Presentation Status: Final Version: 01 Date: 31-Mar-2015 EC - - PDF document
Marine Renewables Infrastructure Network WP2: Marine Energy System Testing - Standardisation and Best Practice Deliverable 2.9 Standards for Wave Data Analysis, Archival and Presentation Status: Final Version: 01 Date: 31-Mar-2015 EC FP7
EC FP7 Capacities: Research Infrastructures Grant Agreement No. 262552, MARINET
D2.9Standards for Wave Data Analysis, Archival and Presentation
Page 2 of 42
MARINET (Marine Renewables Infrastructure Network for emerging Energy Technologies) is an EC-funded network
energy - wave, tidal & offshore-wind. The initiative is funded through the EC's Seventh Framework Programme (FP7) and runs for four years until 2015. The network of 29 partners with 42 specialist marine research facilities is spread across 11 EU countries and 1 International Cooperation Partner Country (Brazil). MARINET offers periods of free-of-charge access to test facilities at a range of world-class research centres. Companies and research groups can avail of this Transnational Access (TA) to test devices at any scale in areas such as wave energy, tidal energy, offshore-wind energy and environmental data or to conduct tests on cross-cutting areas such as power take-off systems, grid integration, materials or moorings. In total, over 700 weeks of access is available to an estimated 300 projects and 800 external users, with at least four calls for access applications over the 4-year initiative. MARINET partners are also working to implement common standards for testing in order to streamline the development process, conducting research to improve testing capabilities across the network, providing training at various facilities in the network in order to enhance personnel expertise and organising industry networking events in order to facilitate partnerships and knowledge exchange. The initiative consists of five main Work Package focus areas: Management & Administration, Standardisation & Best Practice, Transnational Access & Networking, Research, Training & Dissemination. The aim is to streamline the capabilities of test infrastructures in order to enhance their impact and accelerate the commercialisation of marine renewable energy. See www.fp7-marinet.eu for more details.
Partners
Ireland University College Cork, HMRC (UCC_HMRC) Coordinator Sustainable Energy Authority of Ireland (SEAI_OEDU) Denmark Aalborg Universitet (AAU) Danmarks Tekniske Universitet (RISOE) France Ecole Centrale de Nantes (ECN) Institut Français de Recherche Pour l'Exploitation de la Mer (IFREMER) United Kingdom National Renewable Energy Centre Ltd. (NAREC) The University of Exeter (UNEXE) European Marine Energy Centre Ltd. (EMEC) University of Strathclyde (UNI_STRATH) The University of Edinburgh (UEDIN) Queen’s University Belfast (QUB) Plymouth University(PU) Spain Ente Vasco de la Energía (EVE) Tecnalia Research & Innovation Foundation (TECNALIA) Belgium 1-Tech (1_TECH) Netherlands Stichting Tidal Testing Centre (TTC) Stichting Energieonderzoek Centrum Nederland (ECNeth) Germany Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V (Fh_IWES) Gottfried Wilhelm Leibniz Universität Hannover (LUH) Universitaet Stuttgart (USTUTT) Portugal Wave Energy Centre – Centro de Energia das Ondas (WavEC) Italy Università degli Studi di Firenze (UNIFI-CRIACIV) Università degli Studi di Firenze (UNIFI-PIN) Università degli Studi della Tuscia (UNI_TUS) Consiglio Nazionale delle Ricerche (CNR-INSEAN) Brazil Instituto de Pesquisas Tecnológicas do Estado de São Paulo S.A. (IPT) Norway Sintef Energi AS (SINTEF) Norges Teknisk-Naturvitenskapelige Universitet (NTNU)
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 0 of 39
Title Standards for Wave Data Analysis, Archival and Presentation Distribution Work Package Partners Document Reference MARINET-D2.9 Deliverable Leader José Cândido WAVEC Contributing Authors Matthew Finn EMEC Keith Dampney EMEC John Lawrence EMEC Lucia Margheritini AAU
Rev. Date Description Prepared by (Name & Org.) Approved By (Task/Work- Package Leader) Status (Draft/Final) 01
The work described in this publication has received support from the European Community - Research Infrastructure Action under the FP7 “Capacities” Specific Programme through grant agreement number 262552, MaRINET.
The views expressed, and responsibility for the content of this publication, lie solely with the authors. The European Commission is not liable for any use that may be made of the information contained herein. This work may rely on data from sources external to the MARINET project Consortium. Members of the Consortium do not accept liability for loss or damage suffered by any third party as a result of errors or inaccuracies in such data. The information in this document is provided “as is” and no guarantee or warranty is given that the information is fit for any particular
member of the MARINET Consortium is liable for any use that may be made of the information.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 1 of 39
The definition of standards is a dynamic process that follows closely the development of technology, in a constant interchange of information. A continued update and reorientation of recommendations and guidelines to track the evolution of wave energy towards the market is of the essence. The work developed in the scope of the MaRINET project, and in particular the present deliverable, is strategically timely. This report intends to provide an up-to-date contribution to the definition of a set of standards to be applied in the handling of wave data, based on the experience and expertise of MaRINET’s consortium and ensemble of world-class wave test facilities. The report covers a brief review of the standardization efforts carried out to date, the description of the type of wave data commonly considered for wave energy purposes and the typical data processing outputs, common and suggested practices in wave measurement and data analysis, as well as data archival and presentation approaches to be followed. The document uses the MaRINET experience to provide its contribution towards wave data standardization. Methods currently adopted by two open sea test facilities within MaRINET are taken to be up-to-date operational procedures and suggested as recommended approaches. It is concluded that binned tables are a preferable means to present data and one that technology developers commonly use to create the necessary interface between technology performance and available resource, in order to report device efficiency. Specific recommendations for the creation of contingency tables are provided.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 2 of 39
1 INTRODUCTION .............................................................................................................................................3 1.1 CONTEXTUALIZATION OF THE DELIVERABLE ......................................................................................................... 3 1.2 RELATION TO PREVIOUS WORK ................................................................................................................................ 3 2 WAVE DATA ..................................................................................................................................................5 2.1 CHARACTERIZATION OF OCEAN WAVES ..................................................................................................................... 5 2.2 WAVE DATA PROCESSING OUTPUTS ....................................................................................................................... 10 3 WAVE MEASUREMENTS AND DATA ANALYSIS .............................................................................................. 11 3.1 RAW WAVE MEASUREMENTS DATA ........................................................................................................................ 11 3.2 WAVE DATA ANALYSIS PROCEDURES ...................................................................................................................... 15 4 WAVE DATA ARCHIVAL AND PRESENTATION ................................................................................................ 19 5 CONCLUSIONS AND RECOMMENDATIONS .................................................................................................... 37 6 REFERENCES ................................................................................................................................................ 38
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 3 of 39
The need for a consensual set of standards in wave energy, from the several stages of development (numerical modelling, wave flume and/or tank testing, sea trial) to the design of wave energy converters (WECs) itself, has been widely discussed and treated in publications and a variety of research projects (e.g EquiMar). In the scope of MaRINET, in particular, Deliverable 2.14 (MaRINET, 2012a) addressed data storage, including storage requirement and data availability, and wave data representation, focusing on common nomenclature and graphs, both at laboratory facilities and test centres. Deliverable 2.22 (MaRINET, 2014), on the other hand, focused on identifying commonalities between the different methodologies across the MaRINET facilities for data retrieval, storage, processing and presentation, through the collection of the results of a specific survey, thus preparing the field for possible avenues for confluence of approaches. The present document picks up on the findings of these initial efforts and builds towards a tentative proposal of wave data standards, in particular with respect to analysis, archival and presentation. The definition of priority areas where efforts should be focused is a common concern of Deliverable 2.14 and Deliverable 2.22. A common relevant
are controlled or produced on demand and wave data handling relies heavily on client customization, as in the field environment, where conditions must be accepted and there is a wider dispersion of the type of data that is measured and the possible measuring techniques. Furthermore, as highlighted in Deliverable 2.22, obvious similarities exist in the sort of data that is presented and the way it is presented throughout the different test facilities in MaRINET, from laboratories to test sites. Therefore, Deliverable 2.09 focuses primarily on common practices and standards with regard to wave data in the field environment. In other words, measurements, analysis techniques, typical archival procedures and presentation methods in test sites are the prime target of the present
test sites, offers the possibility to develop such intents.
Throughout the last decade, the lack of specific codes of practice, standards or guides in wave energy, as a consequence of the relatively new and unproven technology in the field, has been recognized. As a reaction to this circumstance, and also thanks to the relatively slow but steady path to commercialization, a few works and initiatives contributing to the development of a set of standards in the sector have recently come to light. In this context, the efforts under the EU FP7 project EquiMar and a few publications (e.g. Saulnier et al., 2006, Pitt, 2009a and Tietje et al., 2011) stand out. The insufficient maturity of the technology and the consequent scarcity of data recorded by WECs in actual open sea conditions have no doubt limited the ability of these efforts to produce truly binding results and limited their universal adoption. The reality is that standards and guidelines are bound to continuously evolve as the findings from ongoing research and development lead to their refinement. Such reality is particularly evident in the case of a fast-developing pre-commercial technology, as is the case of wave energy. A continued update and reorientation of recommendations and guidelines to track the development of wave energy towards the market is of the essence. In the light of this, the work developed in the scope of the MaRINET project, and in particular the present deliverable, is strategically timely. A significant part of the work on wave energy standards and guidelines developed to date focus on the actual WEC performance assessment. However, the definition of common methodological avenues for the handling of the underlying environmental data is equally crucial to move the sector forward towards economic competiveness. Although a work of this nature may not directly reflect the standpoint and needs of the technology developer, it will be essential to facilitate the comparison between approaches across different test sites. The resulting harmonized comparative procedures will contribute to strengthening the necessary bridge between developers and test site
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 4 of 39
the technologies. This is precisely the focus of the present document. In the scope of the EquiMar project, protocols for wave energy resource assessment have been produced in Deliverable 2.7 (EquiMar, 2010a). The document provided recommendations for resource assessment at different stages of a wave energy project development, by resorting both to in-situ measurements and numerical modelling
presentation of data obtained from sea trials of ocean energy converters (wave and tidal). The document was elaborated with a view to describe a logical and widely-applicable method to analyze and present data obtained from sea trials. The main focus was placed on the establishment of the foundations for the presentation of results
uncertainty involved with the stated performance. The ultimate objective was to contribute to a more accurate and credible comparison between the performance of devices and to an expeditious application of results to alternative locations. The European Marine Energy Centre (EMEC) has been developing extensive work on the design of standards in wave
storage of wave data. The recommendations can be summarized in the following:
the details may vary;
and/or no-data records;
containing information detailing the sampling configuration, data processing and recording of the data. According to this report, each sample data record should contain the following elements: i) Date and time stamp; ii) Depth in metres (variations associated with tides should be included if they lead to > 5 % changes in group velocity, referred to the maximum energy value of Te in the expected (Hs, Te) scatter diagram); iii) A frequency listing containing information about the variance spectrum and its directional characteristics as well as the power spectrum; iv) A number of derived parameters including power, mean wave direction, wave height and wave period; v) Quality control flags. Such requirements are further expanded in the document. This document focuses on standards for the procedures surrounding the analysis, archival and presentation of wave
agreements found during the project’s execution and the current state-of-the-art of wave energy as a whole, tentative proposals will be presented. The document does not include an exhaustive review of wave data archival and presentation, covered in MaRINET’s Deliverable 2.14, nor an in-depth analysis of the methodologies for data retrieval, storage, processing and presentation applied by the different MaRINET facilities, addressed in Deliverable 2.22. Likewise, details on wave measurement techniques, the different types of equipment and the respective working principles were discussed in Deliverable 2.1 (MaRINET, 2012b). Consequently, these aspects are not considered in Deliverable 2.9, apart from a few occasional exceptions required to elaborate on the matters addressed. Section 2 provides a description of the raw data and parameters typically measured and retrieved in the test sites. The following section describes typical procedures for wave measurements and data analysis. Section 4 addresses common approaches for wave data archival and presentation. In the final section, some conclusions are drawn and a few key recommendations provided.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 5 of 39
The characteristics of ocean waves in the time domain are determined through a wave-by-wave analysis of the surface elevation at a given location (time records). Through this type of analysis it is possible to determine both individual and integrated wave parameters. In the frequency domain, the determination of wave parameters is obtained from the spectral analysis of the wave records, based on the energy spectrum of the records. Parameters are computed through a Fourier analysis of the time series data, typically with a spectral resolution of 0.05 Hz, as recommended in Tucker, 1993. On a scale of tens of kilometres and minutes, the local state of the sea surface in deep water conditions may be accurately described by a stationary Gaussian random process. The local behaviour of the waves is then determined by the 2D wave spectrum S(f,θ). Spectral information is typically condensed in mean height, period and direction wave parameters, expressed as a function of the n-th moment spectral moment
∫ ∫
∞
=
2
) , ( θ θ
π
dfd f S f m
n n
. [ 1 ]
The significant wave height Hs is the most widely used wave height parameter. It is defined as the average of the highest one third of the trough to crest wave heights. The respective wave height spectral parameter is
s
H m Hm ≅ =
2 / 1
4 ,
[ 2 ]
where m is the zero-th spectral moment. As regards wave period, several parameters are commonly used. Spectral parameter
2 / 1 2
02
= m m Tm
[ 3 ]
reasonably corresponds to the mean zero-crossing period (the average time elapsed between two sequential crests). Since it depends on m2,
02
m
T is very sensitive to the high frequency spectral tail, which typically presents significant variability and low energy contents. The mean energy period Tm-10, commonly used for wave energy resource assessment purposes, is defined by
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 6 of 39
1 10
m m Tm
− −
= .
[ 4 ]
Since f T / 1 = , Tm-10 provides an average value of T weighted by the spectral distribution. This parameter mainly depends on the lower frequency band of the spectrum where most of the energy is concentrated. The peak period Tp is the inverse of the peak frequency, fp, to which corresponds the maximum spectral density value, so that
p p
f T 1 = .
[ 5 ]
The wave power level or flux of energy per unit crest length is given by
∫ ∫
∞
=
π
θ θ ρ
2 0 0
) , ( ) , ( dfd h f c f S g P
g
,
[ 6 ]
where ρ is the water density and g the gravity acceleration. The group velocity cg, corresponds to the velocity at which the energy propagates and is defined by k cg ∂ ∂ = ω .
[ 7 ]
In deep water conditions (in practice when the water depth h is larger than half wavelength) cg simplifies to f g cg π 4 =
[ 8 ]
so that the wave power is given by
1 2 2 0 0 1 2
4 ) , ( 4
− ∞ −
= =
∫ ∫
m g dfd f f S g P π ρ θ θ π ρ
π
.
[ 9 ]
This relation may be expressed in terms of Hm0 (i.e., Hs) and Tm-10 by
10 2 2
64
−
=
m m
T H g P π ρ .
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 7 of 39
[ 10 ]
Assuming that the sea water density is 1025 = ρ kgm-3, if Hm0 is expressed in meters and Tm-10 in seconds, the power level in kWm-1 is given by
10 2
4906 .
−
=
m m T
H P .
[ 11 ]
To overcome the limitations of equation [ 11 ], an alternative TTV (Transport of Total Variance) method to calculate the wave power is additionally used by EMEC (Pitt, 2005). This method provides a better representation in shallower water conditions, which is the case of EMEC’s Billia Croo site. The power level estimation using this method is given in kW/m by
2
6286
s e g
,h)H (T c . P = .
[ 12 ]
This expression in full turns to
2 2 1
2 sinh 2 1 tanh 2 1 6286 .
s
H kh kh kh k g P + =
[ 13 ]
with
+ = kh kh c cg 2 sinh 2 1 2 1
,
[ 14 ]
in which the phase velocity is given by the dispersion relation
2 1
tanh = kh k g c
[ 15 ]
A method based on Mollison (1986) to provide an indication of the effect of directionality on the wave spectra is also applied by EMEC. This provides firstly an estimate of the net power flux, Pflux , in direction θ , by multiplying the energy of each spectral component resolved in direction θ by the group velocity
g
c ,
∫
− = f f f S f c g P
g flux
d )) ( cos( ) ( ) ( ) ( θ ψ ρ ψ
[ 15 ]
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 8 of 39 where ψ is the directional alignment to the waves. In practical terms this can be considered as the power available in the waves from all directions projected on to a given WEC heading or directional alignment (ψ). The limits to the integration recognise that waves from opposite directions pass through each other, but only those within 2 / ) ( π θ ψ ± = − are able to transmit power. For each hourly record, it is then calculated (in W/m) by
∑
= =
∆ − =
64 1
) ( )) ( cos( ) (
i i i i g i i flux
f f c f S g P θ ψ ρ ψ
[ 16 ]
where ) ( i
g f
c is the group velocity at the ith frequency, Si is the corresponding spectral density estimate, ) ( i f θ is the spectral wave direction at the ith frequency,
i
f ∆ is the ith frequency increment (slice thickness) and ) ( i f θ ψ − is limited to ± π/2. In order to derive a more useful expression of directionality, the above calculation is repeated for ψ = 0 to 360° in 1° increments, and the power flux for every direction and the best direction Dδ (i.e. the calculated ψ which maximises the power flux Pflux) are obtained by this means. Another useful quantity is the directionality coefficient /P δ = Pflux .
[ 17 ]
The best direction and the directionality coefficient are calculated for a given period of time. In EMEC’s monthly data reports they are calculated for each 30 min record and for all of the monthly records combined, and are thus referred to as the half-hourly or monthly best direction or directionality coefficient. The overall mean wave direction may be obtained from the 2D spectrum by
∫ ∫ ∫ ∫
∞ ∞
=
π π
θ θ θ θ θ θ θ
2 0 0 2 0 0
) cos( ) , ( ) sin( ) , ( arctan df d f S df d f S .
[ 18 ]
The Alliance for Coastal Technologies (ACT)1 has found that the majority of coastal users collecting wave data require the first five Fourier coefficients (Alliance for Coastal Technologies, 2007). This report suggests that wave monitoring instrumentation should be able to provide at least the first five coefficients, which, nevertheless, are insufficient to fully characterize the directional spectrum. Additional equipment may be required to provide further Fourier moments.
1 The ACT (www.act-us.info) comprises research institutions, resource managers and private sector companies to aid the development and adoption of effective
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 9 of 39 Typical post-processing of directional buoys’ measurements provides only frequency spectra ) (f S , related to the directional spectrum by
∫
=
π
θ θ
2
) , ( ) ( d f S f S ,
[ 19 ]
the mean direction per frequency band ) (f θ and its spreading for each frequency band ) (f
θ
σ . Mean wave direction is then computed by
∫ ∫
∞ ∞
= )) ( cos( ) ( )) ( sin( ) ( arctan df f f S df f f S θ θ θ .
[ 20 ]
As suggested by a few authors (e.g. Saulnier et al., 2006), recent technological advances have made clear that the
such as spectral width and wave grouping. The energy spectral broadness parameter (or narrowness parameter)
2 / 1 2 1 2
1 − ⋅ = m m m ε
[ 21 ]
is often used to characterize spectral bandwidth due to its weak sensitivity to the cut-off frequency and its focus on the section of the spectrum with higher energy content (low-frequency waves). It provides an indication of the degree of tuning (or resonance) that a WEC will require to maximize energy extraction from the sea-state. Wave groupiness may be assessed using the wave correlation parameter K, which is related to group formation and is obtained from (Battjes et al., 1984)
2 / 1 2 2
) 2 sin( ) ( ) 2 cos( ) ( 1 + =
∫ ∫
∞ ∞
df fT f S df fT f S m K
z z
π π
[ 22 ]
The wave correlation parameter and Hs govern the bivariate Rayleigh distribution which has been used by Kimura (1980) for modelling the dependence of two successive (zero up-crossing) wave heights in offshore applications. The envelope theory (Longuet-Higgins, 1984) may also be used, although this is applicable only to very narrow spectra. Spectral peakedness may also be a quantity of interest, in particular to relate wave group length statistics to spectral
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 10 of 39
∫
= f f S f m Qp d ) ( 2
2 2
.
[ 23 ]
Typical wave data processing outputs for a detailed wave climate characterization and resource assessment in the test site may include average values of wave height, period, power and direction, scatter tables of Hs, Te and θ , a wave power exceedance curve, as well as seasonal and inter-annual variability tables. Generally, scatter tables present the frequency of occurrence of pairs of Hs, Te and θ . Additionally, energy distribution tables presenting the percentage of the total energy that occurs for each bin of (Hs, Te), (Hs, θ ) and (Te, c) may be computed. Table 1 summarizes long-term statistics commonly used to characterize the wave climate and energy resource in the test sites. Tables Plots Seasonal and Inter-annual Variability
density (scatter diagram) of (Hs,Te)
density (scatter diagram) of (Hs,Tp)
value of Hs (with confidence limits)
value of Te (with confidence limits)
value of P (with confidence limits)
value of θ (with confidence limits)
minimum and maximum value of Hs (with confidence limits)
minimum and maximum value of Te (with confidence limits)
minimum and maximum value of P (with confidence limits)
minimum and maximum value of θ (with confidence limits)
Table 1 Typical wave data post-processing outputs
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 11 of 39
According to EquiMar’s Deliverable 2.7 (EquiMar, 2010a), physical wave measurements should be performed using either wave measurement buoys or acoustic Doppler profilers (ADPs). As suggested by the document, other measurement methods such as remote sensing or pressure transducers are not recommended due of their lack of validation for wave energy resource assessment purposes. Directional buoys, such as Pitch-Roll-Heave (PRH) buoys, follow the surface slope of the waves, measuring time series of the vertical (heave) acceleration and two orthogonal components of the surface slope (pitch and roll). From each triplet of time series wave measurements, a directional spectrum can be estimated using the theory of Longuet-Higgins (Longuet-Higgins et al., 1963) to calculate the auto- and cross-spectra between the records (MaRINET, 2012b). Spectral parameters can thus be computed. Data is typically processed on-board. Both raw and processed data may be stored or transmitted to shore via a radio or satellite link. ADCPs (Acoustic Doppler Current Profilers or simply Acoustic Doppler Profilers) measure the wave orbital velocities below the sea surface. The ADCP measures the flow velocity of the water by transmitting short sound pulses of known frequency and measuring the Doppler shift of the signal reflected in scatterers assumed to be passively following the flow (MaRINET, 2012b). ADCPs use piezoelectric oscillators to transmit and receive the acoustic signals.
Since at least 3 vector components have to be estimated to measure 3D velocities, ADCPs typically include 4
transistors: 3 for measuring velocity (u, v, and w) and a 4th redundant transducer for error checking. For wave measurement purposes, ADCPs are bottom mounted, upward facing and they include a pressure sensor for additional non-directional wave measurement. ADCPs feature the capacity to measure flow velocity at several depth levels, typically up to 128 levels, with each series of pulses. The along-beam component of the orbital velocity for each depth cell is used to construct a virtual array (Strong et al., 2003). In other words, each cell in each beam is taken as an independent sensor in an array. The wave measurement uses the phase differences on the spatially separated virtual array. Cross-spectra, to which the directional spectrum is linearly related for a given frequency, are calculated between every sensor. Linear wave theory is used to translate the measurement from velocity spectra at various depths to surface displacement. As identified in MaRINET (2012a), the large majority of the open sea test facilities pooled in MaRINET use wave buoys to assess the wave climate and provide developers with real wave conditions in the test site, in order to support device design, marine operations and performance assessment. As a consequence, for the purposes of the discussion on data standardization and in the scope of the examples that will be provided, without loss of generality, this report will focus on wave data measured by this type of equipment. EMEC has currently deployed on test sites four Datawell directional Waverider MkIII (DWR-MkIII) buoys. The DWR- MkIII measures wave height by means of an accelerometer mounted onto a platform within the buoy that remains horizontal under any movement expected to be experienced at sea. Wave direction is measured by correlating horizontal motion of the buoy with the vertical motions. Two perpendicular sensors are used, one for horizontal motion when the buoy is upright and another for when the buoy is in tilt. A fluxgate compass converts the acceleration measured in the buoy to north-west coordinates (Datawell, 2014). In total the DWR-MkIII motion sensors measure 8 observables (see Figure 1).
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 12 of 39
Figure 1 Axes and signs of the DWR-MkIII motion sensors
Raw displacement data are generated by the buoy at a rate of 1.28Hz and stored within an onboard data logger flash
data are transferred via a high frequency (HF) radio link although there is capability to pass this data ashore via satellite or GSM link. From the HF radio link data are logged straight onto EMEC’s SCADA network and a non-quality controlled wave data feed is supplied directly from the SCADA and passed onto the EMEC website (EMEC, 2015). The DWR-MkIII’s onboard batteries have autonomy in continuous operation of up to 3 years. Recalibration of the measurement systems is recommended on a regular basis every 3 to 6 years. The internal wave spectrum is calculated at a sampling rate of 1.28Hz. Every 200s a total number of 256 heave samples are collected. In total 8 sets of 256 translational data samples are processed per half-hour to produce a wave / direction spectrum. A Fast Fourier Transform (FFT) is applied to obtain the spectrum in the frequency range 0 to 0.64 Hz, with 0.005Hz resolution. For the wave directional spectrum, the north, west and vertical displacements are included in the processing to give the wave direction, the directional spread, the wave ellipticity and the power spectral density. The Datawell real-time format refers to hexadecimal vectors with vertical, north and west displacements which, compiled, generate a spectrum. The resulting hexadecimal (*.hxv) files contain real-time displacements, full wave spectrum data, a system status data and timings. For each half an hour record, eight spectra with a 200s interval are collected and averaged at the end of the 30 minutes, thus providing a usable smoothed estimate of the spectrum. The transmission of each 30 minute record from the buoy takes 255s. During the following half an hour the file is transmitted 8 times for redundancy.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 13 of 39 The *.hxv files are processed using the Datawell W@ves21 v2.1.18 software (see Figure 2) where raw displacements (*.raw), spectral data (*.spt), waves statistics (*.wvs), spectrum history (*.his), statistics history (*.his) and 3D spectrum (*.dat) files are produced. The resulting files can be used for further processing using customized software routines.
Figure 2 Screenshot of Datawell W@ves21 software
The Danish Wave Energy Center (DanWEC), on the other hand, has a non-directional Datawell directional Waverider MkII buoy deployed in their test site (located off Hanstholm, in the northwest coast of Denmark). The buoy provides both time and frequency domain parameters for 30 minutes interval. The files containing these parameters are automatically generated by the software provided with the Datawell buoy. Table 2 provides an overview of the typical time domain parameters obtained from the processing of thirty minutes wave records (wave elevation time series) such as the one presented in Figure 3 (Lavelle et al., 2012). As seen in this figure, the wave height is defined as the crest to trough distance and the wave period as the time elapsed between two successive crests or troughs. In the light of this, the Hmax and Tmax parameters in Table 2 respectively are the maximum wave height and the maximum wave period in the thirty minute sample period. H1/3 and H1/10 are respectively the average heights of the top one third and the top one tenth of the highest individual wave heights (i.e., H1/3 is the significant wave height). T1/3 and T1/10 are respectively the average periods of the top one third and the top one tenth of the highest individual wave heights.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 14 of 39 Field Unit Description Date and time Hmax Tmax H1/10 T1/10 H1/3 T1/3 Hav Tav Eps #Waves
s cm s cm s cm s
Height of the highest wave Period of the highest wave Average height of 10% highest waves Average period of 10% highest waves Average height of 33% highest waves Average period of 33% highest waves Average height of all waves Average period of all waves Bandwidth parameter Number of waves
Table 2 Time domain wave parameters provided a Datawell non-directional buoy Figure 3 Wave elevation time series sample from the DanWEC test site buoy (individual wave shaded in pink).
Frequency domain parameters are computed from the wave spectrum which, in turn, is obtained from the time domain data using Fast Fourier Transform (FFT). Table 3 provides an overview of the typical frequency domain parameters obtained from wave spectra such as the one presented in Figure 4 (Lavelle et al., 2012). The parameters in the table are described in section 2.1.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 15 of 39 Symbol Unit Title Moment definition Hm0 m Significant wave height
2 / 1
4m Te s Mean energy period
1 / m
m− Tm01 s Mean period
1 0 / m
m Tm02 s Mean zero-crossing period
2 / m
m Tp s Peak period
2 / 1 4 2 2
) / 1 ( m m m − Lp m Peak wavelength
W/m Wave power
Figure 4 Example of wave spectrum from the DanWEC test site buoy
Measurement errors can occur due to instrument (e.g. electronics malfunctioning, knocks from floating object) or transmission failures resulting in ‘spikes’ (large erroneous values) or ‘flat spots’ (reading remains constant for a few samples). Consequently, data filtering and quality control (QC) are key procedures and, as a standard approach, should be carried out before any other data processing. Discriminating the erroneous from the accurate data values may be challenging. Errors can be detected by range and rate of change checks on the time history. They also have an effect on the spectrum, resulting in anomalously high values at low and high frequencies. There is a variety of quality control and wave data filtering procedures that can be adopted with satisfying results, depending on user’s customization. Pitt (2009b) provides an overview of several quality control methodologies. In
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 16 of 39 the scope of this report, the methods currently adopted by EMEC and DanWEC are taken to be up-to-date
DanWEC adopts a data filtering method developed by Tucker et al. (2001) which is based on the identification of erroneous values using the inspection of elevations and velocity of the waves (Lavelle et al., 2012). The diagram in Figure 5 top right plots velocity versus elevation. Velocity is given by dt v
i i i 1 −
− = η η ,
[ 24 ]
where
i
η is the elevation of the i-th sample and dt is the sampling interval.
Figure 5 Top left: selection of se surface elevation time series. Top right: scatter plot of elevation versus velocity for the full 30 min time series. Bottom right: histogram of wave elevations. Bottom right: wave velocity and wave elevations transformed to new coordinates using SVD
Real ocean waves have normally distributed values of η and v. The values of the η and v vectors shown in Figure 5 top right are decomposed into orthogonal vectors (i.e. factorized) using Scalar Vector Decomposition (SVD), as shown in Figure 5 bottom right, thus rendering easier the discrimination of erroneous values. The standard deviation
identified as erroneous. The erroneous values in η are replaced as with the interpolation of the two adjacent values. The Shapiro test for normality (Shapiro et al., 1965) is then applied to the x and y coordinates of the new elevation and velocity vectors (with erroneous values corrected). The Shapiro test gives a metric with values close to 1 for
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 17 of 39 normal distributions. With values inferior to 0.95 for the x or the y coordinate, the time series is rejected and the respective parameters are not added to the database. Data files from EMEC’s DWR-MkIII buoys undergo reception quality and screening checks which are automated by the W@ves21 software. In addition to this, EMEC routinely carries out additional quality control on all data sources in order to determine whether the information that has been recorded is a valid representation of the test site conditions, thus ensuring that a data feed that is free from erroneous observations is being received. The in-house quality control routine developed by EMEC uses MATLAB software to assess the data files before further use. The quality control routine requires that the *.hxv files first be analysed in W@ves21 and makes use of the resulting processed file types. There are two steps in the quality control process: visual QC Part1 and visual QC part 2. In visual QC part 1 a MATLAB script is run which creates a user interface (UI). The UI table is populated with parameters from the *.raw, *.spt and *.his 30 minute records (see 6). Parameters include the number of valid records in each file, Hmax and a number of heave statistics. Two plots are generated for the user to consider. The topmost is the heave, north and west displacements of the buoy with a status indication and a guideline box within which the displacement values should be retained. The lower plot is the power spectral density with direction and directional spread from the same 30 minute time period. Users are invited to select a ‘flag’ value dependant on what the displacements records indicate. A number of flag types are generated automatically such as reception errors (flag 7) or periods where the DWR-MkIII was otherwise unavailable (flag -2). A selection of other flag types which users can select are ‘Data OK’ (flag 1), ‘Filter error’ (flag 2) and ‘Spikes’ (flag 6). The software automatically detects when a heave displacement has exceeded the expected boundary and the user is prompted to select the ‘Spikes’ flag. When the user reaches the end of the records there is a save function that records the information recorded in visual QC Part 1 out with the MATLAB UI.
Figure 6 EMEC’s Visual QC Part 1 Screenshot
In visual QC Part 2 a second MATLAB script is initiated creating a UI generating a plot of the significant and maximum wave heights and wave periods with the associated flags that were generated as part of visual QC part 1. The parameters and flags also populate the table below in 7. Users can change the status of the error flag during visual QC part 2 as the record can be compared with the records immediately proceeding and preceding any suspected
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 18 of 39
wave records are useable and which contain potential errors.
Figure 7 EMEC’s visual QC Part 2 showing instances of error flags
As mentioned above, ensuring that measurements are reliable and free of spurious data is a crucial step in wave data analysis. Besides quality control and data filtering, other procedures are typically carried out, in particular concerning the more useful spectral analysis (frequency domain). Pitt (2009b) provides an overview of the steps to be followed, which the author summarizes in the following:
algorithm so that the number of data values in the input time history should be a binary number.
be 0.005 Hz or less, and certainly no greater than 0.01 Hz.
range may be lower if constrained by the instrumental technique, but should not be less than 0.3 Hz.
be minimized. This requires that the section of data which is Fourier transformed is ideally no shorter than 1000 seconds. This is particularly important for pressure and other subsurface records in which contamination of the higher frequency estimates should be minimized.
estimate should be defined with a minimum of 20 degrees of freedom (d.f.) and ideally with 40 d.f.. These recommendations are applicable to the many different principles upon which wave measuring systems are based and not only to wave buoys.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 19 of 39
Archival and presentation are priority areas within the wave data standards context. The definition of the storage requirements of the data, how it can be accessed and the format in which it is archived should be the first step in the efficient preparation for wave data collection. Pitt (2009b) establishes a set of recommendations to be followed in data archival procedures, notably:
although the details may vary. The information arising from an individual sample measurement is referred to as a data record and these shall be ascribed a series of nominal times which will usually be on the hour, one hour apart. Where the data record for a particular nominal time is missing, a specially formatted ‘no-data’ record shall be substituted.
and/or no-data records. This is not a fixed requirement, but monthly files have proved to be a useful length to handle.
The documentation should be sufficiently comprehensive to ensure the intelligibility of the data to all interested parties. This report further includes details on the recommended information content of the data records and, in particular, a suggestive listing of the spectral and non-spectral data quantities that should be recorded. The EquiMar project also provides a few key recommendations on wave data archival (EquiMar, 2010a). In particular, it is recommended that time series data are recorded and archived for validation. The elaboration of periodic summary reports including metadata should be carried out at appropriate intervals. These reports should be produced on a monthly basis in the case of buoy measurements where data are transmitted to shore. This is, in fact, aligned with the report periodicity suggested by Pitt (2009b). In the cases of wave instrumentation requiring in loco data recovery, longer periods between reports should be acceptable. In light of the above, the monthly wave report template developed at EMEC is here suggested as a standard format against which wave data can be presented across different test facilities. As a consequence, key elements of this report template are included and described in the present report. EMEC’s template report is based on wave data collected using the Datawell MIII Directional Waverider buoy deployed at the Billia Croo test site and its scope is derived from the performance assessment methods outlined in Pitt (2009a). It is suggested that, although the reported parameters and categorisation of results are focussed on the parameters found relevant at the Billia Croo test site, the template should be easily extended at sites where other parameters are shown to be of relevance. It is stressed that notwithstanding the methods outlined in the report being in review with respect to the updated performance assessment methods described in International Electrotechnical Commission (2012), the overall pattern of reporting remains the same. The summary analyses of the prevailing wave and meteorological conditions in the test site provided below refer to October 2014. The following summarizes data analysis procedures adopted by EMEC for the elaboration of the monthly report. Data processing and data files All figures and data for the tables have been generated by code written in Matlab by EMEC. Where possible, efforts have been made to automate the report writing process as much as possible, but some measure of manual intervention is required.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 20 of 39 The wave data used in this report are generated from the raw (.hxv) files recorded by the wave buoys using file regeneration in W@ves21 (version 2.1.18) software. Met data is as recorded from the Met Station at Billia Croo. Missing data records None of the data files contain indicators for missing records. Thus, prior to processing, a data structure which assumes a record every 30 minutes for every day of the month is generated, and then explicitly filled record by record according to the date and time for each. In this way missing records are identified, and a missing data record is one anticipated at that time on that day but which is not present, or which is only partially complete. The detection of missing records and the correct time allocation for all records is aggravated in the case of data from the buoys whose transmission may be interrupted resulting in the drifting of the time date stamps. Special routines have been incorporated to handle this and provide additional safeguards. As a matter of protocol all tabular outputs explicitly account for the number of records, and these plus the number of missing records must equal the number
30 minutes records While the 30 minute record from the buoy data files is a summary of the time series captured during the preceding half-hour, for the met station (with the exception of rainfall) it contains values captured at that instant. Most processing of this data is simple data summarising, finding the maxima, minima and mean values for the day and month for the parameter concerned from the 30 minute records. Missing records are excluded from these, but only prevent their daily summary presentation where a whole day’s data is missing. Excluded records In many of the monthly data reports there is no need to exclude any of the 30 minute records. However, in certain months (for example when maintenance work on the buoys is undertaken, while still recording) the resulting record will contain anomalous data. These data are removed and treated as missing records, with the monthly narrative report describing the timing and nature of any maintenance undertaken. Additionally where EMEC considers a record to be an anomalous outlier (or simply unreliable) it will require that it is excluded from the analysis. While the code also treats these as missing data records, the monthly narrative report will identify any excluded records and provide an explanation for their exclusion by EMEC. Key elements of the template monthly report developed by EMEC are presented below.
Summary statistics
Table 4 below provides a summary of monthly mean wave statistics.
Table 4 Summary of monthly wave statistics
Table 5 provides a summary of the monthly minimum, maximum and mean values for the characteristics of wave- height (Hs), wave period (Tz) and peak direction, together with the derived estimates of wave power. The values Month Year Tz (s) Hm0 (m) Peak direction (o) Spread (o) Mean power (kW/m)
points buoy E October 2014 8.47 2.17 299.1 26.7 34.73 48
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 21 of 39 contained within each of the rows are selected as discrete values for each characteristic and do not necessarily relate to the same half-hour record. Characteristic Wave height [Hm0 (m)] Wave period [Tz (s)] Direction (o) Wave Power Estimate (Daily mean – kW/m) Minimum 0.37 2.77 141.8 0.42 Maximum 7.57 10.01 357.2 332.50 Mean 2.14 6.00 299.1 34.73
Table 5 Summary of mean, min and max wave characteristics
When available, meteorological data may be of use. The monthly report must state: During [Month] there was a meteorological data capture rate of [perc.]. Table 6 provides a summary of the monthly mean meteorological statistics, together with the sea surface temperature, sourced from the Waverider buoy data. Month Year Air Temp. (oC) Wind velocity (m/s) Pressure (mbar)
points Sea surface temp (oC) October 2014 10.8 6.75 1001 4464 12.38
Table 6 Summary of monthly mean meteorological statistics
Table 7 provides a summary of the monthly minimum, maximum and mean values for the meteorological statistics. Characteristic Air temp. (oC) Wind velocity (m/s) Pressure (mbar) Sea surface temperature (oC) Minimum 4.9 0.47 975 11.6 Maximum 17.4 19.71 1021 13.05 Mean 10.8 6.75 1001 12.38
Table 7 Summary of mean, max and min meteorological statistics
Table 8 contains basic summary information from the wave buoy data, giving daily minima, maxima and mean values, and a monthly summary of these. Summarised data for Hs, Tz, Te, peak direction and spread are taken directly from regenerated files provided by the buoy software for the month, or calculated from these. The number
coverage is calculated as a percentage of the actual records to the expected records for the month. Table 9 contains basic summary information from the met station data, giving daily minima, maxima and mean values, and a monthly summary of these. Summarised data for air temperature, rainfall, wind velocity and pressure are taken directly from the logged data. The summary column of average surface sea temperature is taken from the history file for the buoy data processed during that month.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 22 of 39
Table 8 Monthly summary of basic wave statistics
Day min max mean min max mean min max mean min max mean min max mean min max mean 1 2.31 4.91 3.21 6.40 9.22 7.19 9.03 12.08 9.95 271.5 301.2 284.9 15.1 37.9 23.7 7 27.81 165.66 59.34 48 2 3.15 5.09 4.09 6.90 9.56 8.47 10.41 13.50 11.78 272.6 305.9 286.6 16.0 33.5 22.7 7 59.71 169.44 112.54 48 3 2.57 4.63 3.32 6.24 8.38 7.34 8.73 11.32 10.12 276.7 305.8 284.9 15.5 36.4 22.9 7 32.99 131.52 63.83 48 4 1.89 3.80 2.73 6.27 9.49 7.69 9.23 11.63 10.34 270.4 298.4 281.7 15.6 32.4 22.0 7 19.70 95.53 44.17 48 5 1.79 3.39 2.46 6.06 7.90 6.99 8.60 11.45 9.83 270.1 299.4 286.4 15.5 33.5 24.3 7 15.30 67.44 33.95 48 6 0.87 2.14 1.50 3.56 7.07 4.84 6.26 11.32 8.44 265.2 295.8 281.8 15.7 52.0 30.6 7 4.80 27.20 11.17 48 7 0.91 1.80 1.22 3.40 6.69 4.59 6.08 11.06 8.27 262.6 298.0 280.2 14.3 50.2 29.1 7 3.32 14.55 7.08 48 8 0.71 1.08 0.86 3.38 6.35 4.91 6.23 7.99 7.05 270.4 344.6 314.4 21.6 47.7 33.7 7 1.91 4.81 2.77 48 9 0.87 1.78 1.10 4.17 6.36 5.37 6.03 8.34 7.20 307.8 349.7 334.9 15.4 45.5 28.4 7 2.78 10.73 4.59 48 10 1.18 1.99 1.50 4.38 7.20 5.53 5.87 9.28 7.43 326.3 354.3 341.2 14.1 43.2 27.9 7 4.63 18.62 9.03 48 11 0.98 2.09 1.43 4.60 6.79 5.46 6.10 8.61 7.06 314.1 357.2 339.0 20.3 39.3 28.4 7 3.34 19.84 7.84 48 12 0.75 1.64 1.16 4.03 5.25 4.61 5.25 6.66 5.83 319.8 356.5 336.9 16.9 43.8 29.3 7 1.52 8.99 4.20 48 13 0.68 0.89 0.78 4.18 5.19 4.72 5.36 6.41 5.94 316.9 351.3 336.6 18.0 47.0 28.2 7 1.35 2.49 1.79 48 14 0.52 0.96 0.68 4.10 5.46 4.78 5.34 6.75 6.09 318.9 352.3 337.6 20.6 41.1 29.3 7 0.87 2.86 1.43 48 15 0.37 1.16 0.60 3.73 10.01 5.76 6.09 15.43 9.48 272.4 353.5 314.4 15.8 46.9 30.2 7 0.42 11.75 2.52 48 16 0.57 1.16 0.81 2.97 8.27 4.71 5.89 13.78 9.85 272.5 305.0 285.6 18.7 49.1 30.4 7 1.10 10.29 4.16 48 17 0.53 0.77 0.65 2.78 3.77 3.12 4.82 8.17 6.13 261.5 300.0 283.4 26.1 57.3 38.6 7 0.81 2.00 1.40 48 18 0.43 1.14 0.78 2.91 4.84 3.46 3.85 8.78 5.83 141.8 309.8 248.8 14.1 51.6 33.1 7 0.72 6.05 1.91 48 19 1.00 4.73 2.15 3.99 7.03 5.07 6.28 8.39 7.26 262.8 300.0 277.2 14.9 44.1 28.3 7 4.38 91.38 21.32 48 20 1.53 4.29 2.76 5.28 6.96 6.08 6.86 8.77 7.73 268.6 309.1 287.0 13.1 32.3 21.1 7 9.21 77.93 33.82 48 21 1.53 7.69 4.30 4.98 8.73 6.90 6.79 10.44 8.65 276.1 319.8 293.6 15.1 39.5 24.3 7 8.75 342.27 110.21 47 22 2.43 4.52 3.65 6.66 7.79 7.29 8.79 10.34 9.44 282.4 318.5 295.9 15.7 35.0 24.1 7 32.92 108.65 69.70 48 23 2.42 4.39 3.58 6.86 8.53 7.44 9.03 10.54 9.63 264.2 305.9 283.8 13.4 33.4 23.7 7 31.26 95.33 68.46 48 24 2.86 4.44 3.60 6.78 7.89 7.27 9.00 10.18 9.64 271.9 299.0 282.1 15.5 30.9 21.9 7 43.99 103.96 68.99 48 25 2.65 5.63 3.77 6.33 7.98 7.19 8.42 10.44 9.44 268.9 292.4 281.3 13.6 35.5 21.1 7 39.03 165.40 76.97 47 26 4.03 6.42 5.19 7.22 8.83 8.15 9.25 11.68 10.63 266.5 294.6 281.4 14.6 32.5 22.2 7 97.95 256.48 161.43 48 27 2.17 4.48 3.08 7.02 8.49 7.75 8.95 10.88 10.05 271.1 308.2 284.5 14.6 26.7 21.0 7 26.62 111.31 54.39 48 28 1.73 3.38 2.38 5.33 7.39 6.14 6.93 9.90 8.18 272.5 314.7 287.1 16.0 38.2 25.4 7 14.76 46.63 24.66 48 29 1.60 3.53 2.36 5.86 7.39 6.48 7.42 9.08 8.31 269.3 322.9 296.5 16.4 39.0 28.7 7 11.88 48.05 24.94 48 30 1.06 2.02 1.36 4.34 9.18 6.50 7.58 10.70 9.13 281.8 349.1 331.5 14.3 42.6 25.5 7 4.71 18.45 9.25 48 31 0.91 1.51 1.14 4.21 7.59 5.41 7.07 10.29 8.75 280.1 349.9 329.7 14.8 46.3 28.6 7 3.68 11.60 6.31 48 Monthly Summary Battery Records min 0.37 2.78 3.85 141.8 13.1 7 0.42 47 max 7.69 10.01 15.43 357.2 57.3 7 342.27 48 mean 2.20 6.04 8.50 299.1 26.7 7 35.62 48 Total 1486 1488 99.87% Total possible records Actual coverage Power (kW/m) Number of records Hm0 (m), monthly Tz (s), monthly Te (s), monthly Direction, monthly Spread, monthly Power (kW/m) Hm0 (m), daily Tz (s), daily Te (s), daily Peak direction, daily Peak spread, daily Battery status
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 23 of 39
Table 9 Monthly summary of met statistics
Day min max mean min max mean min max mean min max mean 1 10.9 16.5 12.6 3.01 11.50 8.25 1007 1021 1014 144 12.95 13.05 13.01 2 10.5 14.5 12.8 2.40 9.09 5.96 1003 1021 1016 144 12.90 12.95 12.93 3 10.6 14.7 12.4 1.28 11.52 6.43 999 1007 1004 144 12.85 12.95 12.91 4 8.4 11.8 10.4 0.55 8.22 4.35 1003 1008 1005 144 12.75 12.90 12.85 5 8.5 13.4 11.0 2.87 11.92 6.64 1002 1009 1007 144 12.65 12.85 12.78 6 10.8 12.8 11.8 5.17 16.47 12.26 993 1002 996 144 12.60 12.75 12.70 7 10.0 12.2 11.0 4.30 16.23 11.46 990 998 993 144 12.55 12.70 12.65 8 7.2 12.7 9.9 2.02 11.64 4.43 988 990 989 144 12.55 12.65 12.60 9 8.1 11.7 10.4 1.77 6.60 3.66 987 993 989 144 12.55 12.60 12.56 10 8.0 12.4 10.9 0.58 7.58 4.22 993 1002 998 144 12.50 12.55 12.55 11 7.8 11.4 10.6 1.05 6.56 4.03 1002 1005 1003 144 12.45 12.55 12.51 12 5.8 11.5 9.6 2.06 5.51 3.54 1005 1013 1009 144 12.45 12.55 12.51 13 5.1 10.4 7.7 0.92 3.63 2.41 1013 1015 1014 144 12.40 12.60 12.51 14 6.5 11.2 8.7 1.69 3.89 2.86 1011 1015 1013 144 12.40 12.55 12.48 15 5.1 11.9 8.3 2.60 6.24 4.23 1005 1011 1008 144 12.35 12.55 12.46 16 7.1 12.1 10.1 3.32 9.97 7.33 1002 1005 1003 144 12.20 12.45 12.38 17 10.3 12.3 11.2 6.67 11.08 9.02 999 1003 1002 144 12.15 12.40 12.34 18 12.1 17.4 14.4 1.38 13.20 6.30 990 999 992 144 12.25 12.50 12.39 19 10.5 14.0 12.7 2.98 15.23 9.25 984 995 988 144 12.25 12.40 12.38 20 10.3 12.3 11.4 3.21 15.26 7.71 986 999 996 144 12.20 12.40 12.28 21 7.6 12.8 9.4 4.54 18.28 11.46 975 1007 990 144 12.15 12.35 12.28 22 8.1 12.8 10.7 2.65 11.19 7.02 1000 1010 1006 144 11.95 12.25 12.12 23 10.6 13.2 12.0 2.79 12.27 7.44 996 1001 999 144 11.90 12.20 12.11 24 8.9 11.5 10.2 4.37 12.61 7.46 997 1000 998 144 11.90 12.20 12.08 25 9.1 12.2 10.6 3.89 15.46 8.69 990 997 993 144 11.90 12.15 12.01 26 9.4 13.1 11.3 6.38 19.71 11.58 988 999 991 144 11.80 12.15 12.01 27 8.3 13.4 10.6 1.05 12.02 5.33 995 1001 998 144 11.90 12.10 11.96 28 7.5 10.5 9.2 2.03 13.64 7.71 995 1005 999 144 11.85 12.10 11.95 29 4.9 10.6 8.5 1.28 13.38 5.87 1005 1017 1012 144 11.70 12.05 11.85 30 7.3 11.4 10.1 2.65 11.79 6.55 1006 1016 1010 144 11.60 11.95 11.83 31 11.3 15.6 13.1 0.47 9.02 5.87 997 1007 1001 144 11.70 11.95 11.89 Records min 4.9 0.47 975 144 11.60 max 17.4 19.71 1021 144 13.05 mean 10.8 6.75 1001 144 12.38 Total 4464 4464 100.00% Total possible (10 min) met records Actual coverage Daily Temperature (oC) Wind velocity (m/s) Air pressure (mbar) Number of records Sea surface temperature Temperature, monthly Wind velocity, monthly Pressure, monthly Monthly summary Sea temperature
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 24 of 39 Table 10 is generated by code, which constructs a simple 2-way contingency table using Hs and peak direction. Column and row names show the binned ranges, using the notation explained above. All records for each Hs and peak direction range are summed for each cell in the table, so that the sum of rows and columns must equal the total number of records for the month. Any rows or columns that are empty are excluded from the table for ease of reading and printing. The complete data space is represented graphically in Figure 8 at a much finer resolution, and also in a wave rose format in Figure 9. Table 11 is generated by code, which constructs a simple 2-way contingency table using Tz and peak direction. Column and row names show the binned ranges, using the notation explained above. All records for each Tz and peak direction range are summed for each cell in the table, so that the sum of rows and columns must equal the total number of records for the month. Any rows or columns that are empty are excluded from the table for ease of reading and printing. The complete data space is represented graphically in Figure 10 at a much finer resolution. Table 12 is generated by code, which constructs a simple 2-way contingency table using Tz and wave height, Hs. Column and row names show the binned ranges, using the notation explained above. All records for each Tz and Hs range are summed for each cell in the table, so that the sum of rows and columns must equal the total number of records for the month. Any rows or columns that are empty are excluded from the table for ease of reading and
Table 13 is generated by code, which constructs a simple 2-way contingency table using energy period (also referred to as the power period), Te and wave height, Hs. Column and row names show the binned ranges, using the notation explained above. All records for each Te and Hs range are summed for each cell in the table, so that the sum of rows and columns must equal the total number of records for the month. Any rows or columns that are empty are excluded from the table for ease of reading and printing. The complete data space is represented graphically in Figure 12 at the same resolution.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 25 of 39
Table 10 Monthly occurrence table, Hm0 vs Peak direction Figure 8 Distribution of the number of 30 minute records by peak direction and significant wave height
Height [Hm0] in m (140,150] (150,160] (160,170] (170,180] (180,190] (190,200] (200,210] (210,220] (220,230] (230,240] (240,250] (250,260] (260,270] (270,280] (280,290] (290,300] (300,310] (310,320] (320,330] (330,340] (340,350] (350,360] TOTAL (0.0,0.5] 1 1 4 1 1 6 9 1 24 (0.5,1.0] 3 1 2 1 1 1 11 43 59 38 10 10 28 79 55 4 346 (1.0,1.5] 2 3 7 39 40 14 4 8 29 83 64 6 299 (1.5,2.0] 2 17 15 18 15 4 5 15 16 4 111 (2.0,2.5] 2 14 51 41 16 6 1 1 1 133 (2.5,3.0] 1 42 63 17 2 5 130 (3.0,3.5] 5 49 72 26 3 3 158 (3.5,4.0] 7 26 55 20 3 111 (4.0,4.5] 20 33 21 4 1 79 (4.5,5.0] 1 15 19 6 2 1 44 (5.0,5.5] 1 4 11 3 2 21 (5.5,6.0] 1 4 7 1 13 (6.0,6.5] 6 2 2 10 (6.5,7.0] 1 4 1 6 (7.0,7.5] (7.5,8.0] 1 1 TOTAL 5 4 2 1 1 1 39 280 432 213 62 38 64 184 145 15 1486 Peak Direction (degrees)
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 26 of 39
Figure 9 Wave rose
10% 20% 30% W E S N 0 - 0.5 0.5 - 1 1 - 1.5 1.5 - 2 2 - 2.5 2.5 - 3 3 - 3.5 3.5 - 4 >=4
Significant Wave Height (m)
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 27 of 39
Table 11 Monthly occurrence table, Tz vs Peak direction Figure 10 Distribution of the number of 30 minute records by peak direction and period Tz
Period [Tz] in seconds (140,150] (150,160] (160,170] (170,180] (180,190] (190,200] (200,210] (210,220] (220,230] (230,240] (240,250] (250,260] (260,270] (270,280] (280,290] (290,300] (300,310] (310,320] (320,330] (330,340] (340,350] (350,360] TOTAL (0.0,0.5] (0.5,1.0] (1.0,1.5] (1.5,2.0] (2.0,2.5] (2.5,3.0] 1 1 1 3 5 2 6 19 (3.0,3.5] 5 4 1 1 6 15 23 17 3 75 (3.5,4.0] 4 15 23 14 4 1 4 1 66 (4.0,4.5] 3 18 20 10 1 6 13 33 20 124 (4.5,5.0] 5 9 9 4 2 5 20 69 42 5 170 (5.0,5.5] 2 15 18 8 8 5 12 39 33 4 144 (5.5,6.0] 2 19 21 21 10 5 7 20 18 3 126 (6.0,6.5] 9 18 33 18 9 5 9 7 15 2 125 (6.5,7.0] 2 41 62 30 9 5 1 6 6 162 (7.0,7.5] 1 43 100 37 8 6 1 2 198 (7.5,8.0] 2 46 59 24 5 1 2 2 141 (8.0,8.5] 27 41 12 3 5 2 90 (8.5,9.0] 7 16 7 3 33 (9.0,9.5] 1 5 4 1 11 (9.5,10.0] 1 1 (10.0,10.5] 1 1 TOTAL 5 4 2 1 1 1 39 279 432 213 62 38 64 184 145 15 1486 Peak Direction (degrees)
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 28 of 39
Table 12 Monthly occurrence table, Tz vs Hm0
Height [Hm0] in m (2.5,3.0] (3.0,3.5] (3.5,4.0] (4.0,4.5] (4.5,5.0] (5.0,5.5] (5.5,6.0] (6.0,6.5] (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] (10.0,10.5] TOTAL (0,0.5] 3 7 2 4 7 1 24 (0.5,1.0] 18 63 35 57 85 43 20 9 8 4 3 1 346 (1.0,1.5] 7 19 58 66 53 26 26 17 6 7 8 4 1 1 299 (1.5,2.0] 4 15 26 24 12 16 14 111 (2.0,2.5] 2 14 28 32 26 20 7 4 133 (2.5,3.0] 3 17 17 19 35 20 19 130 (3.0,3.5] 5 14 45 49 24 17 4 158 (3.5,4.0] 1 8 14 49 21 11 5 2 111 (4.0,4.5] 2 11 16 29 9 6 5 1 79 (4.5,5.0] 2 8 16 9 6 3 44 (5.0,5.5] 1 13 4 3 21 (5.5,6.0] 1 3 8 1 13 (6.0,6.5] 1 4 5 10 (6.5,7.0] 1 3 2 6 (7.0,7.5] (7.5,8.0] 1 1 TOTAL 18 73 61 121 172 146 122 120 158 203 145 97 37 11 1 1 1486 Period [Tz] in seconds
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 29 of 39
Figure 11 Distribution of the number of 30 minute records by significant wave height and period Tz Table 13 Monthly occurrence table, Te vs Hm0 Figure 12 Distribution of the number of 30 minute records by significant wave height and energy period
Height [Hm0] in m (3.5,4.0] (4.0,4.5] (4.5,5.0] (5.0,5.5] (5.5,6.0] (6.0,6.5] (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] (10.0,10.5] (10.5,11.0] (11.0,11.5] (11.5,12.0] (12.0,12.5] (12.5,13.0] (13.0,13.5] (13.5,14.0] (14.0,14.5] (14.5,15.0] (15.0,15.5] (15.5,16.0] TOTAL (0,0.5] 8 13 1 1 1 24 (0.5,1.0] 1 3 6 23 71 82 51 23 21 10 14 9 3 5 6 3 2 2 2 2 4 2 1 346 (1.0,1.5] 1 3 1 18 40 31 45 41 35 16 15 16 19 4 2 2 3 4 1 1 1 299 (1.5,2.0] 5 13 13 18 19 28 9 2 4 111 (2.0,2.5] 2 8 14 18 12 25 18 14 12 5 5 133 (2.5,3.0] 2 3 3 18 4 2 16 35 27 15 5 130 (3.0,3.5] 1 5 8 7 9 31 39 41 12 5 158 (3.5,4.0] 1 5 5 3 6 27 26 14 8 8 5 1 1 1 111 (4.0,4.5] 6 5 3 15 19 4 5 9 4 3 3 2 1 79 (4.5,5.0] 1 3 2 12 7 5 6 4 2 1 1 44 (5.0,5.5] 1 2 6 5 1 3 3 21 (5.5,6.0] 1 2 1 6 2 1 13 (6.0,6.5] 1 1 3 2 3 10 (6.5,7.0] 1 2 3 6 (7.0,7.5] (7.5,8.0] 1 1 TOTAL 2 6 7 23 89 139 121 109 137 99 105 148 176 146 64 52 20 8 9 9 7 5 3 2 1486 Energy period [Te] in seconds
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 30 of 39 Tables 14 and 15 are 3-way contingency tables generated by code, extending Table 10 by a further directional factor (Mean Wave Direction – MDIR). Each worksheet provides a monthly occurrence table for energy period [Te] against wave height [Hs] and Direction [MDIR]. Column and row names show the binned ranges, using the notation explained above, with the title of each indicating the binned ranges for direction. All records for each Te and Hs range are summed for each cell in the table, so that the sum of rows and columns must equal the total number of records for the month. As the Mean Wave Direction is factored in 15° bins there is in principle a maximum of 24 tables. Two tables were selected to include in the present report: one with very few occurrences within the binned directional range (Table 14) and another with a large number of occurrences (Table 15). Note that, as expected, the number of occurrences for these two directional bins is in agreement with what can be observed in Figure 9. Figure 13 presents a preferred graphical representation of the data contained in Table 13 and Figure 12, showing estimated power contours by both the methods described in equations [12] and [13]. In Figures 14 and 15, the power estimate for each 30 minute record is summed according to the binned range within which its Hs and Te value falls. The summed results are given in kWh/m for each Hs and Te bin combination in Figure 14 and in parts per thousand (ppt) in Figure 15. The highlighted portion of these matrices must correspond with that
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 31 of 39
Table 14 Monthly occurrence table, Te vs Hm0 for mean direction (30°,45°] Table 15 Monthly occurrence table, Te vs Hm0 for mean direction (270°,285°]
Height [Hm0] in m (0,0.5] (0.5,1.0] (1.0,1.5] (1.5,2.0] (2.0,2.5] (2.5,3.0] (3.0,3.5] (3.5,4.0] (4.0,4.5] (4.5,5.0] (5.0,5.5] (5.5,6.0] (6.0,6.5] (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] (10.0,10.5](10.5,11.0](11.0,11.5](11.5,12.0](12.0,12.5](12.5,13.0](13.0,13.5](13.5,14.0] TOTAL (0,0.5] (0.5,1.0] (1.0,1.5] 1 1 (1.5,2.0] (2.0,2.5] (2.5,3.0] (3.0,3.5] (3.5,4.0] (4.0,4.5] (4.5,5.0] (5.0,5.5] (5.5,6.0] (6.0,6.5] (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] TOTAL 1 1 Energy period [Te] in seconds Height [Hm0] in m (0,0.5] (0.5,1.0] (1.0,1.5] (1.5,2.0] (2.0,2.5] (2.5,3.0] (3.0,3.5] (3.5,4.0] (4.0,4.5] (4.5,5.0] (5.0,5.5] (5.5,6.0] (6.0,6.5] (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] (10.0,10.5](10.5,11.0](11.0,11.5](11.5,12.0](12.0,12.5](12.5,13.0](13.0,13.5](13.5,14.0] TOTAL (0,0.5] 1 1 1 3 (0.5,1.0] 1 1 1 1 4 (1.0,1.5] 1 1 1 1 1 4 2 2 1 1 1 16 (1.5,2.0] 1 4 6 1 2 1 2 17 (2.0,2.5] 2 4 3 7 7 5 1 3 2 34 (2.5,3.0] 2 2 2 7 2 13 17 13 8 2 68 (3.0,3.5] 1 2 7 6 8 22 29 25 8 4 112 (3.5,4.0] 1 1 2 3 5 14 23 9 8 5 3 74 (4.0,4.5] 3 3 10 8 3 3 8 2 1 41 (4.5,5.0] 1 2 7 5 4 5 3 2 29 (5.0,5.5] 2 4 3 1 3 3 16 (5.5,6.0] 2 1 5 2 1 11 (6.0,6.5] 1 2 3 6 (6.5,7.0] (7.0,7.5] (7.5,8.0] (8.0,8.5] (8.5,9.0] (9.0,9.5] (9.5,10.0] TOTAL 4 10 14 25 15 29 75 98 69 42 33 12 3 1 1 431 Energy period [Te] in seconds
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 32 of 39
Figure 13 Scatter diagram of significant wave height and energy period
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 33 of 39
Figure 14 Wave energy matrix showing kWh/m by significant wave height and energy period
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 34 of 39
Figure 15 Normalised wave energy matrix in parts per thousand (1 ppt = 25.963 kWh/m) by significant wave height and energy period
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 35 of 39 Figures 16 to 18 show the richness of content in the spectral data files, respectively plotting the spectral energy, the spectral direction (where S(f)>1m2/Hz) and the spectral direction spread (where S(f)>1m2/Hz) at each frequency for all 30 minute records throughout the month. This data in Figure 16 is used to construct the filter applied to Figure 17 and Figure 18.
Figure 16 Time series of wave spectra (m2/Hz) Figure 17 Direction of the wave spectral components where S(f) > 1m2/Hz
Record number in sequence of 30 minute records in October 2014 Frequency (Hz) 200 400 600 800 1000 1200 1400 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55
Spectral energy density (m2/Hz)
10 20 30 40 50 60 70 80 90 100 110 Record number in sequence of 30 minute records in October 2014 Frequency (Hz) 200 400 600 800 1000 1200 1400 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55
Direction (degrees)
50 100 150 200 250 300 350
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 36 of 39
Figure 18 Directional spread by frequency and time where S(f) > 1m2/Hz
Figure 19 provides a simple means of comparison of Mean Wave Direction [MDIR], Peak Direction [Dp] and Best Direction for each 30 minute record during the month.
Figure 19 Comparison of mean wave direction (green), peak direction (pink) and best direction (blue)
Record number in sequence of 30 minute records in October 2014 Frequency (Hz) 200 400 600 800 1000 1200 1400 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 Direction (degrees) 10 20 30 40 50 60 70 200 400 600 800 1000 1200 1400 50 100 150 200 250 300 350 Record number in sequence of 30min records Direction (degrees)
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 37 of 39
This report intends to provide an up-to-date contribution to the definition of a set of standards to be applied in the handling of wave data, based on the experience and expertise of MaRINET’s consortium and ensemble of world-class wave test facilities. It covered a brief review of the standardization efforts carried out to date, the description of the sort of wave data commonly considered for wave energy purposes, wave measurement and data analysis common and suggested practices, as well as data archival and presentation approaches to be followed. The definition of standards is a dynamic process that follows closely the development of technology, in a constant interchange of information. In other words, standards adapt to technology evolution which, in turn, seeks the
the pre-commercial stage of wave energy technology is a clear obstacle in the definition of a stable and unanimous set of standards. Nevertheless, there is a considerable amount of information on guidelines and standards from
The present work uses the MaRINET experience to provide its contribution towards wave data standardization. It has been seen before (MaRINET (2012a) and MaRINET (2014)) that, despite the variety of test facilities across MaRINET, there is a surprisingly large basis of common ground on which to build consensual and unified approaches and
for technology developers to create the necessary interface between technology performance and available
developer to report efficiency and performance of his device through the commonly used power matrix. Occurrence tables and scatter diagrams should be presented in terms of wave height and period. For wave height, significant wave height is unanimously the parameter to choose. As regards period, some developers seem to prefer peak period, others mean energy period. Preferably, a consensus on this matter should be achieved. While this is not the case, it is important to archive wave data both in (Hs,Tp) and (Hs,Te) binned formats. A number of the tabular and graphical outputs presented in this report involved the creation of contingency tables (the term “Occurrence Table” has been preferred for titles of tables, and “Scatter Diagram” for the Figure). In generating binned ranges for these the following protocols have been adopted by EMEC:
use of (a,b] means “greater than a and less than or equal to b” within any range of values. Conversely [a,b) means “equal to or greater than a and less than b”. This notation is growing in popularity, is clear and more elegant than the alternatives.
include certain zero conditions for example of no wind (to include calm days) left breaks are occasionally used to ensure that the total number of records appearing in a table corresponds to the sum of all records
column headings and this is identical to the method for any corresponding figure. The break protocol adopted can be readily changed for any of these if required.
and rows where data is present, i.e. empty columns and rows are excluded. Similarly for the series of 3 way contingency tables only those tables where data are present are included.
which are based on a much greater number of bins, and which show explicitly the entire data range.
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 38 of 39
Alliance for Coastal Technologies, 2007. Wave sensor technologies. 26 p. Battjes, J. A., Van Vledder, G. P., 1984. Verification of Kimura’s Theory for Wave Group Statistics. Proc. ICCE’84, 43, 642-648. Datawell, 2014. Datawell Waverider Reference Manual. EMEC, 2015. Wave Data. Retrieved from EMEC Orkney: http://www.emec.org.uk/facilities/live-data/wave-data/ EquiMar, 2010a. Deliverable 2.7 Protocols for wave and tidal resource assessment. Davey, T., Smith, H., Smith, G., Lawrence, J., Cavaleri, L., Bertotti, L., Prevosto, M., Girard, F., Holmes, B.. European Commission, 2010. 35 pp. EquiMar, 2010b. Deliverable 4.2 Data Analysis & Presentation to Quantify Uncertainty. Kofoed, J. P., Pecher, A., Margheritini, L., Holmes, B., McCombes, T., Johnstone, C., Bittencourt, C., Retzler, C., Meyers, L. E.. European Commission, 2010. 25 pp. Goda, Y., 1970. Numerical experiments on wave statistics with spectral simulation. Rep. Port Harbour Res. Inst., Vol. 9, No. 3. International Electrotechnical Commission, 2012. Marine energy – Wave tidal and other water current converters – Part 100: Electricity producing wave energy converters – Power performance assessment. IEC/TS 62600-100, IEC [Geneva]. Kimura, A., 1980. Statistical Properties of Random Wave Groups. Proc. 17th ICCE, Sydney, 2955-2973. Lavelle, J., Kofoed, J. P., 2012. Providing
wave data for the DanWEC test site (DanWEC Vaekstforum 2011). DCE Contract Report No. 138, Civil Engineering Department, Aalborg University. Denmark. Longuet Higgins, M.S., Cartwright, D.E., Smith, N.D., 1963. Observations of the directional spectrum of sea waves using the motions of a floating buoy. Ocean Wave Spectra, Prentice-Hall, pp. 111-136. Longuet-Higgins, M. S., 1984. Statistical Properties of Wave Groups in a Random Sea State. Phil. Trans. R. Soc. Lond. A, vol. 312, 219-250. MaRINET, 2012a. Deliverable D2.14 Wave data presentation and storage review. Magagna, D., Conley, D., Proenca, B., Greaves, D., Margheritini, L., Finn, M., Lawrence, J., Lopez, J., Enciso, Y., Holmes, B., Paul, M., MacNulty, H., Smith, H., Ashton, I., Lopes, M., Candido, J., Davey, T., Bryden, I.. European Commission, 2012. 42 pp. MaRINET, 2012b. Deliverable D2.1 Wave Instrumentation Database. Lawrence, J., Holmes, B., Bryden, I., Magagna, D., Enciso, Y., Rousset, J., Smith, H., Paul, M., Margheritini, L., Candido, J.. European Commission, 2012. 55 pp. MaRINET, 2014. Deliverable D2.22 Standards for Wave Data Analysis, Archival and Presentation. Cândido, J., Holmes, B., Tetu, A., Bourdier, S., Johanning, L., Dampney, K.. European Commission, 2014. 11 pp. Mollison, D. (1986). Wave climate and wave power resource. pp 133-156 in (D.Evans and A.F. de O. Falcao eds) Hydrodynamics of Ocean Wave-Energy Utilisation, IUTAM Symposium, Lisbon 1985. Springer, Berlin. Pitt, E. G. (2005). Estimating Power from Wave Measurement at the European Marine Energy Centre (EMEC) Test
D2.9 Standards for Wave Data Analysis, Archival and Presentation
Page 39 of 39 Pitt, E., 2009a. Assessment of Performance of Wave Energy Conversion Systems - Marine Renewable Energy Guides. EMEC report. 28 pp. Pitt, E., 2009b. Assessment of Wave Energy Resource - Marine Renewable Energy Guides. EMEC report. 36 pp. Saulnier J.B, Pontes M.T., 2006. Guidelines for Wave Energy Resource Assessment and Standard Wave Climate. WAVETRAIN Deliverable nº 4-7a. Shapiro, S. S., Wilk, M. B., 1965. An Analysis of Variance Test for Normality (Complete Samples). Biometrika, Vol. 52. Strong, B., Brumley, B., 2003. The Application of the Doppler Shifted Dispersion Relationship to Hurricane Wave Data from an ADCP Directional Wave Gauge and Co-Located Pressure Sensor. Proceedings of IEEE 7th Working Conference
Tietje, D., Folley, M., Beatty, S., Raftery, M., Ricci, P., 2011. Standardised Performance Assessment of Wave Energy
Southampton, UK. Tucker, M. J., 1993. Recommended standard for wave data sampling and near-real-time processing. Ocean Engineering, 20(5), 459–474. doi:10.1016/0029-8018(93)90015-A. Tucker, M. J., Pitt, E. G., 2001. Waves in ocean engineering. Oxford: Elsevier Science. 0 08 043566.