Case study feasibility analysis of the Pelamis wave energy convertor in Ireland, Portugal and North America

Renewable Energy 35 (2010) 443–455

Contents lists available at ScienceDirect

Renewable Energy journal homepage: www.elsevier.com/locate/renene

Case study feasibility analysis of the Pelamis wave energy convertor in Ireland, Portugal and North America G.J. Dalton*, R. Alcorn, T. Lewis Hydraulics and Maritime Research Centre (HMRC), University College Cork, Youngline Industrial Estate, Pouladuff Road, Cork, Ireland

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 February 2009 Accepted 1 July 2009 Available online 5 August 2009

The performance and economic viability of the Pelamis wave energy converter (WEC) has been investigated over a 20 year project time period using 2007 wave energy data from various global locations: Ireland, Portugal, USA and Canada. Previous reports assessing the Pelamis quote a disparate range of financial returns for the Pelamis, necessitating a comparative standardised assessment of wave energy economic indicators. An Excel model (NAVITAS) was created for this purpose which estimated the annual energy output of Pelamis for each location using wave height (Hs) and period (Tz) data, and produced financial results dependant on various input parameters. The economic indicators used for the analysis were cost of electricity (COE), net present value (NPV) and internal rate of return (IRR), modelled at a tariff rate of V0.20/kWh). Analysis of the wave energy data showed that the highest annual energy output (AEO) and capacity for the Pelamis was the Irish site, as expected. Portugal returned lower AOE similar to the lesser North American sites. Monthly energy output was highest in the winter, and was particularly evident in the Irish location. Moreover, the difference between the winter wave energy input and the Pelamis energy output for Ireland was also significant as indicated by the capture width, suggesting that Pelamis design was not efficiently capturing all the wave energy states present during that period. Modelling of COE for the various case study locations showed large variation in returns, depending on the number of WEC modelled and the initial cost input and learning curve. COE was highest when modelling single WEC in comparison to multiples, as well as when using 2004 initial costs in comparison to 2008 costs (at which time price of materials peaked). Ireland returned the lowest COE of V0.05/kWh modelling over 100 WEC at 2004 cost of materials, and V0.15/kWh at 2008 prices. Although favourable COE were recorded from some of the modelled scenarios, results indicated that NPV and IRR were not encouraging when using a V0.20/kWh tariff. It is recommended that a tariff rate of V0.30/kWh be considered for Ireland, and higher rates for other locations. In conclusion, Ireland had the most abundant wave energy output from the Pelamis. COE returns for Ireland were competitive for large number of WEC, even at peak costs, but it is recommended that careful analysis of NPV and IRR should be carried out for full economic assessment. Finally, a standardised method of COE reporting is recommended, using fixed WEC number or MW size, as well as standardised learning/production curves and initial costs, to facilitate confidence in investment decisions based on COE. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Pelamis wave energy converter Hs and Tz Annual energy output Cost of electricity Feed-in tariff Learning/production curve

1. Introduction This article is an economic feasibility analysis study of the Pelamis device, using wave energy data collected from 6 case study locations in Europe and North America. The study stems from the increasing need for alternative energy sources as Europe faces a renewable energy target of 20% target by 2020 [1], and some countries such as Ireland setting even higher targets of 40% for the

* Corresponding author. Tel./fax: þ353 21 4250028. E-mail address: [email protected] (G.J. Dalton). 0960-1481/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2009.07.003

same time period [2]. Moreover, there is rising demand by both public [3] and state for increased use of renewables in energy supply. A combination of all renewable sources, including wave energy, will be required to meet those targets. Wave energy is still in the nascent stage [4], requiring substantial subsidies and support for research and development to bring the technology to the commercial stage [5]. However, there is a lack of confidence by business and investors that renewable energy is economically feasible for commercial energy supply [6]. The cost of electricity (COE) is the benchmark criterion by which most renewable energy (RE) projects are judged [7]. COE refers to electricity cost where there is zero revenue or tariff from an electricity

444

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

Average €/ ton, gross No. 1 heavy melt steel Glossary Annual energy output Cost of electricity Capital recovery factor Greenhouse gases Significant average wave height, mean period or zero crossing IC Initial cost IRR Internal rate of return NPV Net present value O/M Operation and maintenance RC Replacement cost RE Renewable energy REC/ROC Renewable energy credits/renewable obligation certificates FIT/REFIT Feed-in tariff/renewable energy feed-in tariff RES Renewable energy system Technology factor Tf USA EC USA east coast USA WC USA west coast WEC Wave energy converter WEI Wave energy input utility company. However, COE can also be interpreted as the electricity tariff rate returning a zero net present value (NPV). This arbitrary method of defining COE is not ideal, and other benchmark financial criterion or indicators such as NPV and internal rate of return (IRR) are preferred to compare project viability [8]. Nevertheless, due to its prevalent use COE will be the major indicator of this report. There have been many studies which have provided COE figures for wave energy projects. However at present, it is difficult to compare COE results from reports and studies due to the large variation in quoted figures. It is essential that correct COE are forecast so that investors can budget for long-term projects, as well as policy makers who are nominating feed-in tariff rates intending to support and accelerate growth in wave energy industry. COE is directly related to the quantity of energy input to the system, and thus is very dependant on location of the energy source. This paper will present wave energy data collected from various global locations and assess the corresponding COE resultant. As with all analysis, there are many variables which can impact on the final COE result. This report will conduct sensitivity on the following variables and assess their impact on COE.

2. Impact of the cost of materials and initial cost (IC) Steel is currently the main material constituent of a WEC, and thus has the largest influence on initial cost (IC). Steel has had major price fluctuations over the past few years (Fig. 1). Recent factors influencing steel prices fluctuation were increasing demand from China for raw materials, which led to a price escalation [9], followed by the credit crunch and global recession in mid 2008, causing steel prices to eventually fall to 2007 prices [10]. The final cost of manufactured steel, typically grade 50 (S355), painted with corrosive protection, can cost anywhere from V5000–7000/ton,1 and this price had not substantially fallen at the time of writing this paper, although is forecast to do so in 2009. US currency conversion to Euro used in this report was 1.57 (July 2008),2 which was the

1 2

Personal communication Paul Collins, Malacky Walsh Engineers, Cork. http://finance.yahoo.com/q/bc?s¼EURUSD¼X&t¼5y&l¼on&z¼m&q¼l&c¼

350 300 250

€/ton

AOE COE CRF GHG Hs, Tz

400

200 150 100 50 0 2001 2002 2003 2004 2005 2006 2007 2008 2009

Fig. 1. Historical heavy melt steel price, American metal market, 2001-2007 (www. scrapmetalpricesandauctions.com/iron-steel).

peak recorded rate over a five year period. As a result of using this rate, costs in Euro will make final COE results in this article appear lower that they would at a lower exchange rate. Various initial costs (IC) will be modelled and their impact on COE assessed. 3. Size/number of the WEC farm and dependence on cable size Large scale purchasing of WEC coupled with learning/production curves (defined later) has a moderating influence on IC. Additionally, a collection of WEC can be serviced by one cable, which has the required kV rating and capacity to cater for the load. Thus savings will accrue if the optimal number of WEC is matched to whatever cable size is chosen. 4. Feed-in tariffs There is much debate at present concerning appropriate price support schemes or power purchase arrangement (PPA) to stimulate growth in the wave energy sector. Fixed tariff or feed-in tariff (FIT) rates are gaining consent as the most successful method to stimulate RE development [11,12]. The main proponents of the scheme are Germany and Spain [13], and has resulted in a RE boom in those respective states, almost exclusively in on-shore large wind and PV projects. The Irish and Portuguese governments are the only two states that have promised an FIT or renewable energy feed-in tariff (REFIT) for electricity produced by wave energy. Ireland is promising V0.22/kWh,3 and Portugal a range of tariffs ranging from V0.07–26/kWh [14] (Table 1). Spain, France, Denmark and Germany have also proposed wave energy FITs, but they are modest and not been implemented at the time of this paper. 4.1. Wave energy device – Pelamis The WEC chosen for analysis in this report was the Pelamis, as it is the only WEC to date that has a published and reliable power performance matrix (Table 6). The Pelamis wave energy converter is developed and manufactured by Pelamis Wave Power (PWP) (formerly known as Ocean Power Delivery Ltd), an Edinburghbased company originating from the Wave Power Group at the University of Edinburgh in 1998 [15,16]. The Pelamis is a semisubmerged snake-like device consisting of articulated cylindrical

3 http://www.ndp.ie/viewdoc.asp?Docid¼2034&mn¼newx&nID¼&UserLang¼ EN&CatID¼15&StartDate¼1þJanuaryþ2008

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

445

Table 1 Countries which have feed-in tariff programs. Country

Application and restrictions

Proposed tariff

Ireland Portugal

Applies to all developments Demonstration projects 0–20 MW Pre-commercial 20–100 MW Commercial 20–100 MW Commercial 100–250 MW Commercial 250 þ MW

V0.22/kWh V0.26/kWh V0.16–0.21/kWh V0.10–0.16/kWh V0.08–0.11/kWh V0.075/kWh

sections linked by hinged joints (Fig. 2). The version used in this assessment was 120 m long, 3.5 m in diameter, rated at 750 kW and weighed 700 ton. PWP has raised some £40 m to fund the development of Pelamis technology from a variety of financial and industry backers. Major shareholders include; Emerald Technology Ventures, Norsk Hydro Technology Ventures, BlackRock Investment Managers, 3i, Carbon Trust, Nettuno Power, Tudor Global and Scottish Enterprise. PWP have their first major demonstration project in Aguçadoura, Portugal. The multiple Pelamis units making up the Aguçadoura wave farm constitute both the world’s first, multi-unit, wave farm and also the first commercial order for wave energy converters. The Aguçadoura project is owned and operated by a joint venture company called Companhia da Energia Oceaˆnica (CEO) which is currently 77% owned by a subsidiary of Babcock and Brown Limited and 23% by Pelamis Wave Power Limited. Since the financial crisis accelerated in the last quarter of 2008, Babcock and Brown Limited has had its shares suspended and has been in a managed process of selling its assets. Pelamis are in the process of searching for a new partner. The E.ON energy group are the second major developer using the Pelamis. E.ON plans to deploy the latest Pelamis design, the ‘P2’, in the European Marine Energy Centre (EMEC) in 2010. Future plans for the EMEC and the Orcadian Wave Farm will consist of four Pelamis generators supplied by PWP to ScottishPower Renewables supported by £4 m funding from the Scottish Executive providing. Original plans by E.ON to test the Pelamis at Wave Hub site have been shelved as of April 2009 until a future date. Other wave energy device contenders currently on the market include two Irish companies, Ocean Energy Buoy4 and Wavebob.5 Both these designs have completed the quarter stage prototype development stage and are planning the full scale pre-commercialised testing in the near future. Table 2 lists the other wave energy companies that are the late stages of prototype, precommercial testing and commercial phases of development.

4.2. Previous COE wave energy studies using the Pelamis Table 3 lists the major wave energy reports where the Pelamis was assessed and COE was quoted. The largest study was conducted by Previsic [18] in 2004 for EPRI in California, US. Two feasibility analyses were conducted; the first was a pilot test with one device near shore in 25 m of water in the location off San Francisco, and the second was a commercial test, with 213 devices in 50 m of water depth in a location off Hawaii. An adjustment was made for the shallow water site in San Francisco adding a power conversion efficiency factor of 80% and device availability6 at 85%. For the second commercial project in

4

http://www.oceanenergy.ie/ http://www.wavebob.com/ 6 Availability is defined probability of the whole system functioning at any specific time, taking into account maintenance and breakdowns.

Fig. 2. Pelamis prototype [15].

Hawaii, where there was a lower predicted wave intensity, the power output of the device was reduced from 750 kW to 500 kW. However the efficiency was increased to 88% and the availability to 95%. Costs quoted by the EPRI report for the single WECS, mooring and cables were used in this present article. The initial cost of the WECS was $2,415,000 or V1,533,207, at 2004 prices. Full COE analysis was only conducted by EPRI on the large commercial project. COE quoted was $0.11/kWh (approximately V0.08/kWh) for an electricity tariff rate of V0.07/kWh. A further report by EPRI was conducted 2 years later (2006) by Bedard [19], and reported a slightly lower COE, with a range of V0.05–0.08/kWh. The most recent report on Pelamis was conducted in Canada examining a proposed 25GWh (27 Pelamis) wave power plant [20]. The capacity factor for the Pelamis at the locations sampled were approximately 20% and cost of electricity ranged from V0.23–0.38/ kWh, depending on the location. The article mentioned that the output was low in comparison to the EPRI report which used the same device design. It commented that the design was intended for North Atlantic waters, which have different sea states to those of Canada. Another Canadian study looking at 15 Pelamis quoted a COE of between V0.10–0.15/kWh [21]. The major Irish report studying the Pelamis was produced by ESBI [22] for SEI. The COE reported was V0.105/kWh. A second Irish report was undertaken by Bacon [23] in 2005. However the per kWh reading quoted was not for cost of electricity but value of the electricity produced to the country, and this was V0.097/kWh based on a scenario of 50kW/m power and a tariff for 2010 of V0.08–.10/kWh, decreasing to V0.05/kWh by 2020. Two British studies forecasted COE of V0.08/kWh and V0.10/ kWh [24,25] (the latter examined 4000 Pelamis and assumed a V0.05/kWh renewable subsidy). Finally, the Pelamis Company itself has published data of the Pelamis performance, and quotes a COE of V0.08–0.16/kWh [16]. Overall, almost all of the studies reported a COE of between V0.05 and 0.20/kWh with or without a subsidy or feed-in tariff.

4.3. Case study locations

5

Four countries were chosen for this study. Ireland has one of the highest wave energy climates in the world on a kW/meter basis, as

446

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

Table 2 Wave energy companies that are presently at the prototype or pre-commercial phase of development. (Phase of development 1 ¼ initial design, 3 ¼ prototype testing, 4 ¼ precommercial testing, 5 ¼ commercial arrays). (Data obtained from [17]). Country

Company

Device

Type

Ireland Ireland UK UK

Ocean Energy Wavebob Pelamis Wave Power Aquamarine Power AWS Ocean Energy Finavera Fred Olsen Oceanlinx OPT Wave Dragon WavePlane Wavestar

OE Buoy Wavebob Pelamis Oyster Wave Swing AquaBuOY FOBOX3 Oceanlinx PowerBuoy Wave Dragon WavePlane Wavestar

Floating Inertia Inertia Inertia Inertia Inertia Inertia Floating Inertia Floating Floating Inertia

Canada Norway Australia USA Denmark Denmark Denmark

OWC

OWC Overtopping Overtopping

well as a promised FIT of V0.22/kWh.7 Portugal is making the most progress in wave energy deployment, having both the Pilot Zone test area, as well as Pelamis currently doing tests in the country. Furthermore, Portugal has promised an FIT of V0.26/kWh for the first 20 MW of wave energy installed [14]. Finally, the North American continent was chosen, due to its interest in wave energy technologies, as well as comparison of wave climates between east and west coasts (Table 4). 5. Assessment methods The platform for NAVITAS was Microsoft Excel spreadsheet [26], and the results from NAVITAS were validated using a similar model called RETScreen [27]. NAVITAS proceeds through the analysis in two stages: 1. Energy and power calculations. 2. Economic analysis.

5.1. Energy calculation 5.1.1. Scatter plot of hours Unprocessed wave energy data consists of a wave energy spectrum. Significant average wave height (Hs) and period (Tz) parameters are extracted from the data, using the Bretschneider spectrum function which has been found to be the best suited function for deep sea long-fetch locations. The data is processed and averaged in either 10, 30 or 60 min units. NAVITAS requires the data in 60 min increments. A ‘scatter plot’ or ‘diagram’ (also known as a joint probability distribution) of hours was created for each month, which is a table showing the frequency in hours of occurrence of each sea state. Not all combinations of height and period occur (or are even possible) in a real sea state, thus most of the scatter plot cells were empty. A total year’s scatter diagram was produced summing the cell points for each month, totalling 8760 h for the entire year (or 8784 in a leap year). Data for the locations was obtained from various wave energy centres and websites, listed in Table 5. Data from USA and Canadian websites was publically available from their websites. The Canadian data was in excellent condition with no modification needed. The NOA data had some blanks recorded (0.05%), but were minor and easily corrected. The Irish data was obtained from HMRC, which collected the data from the M buoys, owned by the Marine Institute.8 The data was available for a fee. The quality of the data was in

Phase

Scale

Test rating

3 3 4–5 3–4 3 3 3 3 3 3 3–4 3

01:04 01:04 01:01 01:01 01:01.8 01:02 1.3 01:03 01:01.5 01:05.2 1:1–2 01:10

15 kW 15 kW 750 kW 500 kW 250 kW 25 kW 50 kW 45 kW 40 kW 20 kW 250 kW 5.5 kW

moderate condition with 1.5% of the data blank, which was readily correctable. The Portuguese data was also private and available for a fee. The data was recorded in 10 min durations. The data was presented in unprocessed form, where unrecorded data was not presented, i.e. unrecorded data was not presented as a blank. Processing of this data required a Matlab [28] program to average the data in hourly increments.9 Blanks that were larger than 1 h and up to 5 h in duration were interpolated. The month of February had 10 days of missing data. The gaps were filled by using data from 5 days prior and 5 days after the gap. 5.1.2. Power matrix and energy matrix The total annual energy output (AEO) for the year was calculated in NAVITAS by multiplying each cell point of the scatter plot of hours with the corresponding cell of a WEC power matrix. For this report, the power matrix of the Pelamis was used [15]. A capacity factor10 (or sometimes called the load factor or utility factor [18]) is incorporated in the power matrix. The Pelamis power matrix is presented in Table 6. Power peaks at 750 kW for a number of sea states. Wave energy input (WEI) is calculated using the following equation:

WEI ¼ 0:55 H2s Tz

(1)

WEI is a product of the energy input and the ‘apparent’ width of a WEC. Although the width of the Pelamis is 3.5 m, the apparent width used was 8.75 m, due to lateral motion of the WEC and wave attenuation along the WEC.

5.2. Economic inputs 5.2.1. Initial cost The project lifespan of 20 years was used. Initial costs (IC) for the WEC only reflected the purchase of the Pelamis device from the manufacturer, and any sundry costs such as planning and installing of the device. The remainder of other costs (except cable costs) was calculated as a percentage of the IC. This method allowed for simplified cost and sensitivity analysis. The IC of the Pelamis WEC chosen for this report was V1,533,00011 obtained from the 2004 EPRI report in California [18]. The figure included costs for both the steel sections and all the internal components. The weight of the Pelamis was 700 ton, giving V2200/ton. Table 7 lists the cost of the other components which were based as a percentage of the IC.

9

Matlab file for data processing was created by Florent Thiebaut, HMRC, Cork. Capacity factor is the ratio of the mean generation to the peak generation of a WEC. 11 WEC $1,565,000 þ steel sections $850,000. US currency conversion to Euro was 1.57 (at July 2008). 10

7 http://www.ndp.ie/viewdoc.asp?Docid¼2034&mn¼newx&nID¼&UserLang¼ EN&CatID¼15&StartDate¼1þJanuaryþ2008 8 http://www.marine.ie/Home/

Prototype rating 2 MW 2 MW 750 kW 500 kW 2 MW 250 kW 2.5 MW 2 MW 150 kW 7 MW 500 kW 5 MW

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

447

Table 3 Wave energy device studies reporting COE using the Pelamis power matrix. Study

Reference

Location

Year

Number of Pelamis

COE V/kWh

Subsidy

EPRI (Previsic) ESBI St Germain EPRI (Bedard) Carbon Trusta

[18] [22] [21] [19] [25]

California Ireland Canada California UK

2005 2005 2005 2006 2006

0.08 0.105 0.10–0.15 0.05–0.12 0.08–0.30

0.06 – – – –

Allan et al. Dunnet and Wallace Pelamis

[24] [20] [16]

Scotland Canada UK

2008 2008 2008

213 209 15 44 13,000 13 4000 15–27 1

0.10 0.18–0.30 0.08–0.16

– – –

a

Pelamis not used in Carbon Trust.

The table also contains comparison costs to another report, showing similarity in costs. The only component that was not based on Previsic [18] IC was cable cost, which were sourced from an ESBI [29] report. Cable cost for 0.5–1.9 MW capacity equalled the IC of one WEC (Table 8). This is in contrast to the cable costs of Previsic [18] and Allan, Bryden et al. [24], who quoted 50% and 5% of IC respectively (Table 7). Such variation in cable cost estimation will have profound impact on economics, and will need to be carefully assessed in analysis. As the cost of steel approximately tripled from 2004 to 2008, IC for the Pelamis in 2008 was similarly based on a tripling of costs (Table 9). The IC for cabling has also increase in that period but not to the same extent. Thus a doubling of cable IC was used in 2008.

WEC, mooring and cable. Since the lifespan of Pelamis WEC used in this report was 20 yrs (as quoted by Allan [24]), and the project lifetime of the analysis is 20 years, there was no replacement costs scheduled. Other lifetimes quoted are 15 yrs by Previsic [18]. No overhaul of the WEC device and infrastructure was included either, as it was deemed that analysis of both overhaul and replacement impacts on COE would be extensive and the subject matter for a further paper. O/M costs are treated as a capital costs. Percentage O/M expenses were take in the range of 1–3% coinciding with studies of St Germain [21] and Dunnett [20]. Previsic [18] and Allan [24] estimated O/M to be around 40% for a project. Further investigation of high O/M impact on COE will be the subject matter of a further paper.

5.2.2. WEC number A sliding scale was used to estimate the cost of WEC components with increasing number. The sliding scale is based on the learning curve [31], and does not factor mass production into the equation. This is because WEC production is deemed never to reach the scale of automotive production. The formula for the learning curve is as follows:

5.2.4. Salvage value The salvage value is a positive credit to the costs once the project has completed. The value assumes a linear depreciation meaning that the salvage value of a component is directly proportional to its remaining life. Salvage value is calculated by the following formula [32]:

ðlnðtf Þ=lnð2ÞÞ

P ¼ A

(2)

Where P is the percentage scaling, A is the number of WEC components and ‘tf’ is the technology factor. The technology factor of most industrial products is mostly between 0.85 and 0.95. For this project, 0.9 was chosen. It is assumed that there will be ideal grid pattern spacing for multiple WECS and that there is no limitation on laying out converter cordons and cabling in the sea so that the technical resource can be evaluated. It is assumed that an electromechanical conversion efficiency of 85% is already incorporated into the energy matrix of the Pelamis. 5.2.3. Overhaul, replacement and operation and maintenance (O/M) Table 10 lists the percentages of IC used to calculate the costs for overhaul, replacement and operation/maintenance for either the Table 4 Case study buoys and location details. Country

Location

Buoy

Depth (m)

Latitude/Longitude

Ireland Portugal USA - West Coast (WC) USA - East Coast (EC) Canada -West Coast (WC) Canada - East Coast (EC)

Belmullet Leixoes California

M4 Leixoes 46012

150 83 213

55 deg N, 10 deg W 41 deg N, 08 deg W 37 deg N, 122 deg W

Cape Canaveral

41010

872

26 deg N, 78 deg W

Graham Island

C46183

no info

50 deg N, 136 deg W

Nova Scotia

C44137

no info

42 deg N, 62 deg W

S ¼

RC*RL Lt

(3)

Where S is the salvage value, RC is the replacement cost of the component, RL is the years remaining and Lt is the components lifetime. The replacement cost is calculated for each replacement time using the discount factor, and thus inflation is factored out of the equation. If the life of the item is less than the project lifespan, then the remaining life (RL) of the component at the end of the project lifetime is given by:

RL ¼ Py  CY

(4)

Where Py is the project years and CY is the total years up to last replacement of the component, and calculated as follows: Table 5 List of case study locations and sources for data. Location

Source of data

Data availability

Ireland

HMRC, Cork, Ireland.http://www.ucc.ie/ research/hmrc National Oceanic and Atmospheric Association (NOA), http://www.ndbc. noaa.gov/data/historical/stdmet Fisheries and Oceans Canada, http:// www.meds-sdmm.dfo-mpo.gc.ca/ MEDS/Databases/Wave/ArchPlot/ ArchPlot_Form_e.asp?medsid¼C44137 Instituto Hidrografico, Lisbon, Portugal http://www.hidrografico.pt/boiasondografo.php

Fee payablea Moderate

USA

Canada

Portugal

a b

Email [email protected]. Email ana.saramago@hidrografico.pt).

Quality of data

Publically available

Moderate

Publically available

Excellent

Fee payableb Poor

448

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

Table 6 Power matrix scatter plot for the Pelamis WECS [15]. Period (Tz)

Height (Hs)

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12

1

2

3

4

5

6

7

8

9

10

11

12

13

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 32 57 89 129 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 29 65 115 180 260 354 462 544 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 37 83 148 231 332 438 540 642 726 750 750 750 0 0 0 0 0 0 0 0 0 0 0

0 38 86 152 238 332 424 530 628 707 750 750 750 750 750 0 0 0 0 0 0 0 0 0

0 35 78 138 216 292 377 475 562 670 737 750 750 750 750 750 0 0 0 0 0 0 0 0

0 29 65 116 181 240 326 384 473 557 658 711 750 750 750 750 750 0 0 0 0 0 0 0

0 23 53 93 146 210 260 339 382 472 530 619 658 750 750 750 750 750 0 0 0 0 0 0

0 0 42 74 116 167 215 267 338 369 446 512 579 613 686 750 750 750 750 0 0 0 0 0

0 0 33 59 92 132 180 213 266 328 355 415 481 525 593 625 750 750 750 750 0 0 0 0

  Py CY ¼ Lt*trunc Lt

(5)

If lifetime of the item is greater than the project lifespan:

RL ¼ Lt  Py

(6)

Any decommissioning costs are subtracted from the final salvage value. Salvage value can be a substantial especially if the replacement of an item is due close to the termination of a project. 5.2.5. Grid sales revenue Feed-in tariff refers to the regulatory, minimum guaranteed price per kWh that an electricity utility has to pay to a private, independent producer of renewable power fed into the grid [33]. It is defined in this report as the full price per kWh received by an independent producer of renewable energy including the premium above or additional to the market price, but excluding tax rebates or other production subsidies paid by the government. Grid sales are a credit, and are added to other negative cost values for the each year. The sales are the product of the following two variables:  The total energy produced each year (referred to as the annual energy output (AEO)). Table 7 Costs of WEC infrastructure, calculated as a percentage of the WEC IC. The majority of Previsic [18] costings were used in the report, except cabling. WEC parameter

% of IC of WEC Previsic [18]

Mooring Cabling Replacement costs Spare parts Sitting and permits GHG investigations Management fees Decommissioning fees Grid connection

Allen, Bryden et al. [24]

10% 10 þ 20% 50% (not used in this report) 5% 90% 2% 2% 0.05% 10% 5% 10% 5% of AEO 1%

 The electricity tariff rate from the utility company. In this report, the following rates were assessed: V0.00, V0.05, V0.10, V0.20, V0.30, V0.40. 5.2.6. Discount factor The discount factor (DF) translates expected financial benefits or costs in any given future year into present value terms. The total nominal profit is adjusted for cash depreciation by multiplying the total nominal profit by a discount factor. DF is calculated using the discount rate12 (DR) in the following equation:

DF ¼

1 ð1 þ DRÞn

(7)

Where n is the number or years. The discount rate is used to convert between one-time costs and annualized costs and is calculated using the following equation:

DR ¼

BR þ f 1f

(8)

Where BR is the borrowing rate given for a loan and f is the inflation rate. By defining the discount rate in this way, inflation is factored out of the economic analysis. All costs therefore become real costs, meaning that they are in defined in terms of constant Euros. The assumption is that the rate of inflation is the same for all costs. A general inflation rate of 5% [32] was used for the project and a borrowing rate of 10%,13 giving a project discount rate of 4.76%. The discount rate can vary up to 12%.

5.3. Economic indicators 5.3.1. Net present value (NPV) The net present value is defined as the present value of investments future net cash flows minus the initial investment [34]. It is

12 The discount rate is an interest rate commensurate with perceived risk used to convert future payments or receipts to present value. 13 Borrowing rate used was at peak 2008 rate.

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455 Table 8 WEC power output, corresponding cable kV and relevant cost V/km [30].

449

Table 9 IC for 2004 and 2008 for WEC and cabling.

Range MW

kV

Cost per km

Number of kms

Cost

Year

IC for 1 Pelamis WEC

IC for Cable for 1 WEC

0.5–1.9 MW 2–5 MW 5–19 MW Over 20 MW

10 10 30 110

V433,333 V433,333 V1,300,000 V3,900,000

4 4 4 4

V1,733,333 V1,733,333 V5,200,000 V15,600,000

2004 Mid 2008

V1,533,000 V4,599,000

V1,733,333 V3,466,666

derived by summing the discounted cash flows over the project lifetime, defined in the following equation:

NPV ¼

X

TNC*DF

(9)

Where TNC is the total net cash for that year and DF is the discount factor, which is defined as:

DF ¼

Where ‘r’ is the discount rate and ‘n’ is the number of years. The discount rate is the reward an investor demands for accepting a delayed payment. In general, only positive NPV are considered viable commercial projects. NPV is a simple mathematical concept that doesn’t include any arbitrary variables. The total annualized cost (TAC) is the sum of the annualized costs of each system component. It is calculated by multiplying the NPV and the capital recovery factor (CRF). CRF is a ratio used to calculate the present value of an annuity (a series of equal annual cash flows). The equation for the capital recovery factor is:

1ð1 þ rÞn 1ð1 þ rÞ  1

(10)

5.3.2. Cost of electricity (COE) The (levelised) cost of electricity (COE) is defined as the average cost per kWh of useful electrical energy produced by the system, and is calculated by dividing the total annualized cost (TAC) of producing electricity by the annual electric output (AEO). The equation for the COE is as follows:

COE ¼

TAC AEO

(11)

5.3.3. Internal rate of return (IRR) Internal rate of return indicates the business return according to alternative return that may be gained on the same investment. The internal rate of return is the discount rate that will create a zero net present value. The IRR is based on the NPV formula (Equation (9)), and is solved iteratively for when NPV ¼ 0.

NPVð0Þ ¼ IC þ

X

cðnÞ ð1 þ IRRÞn

(12)

Where ‘c’ is the cost for year ‘n’ and NPV(0) is the NPV value equal to zero. In this report, an IRR threshold of 10% was required for a project to be considered financially viable.14,15

14

TNC ¼ NP þ interest  fixed annual repayment

(13)

NP is the net profit and is defined as:

NP ¼ GS þ S  IC  RC  Oh  O=M  interest  tax

(14)

Where GS is the grid sales, S is the salvage value, RC is the replacement cost, Oh is the overhaul cost and O/M is the operating and maintenance cost.

1 ð1 þ rÞn

CRF ¼

5.3.4. Total net cash (TNC) Total net cash (TNC) is calculated as follows:

Personal communication Tony Dalton, chief financial accountant for LET Systems, Cork. 15 IRR requirement for investors is a different scenario (not applicable here). Here, investors judge IRR on their initial investment made and the exit price of the company. The investment is more high risk and thus IRR would be expected to be higher, at around 30%.

5.3.5. Equity/debt, tax calculation and depreciation The project assumed 100% equity financing, with no debt. Assessment of the implications of a project funded with a certain percentage of borrowings requiring tax and depreciation analysis would be outside the scope of this paper. 6. Data analysis and results 6.1. Number of hours per sea state Results displayed in Table 11 show that the wave energy period (Tz) recording the highest frequency of hours was 5–7 s in all locations. The wave height (Hs) of approximately 1–2.5 m was most common for all locations, except Ireland, where the highest incidence of wave height occurred between 1–3.5 m. Wave heights in Portugal recorded in 2007 were measurably smaller than the other locations. 6.2. Annual energy output The AEO of the Pelamis at the Irish location had the highest energy return of the global locations selected for comparison in 2007, where the total AEO produced by the Pelamis WEC device was 2.5 GWh (Fig. 3). The Irish energy output figures were approximately 40% higher than the Californian and Nova Scotia site outputs, double the Portuguese output and approximately three times of the remaining sites. The output at the USA WC site was similar to the output quoted in Previsic’s [18] Californian report. The capacity of the Pelamis decreased for each location in proportion to the drop in AOE, with the capacity at the Irish site reaching 38% and dropping to a low of 10% in USA EC. The capacity results indicate that the Pelamis operated closer to its optimal rating in the Irish location, and that perhaps smaller rated devices would be more suitable for locations such as Canada WC and USA EC. The capture width on the other hand did not have much variation, implying that the Pelamis was performing equally in all locations.

Table 10 Operating costs for WEC, mooring and cable. Variable

Calculation method

Overhaul/refurbishment Replacement O/M

Nil Nil 1% or 2% or 3% of IC

450

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

Table 11 Surface plot of hours for some of the case study locations. Period (Tz) 1 Heights (Hs)

Heights (Hs)

Heights (Hs)

2

3

Ireland 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 USA WC 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 Canada EC 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5

7 96 40 2

4

5

6

7

8

9

10

11

12

17 148 75 6

39 312 560 427 99 10

13 93 372 623 508 248 86 14

0 32 187 321 363 394 436 213 83 17

0 13 64 172 182 193 244 246 218 175 88 30

0 0 11 47 67 83 74 97 124 118 122 86 45 24 9 2

0 12

17 271 223 70 5

15 198 516 550 319 77 7

4 224 516 561 445 360 173 86 7

1 222 457 390 287 216 154 86 57 32 25 3

27 176 140 57 17 3

35 498 666 429 249 89 26 5 2

66 591 422 399 435 240 119 67 42 15 5 1

49 308 284 258 256 179 355 146 111 71 43 15 4 2

6 113 83 83 55 68 80 79 80 65 67 50 25 15 9 1 1

13

4 1 2 8 10 18 15 23 22 26 36 37 32 48 30 20 10 5 2 1

0 0 0 0 0 2 7 7 4 3 2 3 4 15 10 16 9 18 8 4 1

0 0 0 0 0 0 1 8 3 5 1 3 4 4 2 4 7 1 2 0 0

0 0 0 0 0 0 1 0 2 3 0 0 1 0 0 0 0 2 0 1 0

0 54 227 279 260 157 107 69 28 21 18 14 5 3

0 19 70 93 97 110 42 23 16 4 0 0 0 2 1 1

0 1 21 84 51 55 34 25 2 3 7 4 2 0 1 2 1

0 0 4 8 10 5 6 14 4 1 0 1 1 1 0 0 0

0 0 3 2 1 0 0 0 0 0 0 0 0 0 0 0 0

18 85 19 16 23 15 20 13 30 15 23 27 19 10 16 11 3

75 88 54 7 2 0 3 3 6 3 4 7 6 6 4 6 1 4 2 1

10 73 46 0 3 0 0 0 1 0 0 1 0 2 0 0 2 0 0 0

0 12 14 3 1 0 0 0 0 0 0 0 3 1 0 0 0 0 0 0

0 0 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

451

Table 11 (continued) Period (Tz) 2

3

Portugal 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5

5

4

5

6

7

8

9

10

11

12

13

135 174 24

297 594 603 142 8

252 506 624 433 203 52 7

110 329 414 341 236 165 35 10

25 177 269 212 160 135 83 40 10 2

0 107 203 219 183 98 83 41 17 16 3 1

0 47 88 101 94 101 75 26 25 29 16 1

0 0 32 47 45 40 62 30 8 3 9 2

0 0 1 1 21 13 13 5 3 4 1 1 6 3

0 0 0 0 0 1 1 0 0 4 8 1 1 2

The monthly energy output produced by the Pelamis device varied considerably throughout the year for each location, as is evident by the trendline graph in Fig. 4. Supply in winter was approximately 3–7 times that of the summer supply. It is interesting to observe the minimum monthly output in the Irish site was in June/July, while in the remainder of the locations was in August/ September. The monthly AEO profile of the Irish site matches the WEI closely except on the winter months (Fig. 5). The capture width was maximum in the summer months, with an apparent width of 7–8, dropping to 5 in the winter months. The capture width results suggest that the Pelamis device is operating more optimally during the summer months and is not using all the available energy during the winter months. This could be due to either to inefficiency of design, or extreme weather conditions necessitating device shutdown (i.e. Pelamis power matrix has a zero power output over a certain height and period). 40%

Annual energy output (GWh)

2.50

()

0.4

0.35

R2 = 0.84

0.3

R2 = 0.81 0.25

R2 = 0.71

0.2

R2 = 0.72 0.15

R2 = 0.67

30%

Capture Width

25%

1.50

20% 15%

1.00

R2 = 0.47

0.05 0

Jan

Mar

May

Jul

Sep

Nov

Poly. (Ireland )

10%

0.50

0.45

0.1

Capacity % 2.00

35%

Capacity (%)

Annual energy output (GWh)

3.00

Electricity demand or load in Ireland varied by only 12% between summer and winter months (2005 data shown in the graph) [35]. The correlation between the AEO and the Irish load was only 0.76. This variation in supply especially for the summer

Monthly energy output (GWH)

1 Heights (Hs)

(6)

(7)

(7)

(6)

(6)

(7)

5%

Poly. (USA WC) Poly. (Canada- EC) Poly. (Portugal)

0.00

EC SA

Poly. (Canada- WC) Poly. (USA- EC)

U

C an W ad C a

ga l Po rtu

C an EC ada

W C SA U

Ire la M nd 4 -

0%

Fig. 3. Annual energy output (AEO) (in GWh) for 6 case studies and their respective capacity and capture width (represented as a number in the energy output bar).

Fig. 4. Monthly energy output trendlines in GWh for the case study locations. 3rd order polynomials were used to calculate the trendlines, except for USA EC where 4th order polynomial was used.

452

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

Irish Load

Energy output

Energy input

1.6

2500 0.6 2000 0.5 1500

0.4 (5)

(5)

1000 0.2

(7)

(6)

(8) (7) (6) (8)

0.1

Monthly Irish load (GWh)

Monthly energy output (GWh)

0.7

500

COE (€/kWh)

3000

0.8

(5)

Ireland USA WC Canada EC

1.8

( ) Capture Width

0.3

Mid-2008 costs

2

Portugal Canada WC USA EC

1.4 1.2 1 0.8 0.6 0.4 0.2 0 1

5

10

20

30

40

50

100 200 500 1000 2000

Number of WEC Fig. 7. Impact on COE for 1–2000 WEC based on mid 2008 prices (zero tariff rate used) for various WEC numbers.

(7) (7)

(8) 0

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

months will have worrying implication for base load supply if WEC power was to be used on a large scale. A healthy energy mix will require a more even energy output profile, and will perhaps need to be supplement by other renewables such as solar which have an inverse annual profile. It also has worrying implications for business which already has the perception that renewable energy does not have the capacity to supply power required or be reliable [6].

6.3. COE, NPC and IRR The COE of Pelamis projects in the various global case study locations is presented in Fig. 6, based on 2004 initial capital cost (IC) of the Pelamis [18] and a zero tariff rate (projected 2008 costs are assessed later). Results showed that COE for 1 WEC varied from a low of V0.16/kWh in the Irish site, up to V0.62/kWh on the USA EC. COE dropped in all cases by approximately 50% when 5 WEC were assessed, with a low of V0.09/kWh for the Irish site. COE savings arose mainly due to the same cable size used for both 1 and

mid 2008 prices

120 100 80 60 40 20 0 -20 -40 -60 -80 -100 -120 -140 -160 -180 -200

1 WEC

20 WEC

5 WEC

50 WEC

10 WEC

100 WEC

COE

NPV

COE

€0.10/kWh tariff

NPV

€0.20/kWh tariff

Fig. 8. Analysis of COE and NPV for various tariff rates and WEC number, 2008 prices. Case study location investigated was Ireland. (Note: COE is in Euro cent, while NPV is in millions of Euro).

2004 costs

2

Mid-2008 costs Portugal Canada WC USA EC

Ireland USA WC Canada EC

1.8 1.6

15

1.4

Portugal Canada WC USA EC

Ireland USA WC Canada EC

10 5

1.2

0

1

IRR (%)

COE (€/kWh)

COE (€c/kWh), NPV (€M)

Fig. 5. Graph of monthly energy input and monthly energy output produced by the Pelamis (2007–2008) and monthly load for Ireland in 2005.

0.8 0.6

1

5

10

20

30

40

50

100 200 500 1000 2000

-5 -10

0.4

-15

0.2 0 1

5

10

20

30

40

50

100 200 500 1000 2000

Number of WEC

-20 -25

Number of WEC Fig. 6. COE results for 1–2000 WEC, based on 2004 IC costs from EPRI [18] and a zero tariff rate.

Fig. 9. IRR for various tariff rates at peak 2008 prices and V0.20/kWh tariff rate.

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

453

Table 12 Table list COE results from three papers, which used multiple WEC numbers and/or tariff rates. The same WEC number, tariff rates and IC (based on year) were used as input in the NAVITAS model, with wave energy data from a similar location, and results compared. Report

Previsic [18] Allan [24] Dunnet et al. [20]

Country

California Scotland Canada

Report findings

NAVITAS simulation

WEC number

Tariff quoted in report

COE quoted in report (V/kWh)

Data Buoy used for NAVITAS

Year used for IC

COE calculated using NAVITAS (V/kWh)

213 WEC 4000 WEC 27 WEC

V0.06/kWh V0.05/kWh None

V0.08/kWh V0.10/kWh V0.30/kWh

California 46612 Ireland M4 Canada EC C44137

2004 2008 2008

V0.07/kWh V0.13/kWh V0.34/kWh

5 WEC as well as scaling of capital costs for multiple WEC. COE tailed off for larger WEC numbers, with values for 100 WEC ranging in Ireland at V0.05/kWh to V0.20/kWh in the USA EC. It is important to remember that there was no tariff paid for COE results just presented. COE value is often mistakenly thought of as the value when profit occurs. A COE of zero is the tariff where the project only pays for itself. NPV and IRR are preferred when discussing profitability, and will be discussed latter. 6.4. Sensitivity analysis Results so far have been based on 2004 costs for Pelamis, sourced from Previsic [18]. Results presented in Fig. 7 show the impact of a threefold increase in steel prices as well as a two fold increase in cable costs on, which are approximately equal to prices at their peak in mid 2008, before global recession began. Impact on COE was greatest for small number of WEC, with COE for 1 WEC in Ireland at V0.40, ranging to V1.60/kWh for USA EC. The range of COE values also increased measurably for WEC arrays over 100 in number, with the range of COE for Ireland and USA EC for 100 devices at V0.15 and 0.58/kWh respectively. As already mentioned, COE is not an indication of profit, and is calculated using a zero tariff rate. Most WEC projects will be supported by an FIT, with rates varying from country to country. Results presented in Fig. 8 which assessed the Irish returns, demonstrate that although attractive COE of V0.10/kWh and below resulted for over 5 WEC with a tariff of V0.10/kWh, a negative NPV was returned, which becomes more negative as WEC number increases. A tariff rate of V0.20 is required to return a positive NPC, which almost exponentially increases over 50 WEC. In conclusion, COE gives an idea of the competitiveness of a project in comparison to other project, but gives no indication of NPV return, which is of prime importance for investors. Negative COE is required for positive NPV values when using FIT. Large numbers of WEC are required to move NPV into the positive at a tariff of 0.20/kWh. IRR is often used as the preferred indicator for financial comparison between projects. For long-term projects such as wave energy projects, only returns above the market borrowing rate of 10% are considered viable. It is observed from Fig. 9 that none of the case study locations returned an IRR above 10%, indicating that none of the sites are economically viable, despite some sites returning attractive COE and NPV. The poor IRR reflect the fact that relatively low returns resulted from the high initial costs incurred. 6.5. Comparison to other reports This section reviews three reports which quoted COE for Pelamis, and examines their WEC number and any tariff rates used and the year in which the study was made. It is observed that low COE were reported for large numbers of WEC in the Californian and Scottish studies [18,24], while the Canadian study, which did not use a tariff, was much higher [20]. Modelling was carried out on the

same WEC number and tariffs for each location, using wave energy data from data buoys in similar locations and the same year for IC. COE results from the NAVITAS matched COE from the reports closely, providing validation of the NAVITAS model (Table 12).

7. Summary and conclusion The performance and economic viability of the Pelamis WEC has been investigated using 2007 wave energy data from various global locations. For this purpose, an Excel based model, NAVITAS, was created which estimated the annual energy output of the Pelamis for each location, and produced financial results dependent on various input parameters. The annual wave energy output (AEO) and capacity the Pelamis was highest from the Irish location followed by USA WC, Canada EC, Portugal, Canada WC and finally USA WC. The AEO results correspond well with widely published wave energy data given in kW/m [22,36–38]. AEO and capture widths varied substantially between winter and summer seasons, particularly in the Irish location. The results imply that although the Pelamis is operating at a better rating during the winter months, much of the available energy is being wasted. This could be either due to inefficiency in design or that the site location was too stormy necessitating shutdown of the Pelamis during those periods (i.e. many sea states recorded during the winter months had corresponding zero in the Pelamis power matrix). The large annual variation in output also did not correlate well with the Irish annual load, implying that other renewables or storage will be necessary to cater for the summer months. Economic returns using the energy data showed the same trends as the energy output; i.e. Ireland returns the most positive economic results while USA EC the worst. Results were very sensitive to initial costs (IC) chosen (either 2004 or 2008 in this article) and the number of WEC. Single Pelamis WEC for the various site locations had COE ranging from V0.16/kWh for the Irish site to V0.62/kWh USA EC at 2004 prices. COE increased substantially when using peak 2008 prices, ranging from V0.40 to V1.60/kWh. High sensitivity of economic performance to IC of materials in this report is similar to those experienced in other renewables such as wind and PV [39]. COE would have been higher if lower exchange rates between US dollar and Euro were used in the modelling. In reality, all commercial ventures contain multiple numbers of WEC, and in all the case study sites for WEC number over 100 units, COE under V0.20/kWh were returned for IC based on 2004 prices. High WEC numbers improved COE for two reasons. The most obvious factor is that high WEC numbers benefit from economies of large scale production with cost reductions estimated using learning or production curves. To a lesser extent, but equally important, are savings accrued from cabling, where one cable size can be used for certain multiples of WEC or MW outputs. Thus in this study, the one 110 kV sized cable could be used for over 26 units Pelamis. When using peak 2008 prices for large number of WEC, the only site location that had COE under V0.20 was the Irish site.

454

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455

COE is the most common indicator used to compare economic viability of WEC. However, it does not give an indication of the projects profitability especially when tariffs are being earned from utilities or from state feed-in tariffs (FITs). Moreover, many reports investigating COE include a tariff already in their analysis. Net present value (NPV) and internal rate of return (IRR) are the better indicators. Results from the present study using various FITs for the Irish site, showed that positive NPV returns only occurred where negative COE eventuated, for high WEC number and tariff rates of V0.20/kWh and over. Importantly, negative NPV were returned despite attractive COE of V0.05/kWh to V0.10/kWh resultant from some of the modelled scenarios. Further analysis inspecting IRR (modelled at V0.20/kWh tariff) determined that the few scenarios which returned positive NPV still did not produce IRR that satisfied the 10% IRR threshold. The IRR analysis demonstrated that although good NPV can be returned, the high IC investment required does not necessarily produce a desired IRR return. The results indicated that for Ireland, a tariff of the order of V0.30/kWh will be needed to produce attractive IRR and financial return. The economic analysis of the Pelamis demonstrated that great care needs to be exercised in interpreting the financial results. When comparing COE between various reports, the same cost basis (often related to the year of the report) and WEC number need to be used. COE quotations for single WEC should be avoided as they are commercially unrealistic, and will appear uncompetitive. The learning curve or production curve which calculates the cost reduction for multiple WEC needs to carefully chosen (the technology factor in this article’s case), as this can have substantial influence on costs when dealing with large number of WEC. Successful validation of the present articles results were obtained by comparing the articles results to those of other reports, using the same variable input values used in their respective reports. The comparative results demonstrated that what might appear to be erroneous COE results may be valid when inspecting the criteria by which the figures were estimated. In conclusion, some of the case study locations had high AEO (especially the Irish site) and returned attractive COE. However, the COE results were highly influenced by IC and device numbers. Higher tariffs rates are recommended to produce more viable returns. It is recommended that some standardisation of input variables be agreed in order to assist in confidently comparing COE between reports, particularly with regard WEC number and production curves and year at which costs are taken. Finally, although the majority of reports use COE as the benchmark for financial comparison, it is essential that other indicators such as NPV and IRR be included in the economic analysis (especially where tariff rates are involved), to provide a more thorough estimate of project viability and profitability.

Acknowledgments Appreciation is extended to Florent Thiebaut for creating a Matlab file to process the data obtained from Portugal. Florent is currently doing research work in HMRC.

References [1] European Renewable Energy Council. Renewable energy technology roadmap. EREC, ; 2005. [2] Commission for Energy Regulation. CER announces unprecedented increase in renewable electricity. CER, ; 2008.

[3] Dalton GJ, Lockington DA, Baldock TE. A survey of tourist attitudes to renewable energy supply in Australian hotel accommodation. Renewable Energy 2008;33(10):2174–85. [4] Cle´ment A, McCullen P, Falca˜o A, Fiorentino A, Gardner F, Hammarlund K, et al. Wave energy in Europe: current status and perspectives. Renewable and Sustainable Energy Reviews 2002;6(5):405–31. [5] AEA. Review and analysis of ocean energy systems development and supporting policies. SEI and IEA, ; 2006. [6] Dalton GJ, Lockington DA, Baldock TE. A survey of tourist operator attitudes to renewable energy supply in Queensland, Australia. Renewable Energy 2007;32(4):567–86. [7] Thorpe TW. A review of wave energy. ETSU report R-72, ; 1992. [8] Dalton GJ, Lockington DA, Baldock TE. Feasibility analysis of stand-alone renewable energy supply options for a large hotel. Renewable Energy 2008;33(7):1475–90. [9] Liebman BH. Safeguards, China, and the price of steel. Review of World Economics 2006;142(2):354–73. [10] Raghavan R. Steel prices fall, but no effect on building prices yet. Express India, ; 2008. [11] Huber C, Ryan L, Gallachoir BO, Resch G, Polaski K, Bazilian M. Economic modelling of price support mechanisms for renewable energy: case study on Ireland. Energy Policy 2007;35(2):1172–85. [12] del Rı´o Gonza´lez P. Ten years of renewable electricity policies in Spain: an analysis of successive feed-in tariff reforms. Energy Policy 2008;36(8): 2907–19. [13] Campoccia A, Dusonchet L, Telaretti E, Zizzo G. Comparative analysis of different supporting measures for the production of electrical energy by solar PV and Wind systems: four representative European cases. Solar Energy 2009;83(3):287–97. [14] Adena. Feed-in tariff for wind energy in Portugal. Agencia para a Energia, ; 2008. [15] Pelamis. Ocean energy, ; 2008. [16] Pelamis Carcas M. The Pelamis wave energy converter. Ocean Power Delivery Ltd, ; 2007. [17] Holmes B, Barrett S. State of the art analysis. HMRC internal report for Waveplam, ; 2009. [18] Previsic M. System level design, performance, and costs of California Pelamis wave power plant. EPRI, ; 2004. [19] Bedard R. EPRI ocean energy program, possibilities in California. EPRI, ; 2006. [20] Dunnett D, Wallace JS. Electricity generation from wave power in Canada. Renewable Energy 2009;34(1):179–95. [21] St. Germain LA. A case study of wave power integration into the Ucluelet area electrical grid. Canada: Department of Mechanical Engineering, University of Victoria, ; 2005. [22] ESBI. Accessible wave energy resource atlas. ESBI Report 4D404A-R2 for Marine Institute (MI) and Sustainable Energy Ireland (SEI), ; 2005. [23] Bacon P. Analysis of the potential economic benefits of developing ocean energy in Ireland. Dublin: Report Prepared for Marine Institute (MI) and Sustainable Energy Ireland (SEI) by Bacon Associates, ; 2005. [24] Allan GJ, Bryden I, McGregor PG, Stallard T, Kim Swales J, Turner K, et al. Concurrent and legacy economic and environmental impacts from establishing a marine energy sector in Scotland. Energy Policy 2008;36(7):2734– 53. [25] Carbon Trust, Callaghan J. CTC601-future marine energy. Carbon Trust, ; 2006. [26] Microsoft. Excel. Washington, US: Software Produced by Microsoft; 2006. [27] RETScreen. Varennes, Canada: Software Produced by Natural Resources, CANMET Energy Technology Centre (CETC), ; 2006. [28] Mathworks. Matlab, ; 2009. Cambridge, UK. [29] ESBI. Standard pricing approach for connecting renewable generators to the distribution network. Ireland: CER, ; 2005. [30] Marine Institute, ESBI. Wave energy network connection project; 2008. [31] Hau E. Wind turbines: fundamentals, technologies, application, economics. Berlin, New York: Springer; 2006. p. 783. [32] Lambert T, Lilienthal P. HOMER: the micro-power optimisation model. Software Produced by NREL; 2004. and .

G.J. Dalton et al. / Renewable Energy 35 (2010) 443–455 [33] Sijm JPM. The performance of feed-in tariffs to promote renewable electricity in European countries. ECN, ; 2002. [34] Khatib H. Economic evaluation of projects in the electricity supply industry. In: Institution of Electrical Engineers (IEE) power and energy series, vol. 44; 2003. London, UK. [35] ESB. Annual review of the Irish electricity market ESB National Grid, ; 2005.

455

[36] Australian Government. Renewable energy atlas, ; 2008. [37] Eurowaves. OCEANOR, ; 2008. [38] Pontes MT. World wave atlas OCEANOR, ; 2006. [39] Dalton GJ, Lockington DA, Baldock TE. Case study feasibility analysis of renewable energy supply options for small to medium-sized tourist accommodations. Renewable Energy 2009;34(4):1134–44.