning, L.: On errors in low frequency wave measurements from wave buoys. Ocean Eng. 95, 11–22 (2015) 1. SM: Population and Vital Statistics—Year 2016. http: //stats mau 20. Joosten, H.: Directional wave buoys and their elastic mooring. riti us.govm u.org/Engl ish/Publ icat ions /Page s/Pop_and_Vita l_ Int. Ocean Syst. 10(4), 18–21 (2006) Stat s_Yr16 .aspx (2017a). Accessed 7 April 2017 21. Sierra, J.P., Martn, C., Msso, C., Mestres, M., Jebbad, R.: Wave 2. SM: Energy and Water Statistics—Year 2016. http://s tatsmau r iti energy potential along the Atlantic coast of Morocco. Renew. us.govm u.org/Engl ish/Stat sbyS ubj/Page s/Ener gy_Wate r_Stat Energy 96, 20–32 (2016) s_Year _2016 .aspx (2017b). Accessed 9 April 2017 22. Rusu, E., Soares, C.G.: Wave energy pattern around the Madeira 3. MPU: Long-Term Energy Strategy 2009–2025. https://sus taina Islands. Energy 45(1), 771–785 (2012) bled evel opme nt.un.org/cont ent/docu ment s/1245 maur itiu sEne 23. Cornett, A. M.: A global wave energy resource assessment. In rgy%20St rate gy.pdf (2009). Accessed 10 April 2017 The Eighteenth International Offshore and Polar Engineering 4. Lannoye, E.: Renewable energy integration: practical manage- Conference. In: International Society of Offshore and Polar ment of variability, uncertainty, and flexibility in power grids Engineers (2008) [book reviews]. IEEE Power Energy Mag. 13(6), 106–107 24. Baird Report: Proposal for Controlling Severe Erosion and (2015) Beach Improvement works in front of The Grand Mauritian 5. Neill, S.P., Lewis, M.J., Hashemi, M.R., Slater, E., Lawrence, Hotel, Balaclava. http://environment.govmu.org/English/ J., Spall, S.A.: Inter-annual and inter-seasonal variability of eia/Documents/Repor ts/blueocean/grand/chapt2.pdf(2003). the Orkney wave power resource. Appl. Energy 132, 339–348 Accessed 20 June 2017 (2014) 25. Alves, J.H.G.: Numerical modeling of ocean swell contributions 6. Ram, K., Narayan, S., Ahmed, M.R., Nakavulevu, P., Lee, Y.H.: to the global wind-wave climate. Ocean Model. 11(1), 98–122 In situ near-shore wave resource assessment in the Fiji Islands. (2006) Energy Sustain. Dev. 23, 170–178 (2014) 26. Samiksha, S. V., Vethamony, P., Aboobacker, V. M., Rashmi, R.: 7. Bernardino, M., Rusu, L., Soares, C.G.: Evaluation of the wave Propagation of Atlantic Ocean swells in the north Indian Ocean: energy resources in the Cape Verde Islands. Renew. Energy 101, a case study. Nat. Haz. Ear. Sys. Sci. 12, 3605–3615 (2012) 316–326 (2017) 27. Nelson, V.C., Starcher, K.L.: Introduction to renewable energy. 8. Iglesias, G., Carballo, R.: Wave resource in El Hierroan island CRC press, Boca Roton (2015) towards energy self-sufficiency. Renew. Energy 36(2), 689–698 28. Sims, R. E., Schock, R. N., Adegbululgbe, A., Fenhann, J. V., (2011) Konstantinaviciute, I., Moomaw, W., Uchiyama, Y.: Energy sup- 9. Nordstrom, K.F.: Estuarine Beaches: an introduction to the ply. In Climate change 2007: Mitigation. In: Contribution of physical and human factors affecting use and management of Working Group III to the fourth assessment report of the Inter- beaches in estuaries, lagoons, bays and fjords. Springer Science governmental Panel on Climate Change. Cambridge University & Business Media, Netherlands (1992) Press (2007) 10. Ludwick, J. C.: Mechanisms of sand loss from an estuarine 29. Rinaldi, G., Thies, P.R., Walker, R., Johanning, L.: On the anal- groin system following an artificial sand fill. In: Technical ysis of a wave energy farm with focus on maintenance opera- Report 87-2 Department of Oceanography, Old Dominion Uni- tions. J. Mar. Sci. Eng. 4(3), 51 (2016) versity, Norfolk, VA (1987) 30. Cruz, J.: Ocean wave energy: current status and future prespec- 11. Lutz, W., Wils, A. B.: People on Mauritius: 1638–1991. In Pop- tives. Springer Science & Business Media, Berlin (2007) ulationDevelopmentEnvironment (pp. 75-97). Springer, Berlin, 31. Sahnoun, M. H., Baudry, D., Mustafee, N., Louis, A., Smart, P. Heidelberg(1994) A., Godsiff, P., Mazari, B.: Modelling and simulation of opera- 12. ASCLME: National Marine Ecosystem Diagnostic Analysis, tion and maintenance strategy for offshore wind farms based on Mauritus. http://www .asclme.or g/repor ts2013/N ational %20ME multi-agent system. J. Intell. Manuf. 1, 1–17 (2015) DAs/MEDA %20Ma urit ius/_Maur itiu sMED APro of.pdf(2012). 32. Van Haaren, R., Fthenakis, V.: GIS-based wind farm site selec- Accessed 19 April 2017 tion using spatial multi-criteria analysis (SMCA): evaluating the 13. JICA: The Project for Capacity Development on Coastal Protec- case for New York State. Renew. Sustain. Energy Rev. 15(7), tion NIPPON KOEI CO., LTD. and Rehabilitation in the Repub- 3332–3340 (2011) lic of Mauritius. http://open_jicareport.jica.go.jp/pdf/12237327. 33. Wengenmayr, R., Buhrke, T.: Renewable energy: sustainable pdf (2015). Accessed 2 May 2017 energy concepts for the future. Wiley, Hoboken (2011) 14. Anthony, A., Atwood, J., August, P., Byron, C., Cobb, S., Foster, 34. Astariz, S., Iglesias, G.: Enhancing wave energy competitive- C., Fry, C., Gold, A., Hagos, K., Heffner, L., Kellog, Q., Lel- ness through co-located wind and wave energy farms. A review lis, K.,Opaluch, J., Oviatt, C., Pfeiffer-Herbert, A., Rohr, N., on the shadow effect. Energies 8(7), 7344–7366 (2015) Smith, L.,Smyth, T., Swift, J., Vinhateiro, N.: Coastal lagoons 35. Castano, M.: Sistema de Monitorización y Supervisión de Una and climate change: ecological and social ramifications in US Boya Para Generación de Energía Undimotriz. Universidad Atlantic and Gulf Coast ecosystems. Ecol. Soc. 14(1), 8 (2009) Politécnica de Cataluña, Barcelona, Spain (2011) 15. Nordstrom, K.F., Jackson, N.L.: Migration of swash zone, 36. OWET: Economic Impact Analysis of Wave Energy: Phase One. step and microtopographic features during tidal cycles on an https://oregonwave.org/oceanic/wp-content/uploads/2013/09/ estuarine beach, Delaware Bay, New Jersey, USA. Mar. Geol. Economic-Im pact- Anal y sis-of- Wa v e-Ener gy -Phase- 92(1–2), 147–154 (1990) 1 3 International Journal of Energy and Environmental Engineering (2018) 9:135–153 153 One%E2%80%94September-2009.pdf(2009). Accessed 15 Dec 40. OPERA: Wave Energy Measurement Methodologies for IEC 2017 Technical Specifications.http://oper a-h202 0.eu/wp-cont ent/uplo 37. Antonio, F.D.O.: Wave energy utilization: a review of the tech-ads/2016 /03/OPER A_D5.1_IEC_TS_Appl icat ion_UCC_2016 nologies. Renew. Sustain. Energy Rev. 14(3), 899–918 (2010)1129 _v2.1.pdf (2016). Accessed 18 Dec 2017 38. TidalEnergyToday: Mutriku wave plant generates over 1 GWh of clean power. http s://tida lene rgyt oday .com/2016 /07/19/mutr Publisher’s Note Springer Nature remains neutral with regard to iku-wave -plan t-gene rate s-over -1gwh -of-clea n-powe r/ (2016). jurisdictional claims in published maps and institutional affiliations. Accessed 16 Dec 2017 39. McCormick, M.E., Ertekin, R.C.: Renewable sea power. Mech. Eng. 131(5), 36 (2009) 1 3

Journal

International Journal of Energy and Environmental EngineeringSpringer Journals

Published: Jan 18, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off